No Carrier @ 56K Connection (x2)

telephone line appears to be clean. site is w/in 2km from telephone exchange.

Are you sure that the answering modem is able to handle 56K?

ISP w/ USR server advertise x2 capability

lyndon1,
Go here for info, and try the USR linetest:

http://www.3com.com/56k/need4_56k/linetest.html

This should tell you if your modem & line are working properly.
Successfull 56k connection here may indicate an ISP problem.
Let me know if you need more.
Regards,
Ralph

i have to pay toll charges for the line test but i'll try it anyway.

to enable x2, the S32 register should be 32. but reading is always s32=0. writing ats32=32 thru hyperterminal returns an error. i tried at+ms=vx2,1,300,0 as adviced by Cirrus Logic but no improvement either.

Are you escaping into modem's command mode in hyperterminal? (+++)
Ralph

Don't forget to use 'AT&F1' before your 'ATDTnnn-nnn-nnnn' command,
in order to correctly-initialize an X2 modem.

i see very few cases of escape sequence in the modemlog file but not in hyperterminal window.

regarding the line test, from the philippines, my call might be routed via satellite.this might affect test results.

I tend to forget that not everyone here is in the US.  Sorry, I hadn't considered where you might be calling from.  
I'm sure that hop will make a substantial difference in the "line quality".
Is there another x2 equipped ISP that you can try?
You don't have to login, just try it to see if there is a carrier and connect.
Please reject the answer proposed, since it has not solved your problem.
Ralph

Since the satellite-link will be "digital",
it should _NOT_ affect line-quality.
A one-minute "international" call should not be _TOO_ expensive,
especially if the money it costs helps you to solve your problem.

will transmission delay inherent in satellite communication not affect the line test results?

Since the dealy is only felt as a delay if you are looking at both sides at once, there should be no difference since everything is delayed by the same amount. In other words, the first bit might be delayed by one second, but the timing on every bit there after is referenced to the bit before it for timing so you would not notice the delay unless you had a reference signal from the far end that was not delayed to compare it to.  Does this make sense?  There is a delay involved in every signal, even those on your motherboard, but in the case of the motherboard delays can be crippling because all signals on the motherboard have the same reference. A good example of how delays may become transparent, try watching TV on cable, and compare the same live signal off of a Satt Dish. You would  not know they were not in sync unless you could compare one end to the other.

i understand echo suppression and compression techniques are applied in satellite circuits. this might affect line test results same as analog- to- digital devices  found in most subscriber lines.

anyway for my modem problem, i have tried the ff:

1. added init string based on the chipset manual at+ms=vx2,1,300,0
2. dial 2 other isp's w/ x2 capability
3. remove telephone set

results:
1.connect@56k,
2.hear continuous tone for few seconds
3.hang-up (no carrier)

what is that continuous tone-carrier signal?

> what is that continuous tone-carrier signal?

I've experienced something like this;
it was a digital "busy" signal from
the ISP's modem-server, after it
"answered" the modem's "call" to it.

i hear a touch tone dialling, then a continuous single tone for a few seconds. after that modem hangs up - no carrier! what is that single continuous tone i hear? is that the carrier signal the modem cannot detect after establishing a 56k connection?

if i don't hear that continuous single tone, i'm connected @33.6k without a problem.

When my USR V.90 modem connects,
I hear a steady-tone, for about 2 seconds,
when the ISP's modem "answers" the call,
and then some hisses, and some pitch-changes (high,low,high),
and then a fading-out tone,
Then, the 'CONNECT 50666' message is produced.
The total elapsed-time is over 5 seconds.
Does this match your experiences??

i hear the monotonous tone after the CONNECT 56000 message. it sounded like a flat line tone of ECG machine and last for about 15 seconds.

i hear the "live" tones during dialling & handshaking. i am interested in the "dead" tone that comes after that.

Although I don't quite understand how your modem can at first detect a carrier, dial and then loose the carrier, I've also had problems detecting a carrier signal whith my 56K modem. What finally solved the problem was downloading a Flash ROM Update from the homepage of my modem manufacturer. You might also be able to upgrade your modem to K56 Flex or V.90 which would, at least that's what I've heard, slightly increase the modem's speed.

simondittami, thank you for sharing your experience. but i am reopening the question based on the case below.

my 56k internal modem w/ CL-MD5650 chipset came under Web Excel drivers and x2 firmware. going to the homepage for v.90 upgrade shows only Rockwell chipset - NO Cirrus Logic.

however, in another manufacturer's homepage, Jaton uses the same Cirrus Logic chipset. the file size of the driver and x2 firmware is the same as Web Excel. can i upgrade tov.90 using Jaton's firmware?

NO- see Q.10068364 (x2 firmware)

how do you explain the unusual tone generated? question increase for 100pts.

any expert on CL-MD5650 chipset? points increase to150

If I read correctly your modem is an X2 only, w/o the V90 upgrade.  Try these strings:
S32=66 or S58=32..I'm not sure which might work since your modem is not a USR X2.
This will try to force X2.  
Judging from the tones you are saying, your ISP's modem is trying to force V90 but your modem can't interprete.  So this should make it more clearer for you modem.

Do you connect correctly to the USR "line-test" server?
Since your modem uses X2, and the X2 technology has been
licensed only from USR, you should try dialing to this server.

Do the sounds emitted by your modem during the "connect"
process match the sounds when connecting to your local ISP?

Does your ISP have any advice?

Drag2-thanks for the advice but i have to reopen the question again.

for CL-MD5650 chipset, s32=32 should enable x2 as commented previously. i tried s32=66 and s58=32 as you instructed. they also gave me an ERROR message.


Otta-nice to hear from you again. if you recall previous comments, line test via satellite is not acceptable.

my ISP does not have line test capability and has not even replied for the list of compatible modems i requested. ISP homepage is www.pworld.net.ph

the sound generated by my modem during the connection process is the same as other modems.

How long do you intend to drag this out?  If your modem came with a manual, you should have the correct commands to try. If the modem is not working correctly and is under a warranty, send it back. You have been given what i think is the correct answer many times. A 56k connection at top line performance is about 52k. When you do not have top performance copper connections your modem will pull back its speed until a reliable connection can be made. Your statement is consistent with these claims, your modem trys at 56k, then pulls back to 33.6. This is normal. How many other experts reading this will agree with this?

>  line test via satellite is not acceptable.

Did you _TRY_ connecting to the USR "line-test"?
If so, please report _EXACTLY_ what you observed.
Did you get connected? If so,
what 'CONNECT xxxxx/yyy/zzz' message was produced?
What was the output from the "line-test"?

Please tell us the telephone-number for the X2 dial-up
port for your ISP.  Then, we can try, with our modem(s),
to get a 'CONNECT xxxxx/yyy/zzz' message.

P.S. After hanging-up from your ISP,
use 'ATI6' and 'ATI7' commands,
and your modem will "report" on the status
of the most-recent connection,
including "speed" of that connection.
I'd like to see that output.

joe_m - your views are well noted. if you're satisfied w/ the answers or felt that this exchange of info is not worth your time, its your choice. but i'm still learning something from this process and i'm sure others too.

one point of clarification, my modem do CONNECT @ 56000 even for just 15 sec. and then hang-up. not fall back to 33.6k as you perceived. don't you think its odd since actual line speed achievable is just 52K due to power limitation?

in most cases, it also negotiates and fall back to 33.6k connections. the occurence is 40% for 56k and 60% for 33.6k.

perhaps you should review question history to get a clearer picture.

joe_m - your views are well noted. if you're satisfied w/ the answers or felt that this exchange of info is not worth your time, its your choice. but i'm still learning something from this process and i'm sure others too.

one point of clarification, my modem do CONNECT @ 56000 even for just 15 sec. and then hang-up. not fall back to 33.6k as you perceived. don't you think its odd since actual line speed achievable is just 52K due to power limitation?

in most cases, it also negotiates and fall back to 33.6k connections. the occurence is 40% for 56k and 60% for 33.6k.

perhaps you should review question history to get a clearer picture.

joe_m - your views are well noted. if you're satisfied w/ the answers or felt that this exchange of info is not worth your time, its your choice. but i'm still learning something from this process and i'm sure others too.

one point of clarification, my modem do CONNECT @ 56000 even for just 15 sec. and then hang-up. not fall back to 33.6k as you perceived. don't you think its odd since actual line speed achievable is just 52K due to power limitation?

in most cases, it also negotiates and falls back to 33.6k connections. the occurence is 40% for 56k and 60% for 33.6k.

perhaps you should review question history to get a clearer picture.

my apologies on repetitive comments above. got problem when server hung-up.

Otta- perhaps i may reconsider the usr line test as last option. i might be routed via the trans Pacific fiber optic cables which has 64k bandwith per circuit. or simply test another x2 modem to prove the line is 56k capable.

meantime, here's the data you asked:

ATZ
OK
AT&F&C1
OK
ATW2
OK
ATDT4354174
CONNECT 56000
                     (15 SEC CONTINUOUS TONE)
NO CARRIER
                      (MODEM HANG-UP)
ATI6
US
OK
ATI7
CIRRUS LOGIC,INC. MODEM ENGINEERING FIRMWARE DEPARTMENT

again, info on isp is at www.pworld.net.ph

if ati6 or ati7 is entered before the modem hangs up (w/in 15 sec) i get this message:

PROTOCOL:LAPM
COMPRESSION:V42B
CONNECT 56000

i'm increasing question to 200 points

Can you use an 'AT' command, peculiar to your modem,
to "limit" your connection-speed to a lower value,
say '50666' or '49333' or '48000', and try that?
The "15-second-squawk" could be a "mis-negotiation",
i.e., the two modems connected at 56000,
but were unable to "sustain" that rate.

Have you ever once called your ISP to make sure you're dialing up to a 56k line?  Many ISPs offer thier 56k services on seperate dial-up numbers.  I would call them and ask if there is a special number you should be using.

ESCHATAR,
Did you read LYNDON1's previous comment:
> my modem do CONNECT @ 56000 even for just 15 sec.
> and then hang-up. not fall back to 33.6k
> as you perceived. don't you think its odd
> since actual line speed achievable is just 52K
> due to power limitation?

This raises a question.  The "limit" of 52000 is a regulation
in the USA, even though the USR Total Control Server
is capable of 56000.  So, I wonder if there is some
"configuration" which your ISP has accidentally/intentionally
omitted, in order to restrict the Server to obey USA
regulations, even though you are not in the USA.

Omitting the "restriction" could get you a 56000 connection,
but also could create the cross-talk problems
which the regulation is intended to prevent.

So, limit your modem to, say, 50666, and see what happens.

at+ms=vx2,1,300,0 usually results 2 outcomes:

1. CONNECT 56000
   NO CARRIER          (PROBLEM)

2. CONNECT 33600       (NO PROBLEM)

forcing the DCE speed to 52000,50666,49333,48000 i tried
at+ms=vx2,1,52000,52000   etc.

results: 2 outcomes

1. CONNECT 33600
   NO CARRIER
2. CONNECT 33333

w/ win95 dial-up i can connect @ 33.6k. how come i'm now limited @33333 after several tests?

Be sure to "factory-reset" your modem,
i.e., something like 'ATZ' or 'AT&F' or 'AT&F1',
before running _each_ of the tests.

Enter the 'ATI5' and 'ATI6' commands after each test,
so that the modem can report more details about each call.

i did several tests using  AT&F&C1W2+MS=VX2,1,48000,48000
                                           ATDT4354174

results are the same:

1. CONNECT 33600                                            2. CONNECT 33333
    LOGIN:                                                                    LOGIN:
    NO CARRIER

1. ati6 after no carrier only responded "US"

2.a. entering login and password @ 33333 responded w/ "PPP session start.....(garbage        characters) NO CARRIER"

2.b. trying escape sequence +++ at login then ati6 responded w/ CONNECT 33333 or        CONNECT 33600 (increased speed)

is the message NO CARRIER the problem here?

Lyndon1,

You say the ISP advertises x2 capability. Have you contacted them to verify that the dial-in number you are using is correct? Some ISPs have separate numbers for 56k and 33.6/less.

Seems this thread and all these experts have you jumping through hoops and no one has suggested starting with the other end of the line.

Colleen

> LOGIN: LOGIN:
> NO CARRIER

This is perfectly normal.

Your ISP will automatically "drop" the connection
if you do not "login" within a few seconds.
When they "hang-up", your modem detects this,
and displays 'NO CARRIER'.

> entering login and password @ 33333 responded w/ "PPP
> session start.....(garbage characters) NO CARRIER"

This is perfectly normal.
Those "garbage" characters are sent by your ISP;
you need to run PPP in order to "decode" them.
Since you are not running PPP, the ISP notices
that you are not responding to those encoded
messages.  So, it decides to "hang-up",
and your modem issues 'NO CARRIER'.

Have you installed the PPP software on your computer?

colleen, if you check my ISP @ www.pworld.net.ph, there's a separate number for x2, v90/56kf,v34.

otta, i don't know about ppp software but that's the result where i don't have a problem. meaning i can connect.

using w3, i tried this string: at&f&c1w3+ms=vx2,1,300,48000
                                               atdt4354174

results:
1. CONNECT 115200/V34/LAPM/V42B/TX=36000/RX=33600
    LOGIN:
    NO CARRIER

2. CONNECT 115200/X2/LAPM/V42B/TX=28000/RX=48000
    NO CARRIER

in result #2, notice no login and connection @ 48000. there is a continuous tone here.

Then we can assume that while you're attempting to connect, you're using the 56k number...

You might find some answers concerning your particular modem at http://www.56k.com/  They seem to have informtion about issues with specific modems, etc.

I don't have all the answers, but I know where to find them.

Colleen

colleen, thanks but i'd been there already and perhaps most of the links it points. maybe there is a design flaw in CL-MD5650 modem chipset. cirrus logic website has good documentation. i have the impression they're still revising.

I have a Com56k x2/v90 modem made by Jaton with the exact same chipset, I hate it, I put "W2" in the advance properties of my modem dialog, I'm able to get up to 44K and thats it.
I called Jaton, they were useless, I tried a Different chipset modem, I think it's "Tex Inst" can't remember for sure, anyway I can Connect at 53K 90% of the time, So I got a hold of Cirus Logic rep, and they basically told me that there is a problem with that chipset.  Have you tried to bump it up to V90 yet, they said that it should correct it, My Isp only supports X2 at the current time, so I havn't yet...  

i already included w2 command in the previous tests. it just report the DCE rate.

i have downloaded v90 from jaton. against experts' advice, do you think i can flash this firmware upgrade without damaging my modem? please refer to my previous comment to simondittami regarding this.

did you get a technical overview of the problem from cirrus logic?

> 1. CONNECT 115200/V34/LAPM/V42B/TX=36000/RX=33600
>   LOGIN:
>   NO CARRIER

This is perfectly normal;
if you do not respond to the 'LOGIN:' prompt,
then, about 30 seconds later,
the ISP will "hang-up" their modem.

Use the '+++' command,
as soon as you see the 'LOGIN:' prompt,
and then the 'ATI6' and/or 'ATI7' commands,
to see the "actual" results.

Then, use 'ATO' to return to "online" mode.

otta, i don't think i have a problem w/ result #1. typograhical error tx=36000 should be tx=33600. anyway i already tried that. please see previous comments.

it's result #2 where the problem is.

I would either take another comptuer with currently correctly configured X2 connection to the internet and try it from your location and if this is not an option then try your current configuration in another location.  Also the other suggestion that I have is to purchase a inexpensive K56FLEX modem.  Most ISPs can support an K56 connection a lot easier becuase the X2 server equipment have to be X2 only. K56 server equpiment is backward compablible with 33.6 speed modems. Let me know if you can perform any of these tests.

michaelham, thanks for the advice but i just got  2 users of same modem having similar problems. i'll just experiment by flashing the v90 firmware from jaton to my modem.

raider187 or any expert, can you confirm design flaw of CL-MD5650 modem chipset and give a technical overview of the problem?

i would like that as a proposed answer

It's obvious your ISP can not communicate at 56k.  The tones you hear are hand shake tones.  If the two modems can not agree on speed and compression mode the receiving modem disconects.

This is what the Jaton tech support page says about connecting at high speeds with their modems.


http:www.jaton.com/web/support2.htm

Q: Tips for high speed connections using a V.34 or faster modem.

Explain and Suggestion:
Very few people can reach a consistent 33,600 or 56K(x2 modem) bps connection. Speeds of 28,800 bps or faster, require perfect, almost ISDN quality, line conditions along the entire length of the connection. However, V.34 and 33.6k modems are capable of pushing the limits of analog phone lines, commonly offering connect speeds of 21600,24000, and even 26400 bps or higher.

Variation in line quality are typically the cause of low connection rate. At one point or another, everyone will experience " a bad line" connection, and have to hang up and call again. However, if you find that you never or rarely connect at rates above 19200 bps, you will need to investigate the line quality of your connections.

If you encounter the same low connection rate, the problem may be resulting from impairments along the lines running to the local telephone company or within your home or office. Your telephone company or a private consultant may be able to help.

This is what the Jaton tech support page says about connecting at high speeds with their modems.


http:www.jaton.com/web/support2.htm

Q: Tips for high speed connections using a V.34 or faster modem.

Explain and Suggestion:
Very few people can reach a consistent 33,600 or 56K(x2 modem) bps connection. Speeds of 28,800 bps or faster, require perfect, almost ISDN quality, line conditions along the entire length of the connection. However, V.34 and 33.6k modems are capable of pushing the limits of analog phone lines, commonly offering connect speeds of 21600,24000, and even 26400 bps or higher.

Variation in line quality are typically the cause of low connection rate. At one point or another, everyone will experience " a bad line" connection, and have to hang up and call again. However, if you find that you never or rarely connect at rates above 19200 bps, you will need to investigate the line quality of your connections.

If you encounter the same low connection rate, the problem may be resulting from impairments along the lines running to the local telephone company or within your home or office. Your telephone company or a private consultant may be able to help.

Also, I researched the chipset and found no reference to anything being wrong with it.  I used IC Master, and the web for this.

MICHAELHAM wrote:
> Most ISPs can support an K56 connection a lot easier
> because the X2 server equipment have to be X2 only.

Incorrect -- "older" modems (even 14.4Kb) can connect to X2 servers.

ruble, sorry your answer is not acceptable.

please review the history questions. you'll find description of dialling and handshaking tones as well as the continuous tone after 56k connection is established.

joe_m, i appreciate some research you did.

my modem statistics show about 40% for 56k and 60% for 33.6k connections (31k negligible). not bad even though i still have a problem with the 56k connection.

if the problem is not in the hardware, could it be the firmware?

cirrus logic is quite evasive with my inquiries (i.e., cannot write s32=32 to enable x2, no carrier message, etc.). i was just adviced to use the fcc number to trace the manufacturer even though the ati7 and at+gmi both points to cirrus logic.

i still believe there's something wrong with it. unless you can prove otherwise...

The only sure file way I know of to prove this one way or another would be for you to borrow another 56k V90 X2 modem, and see if yields the same result from the same location.  I wish there was more to tell you, but as I said in my first post, you are experiencing the same results 1000's of people experience with 56k modems, they do not get 56k, (52k in US). There is much published on this, but you keep thinking that you must have a faulty modem.  So the best and cheapest way to determine your problem, or narrow the possibilities down is to do a comparison. The person who loans you the 56k modem should be able to tell, and show you his connect speed.
Here in the US, you can buy a USR 56k modem and if it doesn't perform to your expectations, they will take it back. I've done that, they took it back with no questions asked.

joe_m, i don't think it's just a coincidence that 2 more users of the same modem in different locations have the similar peculiar problem as mine.

do the 1000 users in US also experience the unusual tone emitted once the temporary 56k connection is established?

in our country, maybe we could not fully monitor the products that's being dumped here.

i just want to clear the facts.

I agree with "joe_massimino" -- obtain a different X2 or V.90 modem, and try it.
The "Cardinal Connecta X2 internal modem" has identical hardware
to the USR Robotics V.90 internal Sportster, but costs much less,
i.e., I bought one for $47 (US), and upgraded it,
via Cardinal's web-site to V.90, for "free"
(the telephone-company billed me for the long-distance call
to Cardinal's office in Georgia (USA).

As a matter of fact, when my 56k modem connects I do hear an unusual tone very unlike any tone I've ever heard from my 33.6 modem on my other system. The usual connect from the 56k modem is 50.6 at work, and when I use a 33.6 I can get 28.8. On the other system I use at home, I get 26.4 no matter what modem I use.
I'm sure that the tones you hear are the same odd tones that I hear on my USR modem, I never gave it a thought because the connect was what I expected, and 56k is a different protocol, so the hand shaking is bound to sound odd compared to what I usually hear on a 33.6 or 28.8. I would say that it sounded a bit musical, but brief, unlike any other modem sounds before 56k. Even the 33.6 modem sounds different than a 14.4 modem, and so does 2400 baud sound different than all the rest. This all makes sense, the hand shaking of each of these modems is different so it should sound different since they work with audio frequencies. Maybe I could send you a sound clip from my modem at work, would that be helpful?

joe_massimino comments on different "sounds".
When I upgraded my US Robotics Sportster
from X2 to V.90, the "negotiation" sounds changed.
When I borrowed a Cardinal Connecta X2 modem,
it produced the same sounds as the USR X2 modem.
After upgrading the Cardinal to V.90,
it produces the same sounds as the USR V.90 modem.

However, I still can't dance to any of the tunes.  :-)

I have the sound clip of my 56k modem connecting. Where can I send you an E-mail attachment?

When your modem tries to connect, you hear the normal training tones. Then you see the message connected 56000. It does not mean that you are actually connected. Your modem passed the negotiation, and thinks the host can communicate at that speed. When you hear a continuous tone from the host, it means it is not connected at all. It is either trying to renegotiate the connection, or is not accepting logins. When you are really connected, you should hear a hissing sound corresponding to the efficient modulation scheme.
To conduct a better test, you need to dial to another POP number. Since you now have v.90 firmware, why not try the other number?

I have posted this before in the win95 area.... I have the same problems you are talking about when I try to dial up my x2 service I have a courier v.everything with true v.90 external.. the problem you are talking about is common on alot of phone lines....
QUESTION.... DID IT EVER WORK FOR YOU AND NOW DOESN'T ??????

If so here might be the answer.......
I did alot of checking with USR and my phone company after doing some extensive line quality testing with my provider I kept getting a result of to many codecs in line ..... what this means acording to USR is there are to many digital switches between me and my provider ( the line quality has to be premo for X2 to work correctly....) according to USR this is a cheap way out for phone companies to set up digital phone lines.... the only option that I had after fighting with the phone company was to find a true 56Kflex server now all works fine getting constant 42,000 connects..... this will only work though if you have true v.90 to do the challenge hand shaking tones.....

I hope this is an answer for you.... or your modem could be shot...

dmckee1, at least you're right in one thing- my modem should be shot!

i borrowed a USR external sporter x2 56k and been getting consistent 48000 connections. so my line is definitely not the problem.

public, i guess i cannot connect @x2 modulation. the fallback is always v34 where i don't have a problem.

48000 is very nice, I wish I could get that at home.  Did you read the other comments I posted? Are you ready to except one of the above answers? You appear to have a poor performing modem, and when you got a good one, you see that your lines aren't good enough to yield 56k. I'll trade my 26400 line for your 48000 anytime. :)

              Joe

joe_m, as i said before i will accept an expert opinion on CL-MD5650 modem chipset.

if one of the expert before mentioned something is wrong w/ the chipset, i like to hear more about it. perhaps just a technical overview of the problem.

regarding the sound of the modems, more or less, both seem to dance to the same tune up to the hissing part. only difference is, my modem, once a 56k connection is established (?) start emitting the "dead" tone for 15 sec & hang-up.

So, you are saying that you have changed the question and the only acceptable answer to the new question is the report on a flaw in your chipset. So if a flaw doesn't exist, or is not reported you are going to withdraw the question?  Think again, you need to reread you original question and all comments and answers up to this point and figure which one explained your situation best.
So you have learned that not every modem is as good as every other regardless of the specifications.  This is why some modems cost many times more than a Jaton, or Cardinal, or even the cheapest of them all, the one that has no name.
I'm not saying who to award it to, just that if you read the question, you will see that there are a few answers that give explanations, and/or suggestions, that was the question, now pick the best answer. Grade it a D if you feel it was not all that you hoped it would be.

Lyndon1 I need some info so I might help you out...
1- what is your flow control settings ? ( i.e. Hardware or Software )
2- how is your compression set ( please let me know exectaly how it is writen ) beside that it is set to V42B as you mentioned befor.
3- do you use IP header comression ?
4- is your local PBX ( telephon company switch board ) digital or analog ?
5- the same with the ISP's PBX ?
6- and if the ISP located @ another town, the regonal PBXs?

Beside that, just some general info : you NEVER been connected to your ISP @56K !
the continuous sound you heared was your modem TRYING to negotiate with the ISP modem, there was the initial "hand shake" and thenyou got "connected @56K hwile the tow modems were trying to establish connection, since you were trying 56K it took more time then it took you to establish connection @33.6K, but after the preset time for the ISP's modem for establishing connection, it just hung up and the you got "no carrier".
Ofcourse there is the issue of the telephone lines quality, and verry seldom you will get a true 54K to 56K connection and even at the best conditions, with a fully digital PBX telephony system, but you SHOULD get atleast 42K to 43K ( with good lines ) but still not with your modem 9 nor the other 2 identical modems... but you got 48K with a better modem, so we are back to you modem or even the brand itself ( i.e. CL's modem...)

joe_m, if you want to dictate the outcome of the exchange of information with your own set of procedures and protocols, sorry to disappoint you!

as far as i'm concern, the question remains valid and i have no intention of withdrawing it. i'm sure the right expert will come along.

for every proposed answer, i always give a comment which at times is also informative.

ACScomm, here's the data you asked:

1-flow control -hardware
2-compression as seen from w3 response code:
   CONNECT 115200/X2/LAPM/V42B/TX=28000/RX=56000
3-IP header compression checked
4-line is connected to digital switch EWSD
5-ISP connected to another telco w/ digital switch
6-both telcos located in metro-manila interconnected w/ E1 digital trunks

the handshaking tones for v34 & x2 are similar. only the x2 emits additional tone as describe above. why can't i connect @x2?

Lyndon1 :
I hope the following info wil help, and it will ONLY if your ISP DO support x2 technology !

1- make sure that your modem setup enables software compression.
2- make sure that you are working only with TCP/IP and NOT with IP/IPX and NetBEUI
3- make sure that your modem initial string allows you to work with V90, you need to check the modem manual for the EXACT string needed for your modem.
( you can ask your ISP for the string needed aswell. )
Check all the steps mentioned, try to set up a ceonnection once more with your ISP
and let me know the outcome.
( send me the connection log, alow about up to ~15 sec' for connection to be fully established )
If you can find out which kind of RAS your ISP use for dailup connections... brand and model, it will help me to have a better of the problem.
ACScomm

ACSComm, my isp has 2 servers:
1) US Robotics x2 (not v90 compatible)
2) Portmaster 3 k56 flex (v90 compatible)

i think i have tried #1&2 before with same results. regarding #3, no init string for v90 from any web page or isp.

against expert advice, i forced flashed the v90 firmware from jaton w/ interesting results. i no longer connect @ x2 or v90. connection is consistent @v34 (33.6k).

i'll just flash back the original x2 firmware and follow procedure 1&2 to be sure. by the way, what does RAS mean?

Lyndon1 :
RAS = Remote Access Server...
> 1) US Robotics x2 (not v90 compatible)
> 2) Portmaster 3 k56 flex (v90 compatible)
Can you tell me the exact models ?

ACSComm, flashing the original x2 firmware, hereunder is the modemlog w/ control flow set to hardware. Only difference when set to software is data compression & error control are off.

09-09-1998 22:36:31.70 - Cirrus 56K bps Modem Internal (CL-MD5650) Modem in use.
09-09-1998 22:36:31.71 - Modem type: Cirrus 56K bps Modem Internal (CL-MD5650) Modem
09-09-1998 22:36:31.71 - Modem inf path: CIRRUS~1.INF
09-09-1998 22:36:31.71 - Modem inf section: Gen56kInt
09-09-1998 22:36:31.94 - 115200,N,8,1
09-09-1998 22:36:31.94 - 115200,N,8,1
09-09-1998 22:36:31.94 - Initializing modem.
09-09-1998 22:36:31.94 - Send: ATZ<cr>
09-09-1998 22:36:31.95 - Recv: ATZ<cr>
09-09-1998 22:36:32.19 - Recv: <cr>
09-09-1998 22:36:32.19 - Interpreted response: Informative
09-09-1998 22:36:32.19 - Recv: <lf>
09-09-1998 22:36:32.19 - Interpreted response: Informative
09-09-1998 22:36:32.19 - Recv: OK
09-09-1998 22:36:32.19 - Interpreted response: Ok
09-09-1998 22:36:32.20 - Send: AT &F E0 V1 W4 &D2 &C1 S0=0 -C1 <cr>
09-09-1998 22:36:32.20 - Recv: <cr>
09-09-1998 22:36:32.20 - Interpreted response: Informative
09-09-1998 22:36:32.20 - Recv: <lf>
09-09-1998 22:36:32.20 - Interpreted response: Informative
09-09-1998 22:36:32.21 - Recv: AT &F E0 V1 W4 &D2 &C1 S0=0 -C1 <cr>
09-09-1998 22:36:32.50 - Recv: <cr>
09-09-1998 22:36:32.50 - Interpreted response: Informative
09-09-1998 22:36:32.50 - Recv: <lf>
09-09-1998 22:36:32.50 - Interpreted response: Informative
09-09-1998 22:36:32.50 - Recv: OK
09-09-1998 22:36:32.50 - Interpreted response: Ok
09-09-1998 22:36:32.50 - Send: ATS7=60\T0L1M1\N3-J1%C1"H3\Q3B1N1X4<cr>
09-09-1998 22:36:32.50 - Recv: <cr>
09-09-1998 22:36:32.50 - Interpreted response: Informative
09-09-1998 22:36:32.50 - Recv: <lf>
09-09-1998 22:36:32.50 - Interpreted response: Informative
09-09-1998 22:36:32.61 - Recv: <cr>
09-09-1998 22:36:32.61 - Interpreted response: Informative
09-09-1998 22:36:32.61 - Recv: <lf>
09-09-1998 22:36:32.61 - Interpreted response: Informative
09-09-1998 22:36:32.61 - Recv: OK
09-09-1998 22:36:32.61 - Interpreted response: Ok
09-09-1998 22:36:32.61 - Send: AT+MS=VX2,1,300,0<cr>
09-09-1998 22:36:32.61 - Recv: <cr>
09-09-1998 22:36:32.61 - Interpreted response: Informative
09-09-1998 22:36:32.61 - Recv: <lf>
09-09-1998 22:36:32.61 - Interpreted response: Informative
09-09-1998 22:36:32.68 - Recv: <cr>
09-09-1998 22:36:32.68 - Interpreted response: Informative
09-09-1998 22:36:32.68 - Recv: <lf>
09-09-1998 22:36:32.68 - Interpreted response: Informative
09-09-1998 22:36:32.68 - Recv: OK
09-09-1998 22:36:32.68 - Interpreted response: Ok
09-09-1998 22:36:32.69 - Dialing.
09-09-1998 22:36:32.69 - Send: ATDT;<cr>
09-09-1998 22:36:34.12 - Recv: <cr>
09-09-1998 22:36:34.12 - Interpreted response: Informative
09-09-1998 22:36:34.12 - Recv: <lf>
09-09-1998 22:36:34.12 - Interpreted response: Informative
09-09-1998 22:36:34.12 - Recv: OK
09-09-1998 22:36:34.12 - Interpreted response: Ok
09-09-1998 22:36:34.13 - Dialing.
09-09-1998 22:36:34.13 - Send: ATDT#######<cr>
09-09-1998 22:36:57.35 - Recv: <cr>
09-09-1998 22:36:57.35 - Interpreted response: Informative
09-09-1998 22:36:57.35 - Recv: <lf>
09-09-1998 22:36:57.35 - Interpreted response: Informative
09-09-1998 22:36:57.35 - Recv: PROTOCOL:LAPM
09-09-1998 22:36:57.35 - Interpreted response: Informative
09-09-1998 22:36:57.35 - Recv: <cr>
09-09-1998 22:36:57.35 - Interpreted response: Informative
09-09-1998 22:36:57.35 - Recv: <lf>
09-09-1998 22:36:57.35 - Interpreted response: Informative
09-09-1998 22:36:57.35 - Recv: COMPRESSION:V42B
09-09-1998 22:36:57.35 - Interpreted response: Informative
09-09-1998 22:36:57.35 - Recv: <cr>
09-09-1998 22:36:57.35 - Interpreted response: Informative
09-09-1998 22:36:57.35 - Recv: <lf>
09-09-1998 22:36:57.35 - Interpreted response: Informative
09-09-1998 22:36:57.37 - Recv: CONNECT 56000
09-09-1998 22:36:57.37 - Interpreted response: Connect
09-09-1998 22:36:57.37 - Connection established at 56000bps.
09-09-1998 22:36:57.37 - Error-control on.
09-09-1998 22:36:57.37 - Data compression on.
09-09-1998 22:37:19.06 - Remote modem hung up.
09-09-1998 22:37:19.07 - Recv: <cr>
09-09-1998 22:37:19.07 - Interpreted response: Informative
09-09-1998 22:37:19.07 - Recv: <lf>
09-09-1998 22:37:19.07 - Interpreted response: Informative
09-09-1998 22:37:19.07 - Recv: NO CARRIER
09-09-1998 22:37:19.07 - Interpreted response: No Carrier
09-09-1998 22:37:23.20 - Hanging up the modem.
09-09-1998 22:37:23.20 - Send: ATH<cr>
09-09-1998 22:37:23.81 - Recv: <cr>
09-09-1998 22:37:23.81 - Interpreted response: Informative
09-09-1998 22:37:23.81 - Recv: <lf>
09-09-1998 22:37:23.81 - Interpreted response: Informative
09-09-1998 22:37:23.81 - Recv: OK
09-09-1998 22:37:23.81 - Interpreted response: Ok
09-09-1998 22:37:23.81 - 115200,N,8,1
09-09-1998 22:37:23.84 - Session Statistics:
09-09-1998 22:37:23.84 -                Reads : 183 bytes
09-09-1998 22:37:23.84 -                Writes: 126 bytes
09-09-1998 22:37:23.84 - Cirrus 56K bps Modem Internal (CL-MD5650) Modem closed.

hope you can solve the problem. i'll get back for the model of RAS.

The "interesting" bits:
    09-09-1998 22:36:34.13 - Send: ATDT#######<cr>
    09-09-1998 22:36:57.35 - Recv: <cr>
Note the 23.22 seconds between "send" and "receive".
This indicates that the two modems have taken this time
to "negotiate" successfully.

    09-09-1998 22:36:57.37 - Connection established at 56000bps.
    09-09-1998 22:36:57.37 - Error-control on.
    09-09-1998 22:36:57.37 - Data compression on.
    09-09-1998 22:37:19.06 - Remote modem hung up.
Note the 21.69 seconds between the last two messages.
Why such a long time?

I think that your ISP is waiting for your PPP-client
to "communicate" with the ISP's PPP-server.
After _NOT_ receiving any messages from your computer,
for '20' seconds, your ISP is just "hanging-up" on your modem.

I must agree with Otta, farther more, remove the compresstion string at your modem initial string commands, and tell us what was the log and outcome.

I am still waiting for the EXACT string needed for your modem to work with V90 and x2 according to the modem manual, as I asked befor 2 comments ago...

ACSComm, i already icluded the x2 init string +ms=vx2,1,300,0 excluding s32=32 which gave an error message in hyperterminal. %C0 to remove compression has no effect so i unchecked error control and data compression @ modem properties. w/ compression off, the result is still the same. tone heard last 20 sec.


otta, w/ compression off, i see garbage characters before login message. the sequence is reverse when compression is on.

If you _DO_ see the "login" message,
but do not respond to it, for, say, 20 seconds,
then your ISP will "drop" the connection.

If you _DO_ respond to to "login" message,
then your ISP's PPP-server will enter PPP-mode,
and will start sending PPP-messages to you.
These PPP-messages *DO* appear to be "garbage",
but they are _NOT_ !

i have to correct the last reported result. with compression off, protocol-none, there is no connection problem. except the max is only 33.6k. sorry about that.

Could you produce a "detailed-log", like the one above,
for a "compression-off-plus-protocol-none" session? Thanks.

We have extensive experience here with the Cirrus CL-MD5650 modem chipset.We have found them to be excellent for:
(1)Preventing loose papers from being blown off desks in areas close to open windows.
(2)Nothing else.

These modems DO NOT WORK! (above 33600).We're currently stuck with a pile of them,as is our supplier.The word has gone out in the channel to avoid them.Ours were manufactured by Jaton,but the problem seems to be traceable to the Cirrus chipset.In defense of Jaton and Cirrus: thier 33600 modem works just fine,but these 5650's were simply a big mistake.In their price range,most suppliers have switched to the TI chipset generics.

otta, here's the log you asked:

09-14-1998 22:22:14.64 - Cirrus 56K bps Modem Internal (CL-MD5650) Modem in use.
09-14-1998 22:22:14.64 - Modem type: Cirrus 56K bps Modem Internal (CL-MD5650) Modem
09-14-1998 22:22:14.64 - Modem inf path: CIRRUS~1.INF
09-14-1998 22:22:14.64 - Modem inf section: Gen56kInt
09-14-1998 22:22:14.89 - 115200,N,8,1
09-14-1998 22:22:14.89 - 115200,N,8,1
09-14-1998 22:22:14.89 - Initializing modem.
09-14-1998 22:22:14.89 - Send: ATZ<cr>
09-14-1998 22:22:14.90 - Recv: ATZ<cr>
09-14-1998 22:22:15.14 - Recv: <cr>
09-14-1998 22:22:15.14 - Interpreted response: Informative
09-14-1998 22:22:15.14 - Recv: <lf>
09-14-1998 22:22:15.14 - Interpreted response: Informative
09-14-1998 22:22:15.14 - Recv: OK
09-14-1998 22:22:15.14 - Interpreted response: Ok
09-14-1998 22:22:15.14 - Send: AT &F E0 V1 W4 &D2 &C1 S0=0 -C1 <cr>
09-14-1998 22:22:15.15 - Recv: <cr>
09-14-1998 22:22:15.15 - Interpreted response: Informative
09-14-1998 22:22:15.15 - Recv: <lf>
09-14-1998 22:22:15.15 - Interpreted response: Informative
09-14-1998 22:22:15.16 - Recv: AT &F E0 V1 W4 &D2 &C1 S0=0 -C1 <cr>
09-14-1998 22:22:15.43 - Recv: <cr>
09-14-1998 22:22:15.43 - Interpreted response: Informative
09-14-1998 22:22:15.43 - Recv: <lf>
09-14-1998 22:22:15.43 - Interpreted response: Informative
09-14-1998 22:22:15.43 - Recv: OK
09-14-1998 22:22:15.43 - Interpreted response: Ok
09-14-1998 22:22:15.43 - Send: ATS7=60\T0L1M1\N0%C0"H0\Q3B0N1X4<cr>
09-14-1998 22:22:15.43 - Recv: <cr>
09-14-1998 22:22:15.43 - Interpreted response: Informative
09-14-1998 22:22:15.43 - Recv: <lf>
09-14-1998 22:22:15.43 - Interpreted response: Informative
09-14-1998 22:22:15.52 - Recv: <cr>
09-14-1998 22:22:15.52 - Interpreted response: Informative
09-14-1998 22:22:15.52 - Recv: <lf>
09-14-1998 22:22:15.52 - Interpreted response: Informative
09-14-1998 22:22:15.52 - Recv: OK
09-14-1998 22:22:15.52 - Interpreted response: Ok
09-14-1998 22:22:15.52 - Send: AT+MS=VX2,1,300,0<cr>
09-14-1998 22:22:15.52 - Recv: <cr>
09-14-1998 22:22:15.52 - Interpreted response: Informative
09-14-1998 22:22:15.52 - Recv: <lf>
09-14-1998 22:22:15.52 - Interpreted response: Informative
09-14-1998 22:22:15.59 - Recv: <cr>
09-14-1998 22:22:15.59 - Interpreted response: Informative
09-14-1998 22:22:15.59 - Recv: <lf>
09-14-1998 22:22:15.59 - Interpreted response: Informative
09-14-1998 22:22:15.59 - Recv: OK
09-14-1998 22:22:15.59 - Interpreted response: Ok
09-14-1998 22:22:15.59 - Dialing.
09-14-1998 22:22:15.59 - Send: ATDT;<cr>
09-14-1998 22:22:16.99 - Recv: <cr>
09-14-1998 22:22:16.99 - Interpreted response: Informative
09-14-1998 22:22:16.99 - Recv: <lf>
09-14-1998 22:22:16.99 - Interpreted response: Informative
09-14-1998 22:22:16.99 - Recv: OK
09-14-1998 22:22:16.99 - Interpreted response: Ok
09-14-1998 22:22:17.00 - Dialing.
09-14-1998 22:22:17.00 - Send: ATDT#######<cr>
09-14-1998 22:22:45.63 - Recv: <cr>
09-14-1998 22:22:45.63 - Interpreted response: Informative
09-14-1998 22:22:45.63 - Recv: <lf>
09-14-1998 22:22:45.63 - Interpreted response: Informative
09-14-1998 22:22:45.63 - Recv: PROTOCOL:NONE
09-14-1998 22:22:45.63 - Interpreted response: Informative
09-14-1998 22:22:45.63 - Recv: <cr>
09-14-1998 22:22:45.63 - Interpreted response: Informative
09-14-1998 22:22:45.63 - Recv: <lf>
09-14-1998 22:22:45.63 - Interpreted response: Informative
09-14-1998 22:22:45.63 - Recv: COMPRESSION:NONE
09-14-1998 22:22:45.63 - Interpreted response: Informative
09-14-1998 22:22:45.63 - Recv: <cr>
09-14-1998 22:22:45.63 - Interpreted response: Informative
09-14-1998 22:22:45.63 - Recv: <lf>
09-14-1998 22:22:45.63 - Interpreted response: Informative
09-14-1998 22:22:45.64 - Recv: CONNECT 33600
09-14-1998 22:22:45.64 - Interpreted response: Connect
09-14-1998 22:22:45.64 - Connection established at 33600bps.
09-14-1998 22:22:45.64 - Error-control off or unknown.
09-14-1998 22:22:45.64 - Data compression off or unknown.
09-14-1998 22:29:37.75 - Hanging up the modem.
09-14-1998 22:29:37.75 - Hardware hangup by lowering DTR.
09-14-1998 22:29:39.95 - Send: ATH<cr>
09-14-1998 22:29:39.97 - Recv: ATH<cr>
09-14-1998 22:29:40.63 - Recv: <cr>
09-14-1998 22:29:40.63 - Interpreted response: Informative
09-14-1998 22:29:40.63 - Recv: <lf>
09-14-1998 22:29:40.63 - Interpreted response: Informative
09-14-1998 22:29:40.63 - Recv: OK
09-14-1998 22:29:40.63 - Interpreted response: Ok
09-14-1998 22:29:40.63 - 115200,N,8,1
09-14-1998 22:29:40.67 - Session Statistics:
09-14-1998 22:29:40.67 -                Reads : 46632 bytes
09-14-1998 22:29:40.67 -                Writes: 20996 bytes
09-14-1998 22:29:40.67 - Cirrus 56K bps Modem Internal (CL-MD5650) Modem closed.

in the post terminal window, the PPP session starts first then login message then PPP session again. unlike when compression is on, login first then PPP session. not sure if this important.

mattcei, you're the second expert to mention problem on cl-md5650 chipset. any explanation?

Adjusted points to 300

Hi lyndon1,

Don't really know if this will help you. Due to my job nature, I need to login remotely from home to office which is about 10 km away.

Due to our low ISP speed, I found that I can login with speed of 1200 bps in my area whereas my colleague have no difficulties login using 9600 bps. Of course he's staying somewhere else.

My scenario : can login using 9600 (Hayes modem), after established the line, hand-shaking, I will have "Connect 9600". Pressing enter will bring me to the login prompt (I'm logging in to UNIX server). After login, I can manage to do some other stuff  (ie vi, some unix command etc) but it only lasts for not more than 10 minutes. Then the mesg "No Carrier" will appear. This happen regardless of how many times you try.

Another colleague who's staying somewhere nearby face the same problem. So only the other fella (above) who's staying about 20 km manage to use higher speed.

My point is : it might depends on your phone's node at your area.

fredfan, if you are referring that my phone line is the problem, pls. review history questions again.

Excerpted:

22:17.00 - Send: ATDT#######<cr>
22:45.63 - Recv: <cr>

Seems "normal" -- the 2 modems took ~30 seconds to "negotiate".

22:45.63 - Recv: PROTOCOL:NONE
22:45.63 - Recv: COMPRESSION:NONE
22:45.64 - Recv: CONNECT 33600

Do you get this result *EVERY* time you
connect to this particular telephone-number?
If not, you may have just hit a "temporarily-noisy"
connection via your telephone-company,
and the results of the "negotiations"
have disabled the (extra) PROTOCOL(s) and COMPRESSION.

22:45.64 - Interpreted response: Connect
22:45.64 - Connection established at 33600bps.
22:45.64 - Error-control off or unknown.
22:45.64 - Data compression off or unknown.
29:37.75 - Hanging up the modem.
29:37.75 - Hardware hangup by lowering DTR.
29:39.95 - Send: ATH<cr>

Seems fine -- the connection stayed "up" (for 7 minutes)
until, apparently, you told it to go "down".

I'll repeat the important question:
do you _ALWAYS_ get these three values:
 PROTOCOL:NONE
 COMPRESSION:NONE
 CONNECT 33600
when using _THIS_ modem and _THIS_ telephone-number?

lyndon1, I'm not blamming your phone line but your phone's node. A node is required to serve a certain area.  Do a simple test. Try using your neighbour's phone line. If all of them face the same problem then you're having the same problem which I'm having now.

fredfan, whether my access connection is pure copper or via digital loop carrier, it boils down to one phone line.
recall:
1-w/ x2 usr sportster , my connection is consistent 48k
2-w/ cl-md5650 on 2 different locations - same problem

your suggestion falls on #2 but on different nodes only.

to get conclusive results (modem,line,node) you have to conduct the 2 tests above.

otta, i really thought disabling error control & compression would always give 33.6k good connections. but after taking several more tests, 56k bad connections appear.

09-16-1998 21:35:16.43 - Cirrus 56K bps Modem Internal (CL-MD5650) Modem in use.
09-16-1998 21:35:16.43 - Modem type: Cirrus 56K bps Modem Internal (CL-MD5650) Modem
09-16-1998 21:35:16.43 - Modem inf path: CIRRUS~1.INF
09-16-1998 21:35:16.43 - Modem inf section: Gen56kInt
09-16-1998 21:35:16.66 - 115200,N,8,1
09-16-1998 21:35:16.66 - 115200,N,8,1
09-16-1998 21:35:16.66 - Initializing modem.
09-16-1998 21:35:16.66 - Send: ATZ<cr>
09-16-1998 21:35:16.67 - Recv: ATZ<cr>
09-16-1998 21:35:16.93 - Recv: <cr>
09-16-1998 21:35:16.93 - Interpreted response: Informative
09-16-1998 21:35:16.93 - Recv: <lf>
09-16-1998 21:35:16.93 - Interpreted response: Informative
09-16-1998 21:35:16.93 - Recv: OK
09-16-1998 21:35:16.93 - Interpreted response: Ok
09-16-1998 21:35:16.93 - Send: AT &F E0 V1 W4 &D2 &C1 S0=0 -C1 <cr>
09-16-1998 21:35:16.93 - Recv: <cr>
09-16-1998 21:35:16.93 - Interpreted response: Informative
09-16-1998 21:35:16.93 - Recv: <lf>
09-16-1998 21:35:16.93 - Interpreted response: Informative
09-16-1998 21:35:16.96 - Recv: AT &F E0 V1 W4 &D2 &C1 S0=0 -C1 <cr>
09-16-1998 21:35:17.21 - Recv: <cr>
09-16-1998 21:35:17.21 - Interpreted response: Informative
09-16-1998 21:35:17.21 - Recv: <lf>
09-16-1998 21:35:17.21 - Interpreted response: Informative
09-16-1998 21:35:17.21 - Recv: OK
09-16-1998 21:35:17.21 - Interpreted response: Ok
09-16-1998 21:35:17.21 - Send: ATS7=60\T0L1M1\N0%C0"H0\Q3B0N1X4<cr>
09-16-1998 21:35:17.21 - Recv: <cr>
09-16-1998 21:35:17.21 - Interpreted response: Informative
09-16-1998 21:35:17.21 - Recv: <lf>
09-16-1998 21:35:17.21 - Interpreted response: Informative
09-16-1998 21:35:17.31 - Recv: <cr>
09-16-1998 21:35:17.31 - Interpreted response: Informative
09-16-1998 21:35:17.31 - Recv: <lf>
09-16-1998 21:35:17.31 - Interpreted response: Informative
09-16-1998 21:35:17.31 - Recv: OK
09-16-1998 21:35:17.31 - Interpreted response: Ok
09-16-1998 21:35:17.31 - Send: AT+MS=VX2,1,300,0<cr>
09-16-1998 21:35:17.31 - Recv: <cr>
09-16-1998 21:35:17.31 - Interpreted response: Informative
09-16-1998 21:35:17.31 - Recv: <lf>
09-16-1998 21:35:17.31 - Interpreted response: Informative
09-16-1998 21:35:17.38 - Recv: <cr>
09-16-1998 21:35:17.38 - Interpreted response: Informative
09-16-1998 21:35:17.38 - Recv: <lf>
09-16-1998 21:35:17.38 - Interpreted response: Informative
09-16-1998 21:35:17.38 - Recv: OK
09-16-1998 21:35:17.38 - Interpreted response: Ok
09-16-1998 21:35:17.38 - Dialing.
09-16-1998 21:35:17.38 - Send: ATDT;<cr>
09-16-1998 21:35:18.84 - Recv: <cr>
09-16-1998 21:35:18.84 - Interpreted response: Informative
09-16-1998 21:35:18.84 - Recv: <lf>
09-16-1998 21:35:18.84 - Interpreted response: Informative
09-16-1998 21:35:18.84 - Recv: OK
09-16-1998 21:35:18.84 - Interpreted response: Ok
09-16-1998 21:35:18.85 - Dialing.
09-16-1998 21:35:18.85 - Send: ATDT#######<cr>
09-16-1998 21:35:41.08 - Recv: <cr>
09-16-1998 21:35:41.08 - Interpreted response: Informative
09-16-1998 21:35:41.08 - Recv: <lf>
09-16-1998 21:35:41.08 - Interpreted response: Informative
09-16-1998 21:35:41.08 - Recv: PROTOCOL:NONE
09-16-1998 21:35:41.08 - Interpreted response: Informative
09-16-1998 21:35:41.08 - Recv: <cr>
09-16-1998 21:35:41.08 - Interpreted response: Informative
09-16-1998 21:35:41.08 - Recv: <lf>
09-16-1998 21:35:41.08 - Interpreted response: Informative
09-16-1998 21:35:41.09 - Recv: COMPRESSION:NONE
09-16-1998 21:35:41.09 - Interpreted response: Informative
09-16-1998 21:35:41.09 - Recv: <cr>
09-16-1998 21:35:41.09 - Interpreted response: Informative
09-16-1998 21:35:41.09 - Recv: <lf>
09-16-1998 21:35:41.09 - Interpreted response: Informative
09-16-1998 21:35:41.10 - Recv: CONNECT 56000
09-16-1998 21:35:41.10 - Interpreted response: Connect
09-16-1998 21:35:41.10 - Connection established at 56000bps.
09-16-1998 21:35:41.10 - Error-control off or unknown.
09-16-1998 21:35:41.10 - Data compression off or unknown.
09-16-1998 21:36:03.43 - Remote modem hung up.
09-16-1998 21:36:03.43 - Recv: <cr>
09-16-1998 21:36:03.43 - Interpreted response: Informative
09-16-1998 21:36:03.43 - Recv: <lf>
09-16-1998 21:36:03.43 - Interpreted response: Informative
09-16-1998 21:36:03.43 - Recv: NO CARRIER
09-16-1998 21:36:03.43 - Interpreted response: No Carrier
09-16-1998 21:36:03.51 - Hanging up the modem.
09-16-1998 21:36:03.52 - Send: ATH<cr>
09-16-1998 21:36:04.12 - Recv: <cr>
09-16-1998 21:36:04.12 - Interpreted response: Informative
09-16-1998 21:36:04.12 - Recv: <lf>
09-16-1998 21:36:04.12 - Interpreted response: Informative
09-16-1998 21:36:04.12 - Recv: OK
09-16-1998 21:36:04.12 - Interpreted response: Ok
09-16-1998 21:36:04.12 - 115200,N,8,1
09-16-1998 21:36:04.15 - Session Statistics:
09-16-1998 21:36:04.15 -                Reads : 163 bytes
09-16-1998 21:36:04.15 -                Writes: 110 bytes
09-16-1998 21:36:04.15 - Cirrus 56K bps Modem Internal (CL-MD5650) Modem closed

> after taking several more tests, 56k bad connections appear.

How can you tell that they are "bad" connections ?

 21:35:41.09 - Recv: COMPRESSION:NONE
 21:35:41.10 - Recv: CONNECT 56000
 21:35:41.10 - Connection established at 56000bps.
 21:35:41.10 - Error-control off or unknown.
 21:35:41.10 - Data compression off or unknown.
 21:36:03.43 - Remote modem hung up.

This looks "normal".

What happens in the 22.33 seconds between 35:41 and 36:03 ??

What _should_ happen is that the PPP-client on your
computer should "connect" to the ISP's PPP-server.
That seems to _NOT_ be happening.
The ISP's PPP-server seems to detect "no response" from your PPP-client,
and then tells the ISP's modem to issue a "hang-up",
to release the telephone-line and modem for another customer.

I'm still not getting your point. You admit that your modem is of substandard quality, and that by changing the modem you can get much better perforance, So, why is this still being discussed?  Do you hope to redesign the chipset and forward it to Jaton so they can make a better modem for you? What is the point, I don't see a question here that hasn't been answered.

otta, that 22.33 sec period is where the "dead tone" is heard.i think.

> that 22.33 second period is where the "dead tone" is heard.i think.

Can you please try again, to measure the length of the tone,
to try to verify your conjecture?

joe_m, this is an exchange of info, remember? i'm giving 300 pts. to an expert giving a brief technical explanation of the problem.

what i do w/ the info is none of your  concern.

otta, i timed the continuous tone @ 20 sec.

I have to go with Joe. The 56k standard has been problematic since it's inception. It works at 33.6 but not at 56k. The only logical assumption is that the isp doesn't support 56k x2. It's not uncommon for isp to claim x2 support. In reality there is no support. If you are not going to accept what appears good advice, try a another 56k modem and see what happens.

I disagree with DANPAUL about being "problematic" and having "no support".

The ISP I use has US Robotics Total Control Servers,
and their X2 support worked fine.

They upgraded to USR's V.90 support,
and had some initial problems.
Once they fixed those "teething" problems,
with support from US Robotics, V.90 has been working fine.

I usually get connected at 48000 or 49333,
with a US Robotics V.90 Sportster,
and have seen 50666 with a Cardinal Connecta X2 modem
(this modem has been "flashed" to V.90).

The nice things about the Cardinal modem are:
 - it works just as well as the USR Sportster,
 - it costs 1/3 as much,
 - it is _NOT_ a "Winmodem" (i.e., can be used with Linux or OS/2)
 - it has jumpers for Plug-and-Play versus "legacy" use,
   so it can be used with non-PnP systems.

if there's a way to eliminate that unusual tone, maybe i can connect @ x2 modulation.

> if there's a way to eliminate that unusual tone,
> maybe i can connect @ x2 modulation.

Did you "flash" your X2 modem to the V.90 standard?

Your "modem-initialization" string(s) are the key to
allowing the modem to try using X2 or V.90 "negotiations".
Just tell your modem to _NOT_ use V.90,
and it should use the X2 protocol.

Also, try limiting the modem to, say, 48000,
as the maximum connection-rate, rather than allowing
it to try to use 56000.  Very occasionally, my modem has
negotiated to 50666, and then "squawked" like yours.
When I hang-up, and re-dial, it selected 49333,
and connected successfully.  After all, a "good" connection
at 49333 is much better than "no" connection at 50666.

otta, if you review history questions, i have tried the ff:

1-flash to v90 using jaton but flash back to original x2
2-forced 48k connection w/ +ms=vx2,1,300,48000

w/ #2 i still hear continuous tone

> if you review history questions

There's over 150K of text/HTML, to review!!!

Suggestion: try different speeds, i.e.,
24000, 26400, 28800, 31200, 33600, and
28000, 29333, 30666,
32000, 33333, 34666,
36000, 37333, 38666,
40000, 41333, 42666,
44000, 45333, 46666,
48000, 49333, 50666,
52000, 53333, 54666,
56000, 57333.

to determine how "fast" your modem _CAN_ go.

otta, w/ v34 modulation, max. rate is 33.6k. forcing lower rate is not a problem. (i.e., +ms=v34,1,300,28000)

w/ x2 modulation, tests made from 56k to 33.3k shows continuous tone and no carrier message.(i.e,+ms=x2,1,300,48000)

you have suggested these tests before.

It's time to buy a better modem.

I recommend the Cardinal Connecta X2 internal modem.
Best price I've seen is $46 US, before the retail-store's
own $14 US mail-in rebate.  It's just as good as a US Robotics
V.90 Sportster, and there's a free upgrade to V.90,
via Cardinal's web-site.  It can be used as "plug-and-play",
or the jumpers can be set (COM4/IRQ3 works for me).

It is the same X2/V.90 technology as the USR,
and even the same chip-set (from Texas Instruments),
but at 1/3 of the price!

otta, could you at least read the cl-md5650 documentation @ their website and in your expert opinion - what went wrong. is it the firmware, hardware (power , registers,etc.) software? is 350 points not enough?just a brief explanation, no recommendation.

What's the URL for their web-site?

http://www.cirrus.com

> w/ x2 modulation, tests made from 56k to 33.3k shows
> continuous tone and no carrier message.
> (i.e,+ms=x2,1,300,48000)

Can you software-upgrade ("flash") your modem to V.90,
and repeat the above range of tests (from 56K to 33.3K)?

Is your ISP running the lastest "flash" of software
on their X2-servers?

I looked at the URLs at:
<A HREF="http://www.cirrus.com/products/overviews/md56xx.html">
CL-MD5650/MD5650T/MD5651T </A>
<A HREF="http://www.cirrus.com/products/overviews/md56xx.html">
CL-MD5652/MD5652T/MD5653T </A>
<A HREF="http://www.cirrus.com/products/overviews/md56xx.html">
CL-MD5662T/MD5663T </A>
<A HREF="http://www.cirrus.com/products//ftp/pubs/md56xxmb.pdf">
Marketing Brief</A>
(83 Kb)
<A HREF="http://www.cirrus.com/products//ftp/pubs/md565Xpb.pdf">
Product Bulletin</A>
(June '98: 160 Kb)
<A HREF="http://www.cirrus.com/products//ftp/pubs/md56xxdb.pdf">
Preliminary Data Book</A>
(May '97, v0.8: 1.2 Mb)
<A HREF="http://www.cirrus.com/products//ftp/pubs/an_md55.pdf">
Application Alert</A> Important corrections to the CL-MD56XX Data Book
(July '97, v1.0: 33 Kb)

and: http://WWW.CIRRUS.COM/products/overviews/md56xx.html

but didn't see anything relevant.

On the page at:
http://WWW.CIRRUS.COM/drivers/modemdrv/
I noticed that Cirrus makes modems for several retail
manufacturers, and you should "contact that manager
for software upgrades".  So, which retail manufacturer's
name appears on your modem?

Use the 'ATI0', 'ATI1', 'ATI2', through 'ATI13' commands,
while in some application like HyperTerminal,
to help you find the answer to the above question.

again, my 56k cl-md5650 modem came under web excel drivers & x2 firmware. but ati7 response is cirrus logic, inc. modem engineering dept.

flashing the v90 from jaton only resulted to a consistent 33.6k connection (v34). no more continuous tone. forcing speeds from 56k to 33.3k has no effect.

in the cl-md5650 documentation, there is one revision: addition of s32 register. i don't have this in my active & stored profile. can this be corrected with a firmware patch or what.

Is your ISP running the latest "flash" of software
on their X2-servers?

What's the response from 'ATI3'?

> can this be corrected with a firmware patch or what.

Some of the other 'ATInn' commands will give
the "flash-date" of the BIOS. Maybe, there is
a newer X2 (not V.90) "flash" for your modem?

if the x2 sportster i tested can connect @ 48k, i can assume their server runs on the standard x2 firmware. are there various version of x2?

ati3 reponse is cl-md56xx

regarding the latest x2 firmware of cl-md5650, i'm still searching if there is one.

This may be way out in left field, but I just had a similar problem after changing ISP

Eventually answer was amazingly to delete the "network adapter" in Settings / Control Panel / System.  The next time after doing this, everything came right

Are you using a USR Courier modem??

Are you using a USR Courier modem??

The CL modem evidently has two problems. First, it does not work at over 33k6. The second problem is the renegotiation firmware. Your modem never connects at over 33k6. It gives you connected message, but the host does not agree, and tries to renegotiate. That is the tone you hear. The firmware of your modem does not respond, and the host times out.

public, i'm not sure if this is a hardware or firmware problem. i'm getting an error message if i write ats32=32 to enable x2. can this be corrected with a firmware patch?

public, please clarify when the modem renegotiates. if connection fails with x2 modulation does it not fallback to v34?

> I'm getting an error message if i write ats32=32 to enable x2.

Strange.  If your modem really is an X2 modem,
as you claim, then using 'AT&F' or 'AT&F1'
should reset it to the "factory" settings,
which should enable all the features of the modem,
including, of course, X2.

> can this be corrected with a firmware patch?

Haven't you already tried this, by "flashing" to V.90 ?

otta, as i said before the s32 register is a revision found in the cl-md5650 documentation. i requested cirrus logic for a generic firmwares but was adviced to use fcc numbers to trace manufacturer instead.

you have been @ cirrus website & verified cl-md5650 to be x2 modem.

flashing v90 firmware from jaton only resulted to v34 connection.

> was advised to use FCC numbers to trace manufacturer ...

Use the "FCC ID Search Form", at:
http://www.fcc.gov/oet/fccid/
to perform this search.

What error do you get when you try to set S32?
Why are you trying to set this switch?
What happens if you simplify the modem-initialization
string, and just use 'AT&F' or 'AT&F1' ?

lyndon,  I have read the tech doc's you want expert opinion on, according to the tech notes, they expect their modem to perform like every other 56k modem, only you suggest that it doesn't. Now, do you expect someone else to read between the lines and suggest that this modem manufacture made a mistake in engineering the chip set used in your modem, or maybe more likely a small faulty capacitor, or line matching  transformer? Well that is not something somone, or any expert can tell you from the documents on the manufactures web page. Now, will you suggest that my expert opinion is not good enough to settle this? I think you will find a few experts that will agree with me. You have already stated that the USR modem gives somewhat satisfactory results from your location, this is further proof in the poor design of the Jaton modem. The reason I sugest the faulty  hardware of the modem, and not the chipset is because a manufacture can get a bad run of parts that only effects a certain quantity of product, but not all. If it were the chipset, it would effect every modem every made with that chipset that would be a lot of bad 56k modems, and angry people. We are moving away from suspecting the firmware because you said you were able to flash upgrade it. If Jaton was aware of any problems with firmware, they would have addressed them in that flash upgrade. I would see if you could get that modem exchanged for any other modem, even one of the same (if you have to), you could end up with a modem from another manufacturing date that may work fine.

> public, please clarify when the modem renegotiates. if
> connection fails with x2 modulation does it not fallback to
> v34?-
When the modem renegotiates depends on the firmware. Evidently this does not work properly. Hence the firmware problem. There may also be a hardware problem, since you never connected at over 33k6. However this condition also may be due to the firmware not configuring the CL chips properly. I doubt that jaton does any software development. They just may be passing on a buggy patch.

joe_m, let me rephrase your opinion:

1-defective modem from production run rather than poor chipset design
2-not a firmware problem

my comments:

for #1
a.there are 2 other modems i mentioned before w/ the same problem & they are not jaton. i only use jaton because of commonality-cirrus logic chipset
b.addition of s32 register-a revision. perhaps my modem is an earlier model-no s32 register

for #2
a.connection @ v34 but not @ v90
b.jaton tech support page for firmware update is still blank?! are they working on the problem?

i appreciate the experts who are more interested in the exchange of info. rather than those who are so eager to close the subject even if the other party is not satisfied.

otta, i im not sure where the fcc number is but all the chips except the intel flash rom are cirrus logic (cl-mdxxxx). at commands also points to cirrus logic.

ats32=32 just respond w/ ERROR

even w/ at&f command, there is no s32 in the stored & active profile which will enable x2. in addition at+ms=vx2,1,300,0 must be entered manually.

public, is the absence of s32 register a firmware or hardware problem?

> I'm not sure where the FCC number is ...

It's silk-screened onto the modem's circuit board.
Keep looking.
> ats32=32 just respond w/ ERROR
> there is no s32 in the stored & active profile

Your modem doesn't seem as "new"
as the online documents.
Flash it to V.90, and see if 'S32' appears.

> is the absence of s32 register a firmware or hardware problem?

Yes, it's one of those two.
It's just a "application programming interface" (API)
to the modem's firmware.

I would consider it a firmware problem. It is also possible that the necessary action has no function in hardware, or does not work properly in hardware, and therefore was not implemented in firmware.

public, 2 more clarifications please:

1-addition of s32 register is a latest revision in cl-md5650 documentation. my modem is perhaps an earlier model. is a firmware update incorporating new control codes (x2) the solution?

2-is it possible i did connect @ 56k but due to power exceeding the limit , it triggers an alarm signal from the telephone company equipment. that continuous tone may be interfering with the modems communications.

otta, i only found md-120/220 in the board & it is not in the fcc database. only the at commands points to cirrus logic.

flashing the v90 from jaton did not result s32 to register. only change is the ati1 response from CD04.8-540 PARALLEL-SP 04 to
CD04.8-600 PARALLEL (V.90 CAPABLE)-SP 04-DSP PATCH:76

> to enable X2, the S32 register should be 32.
> but reading is always s32=0.

Odd.

My US Robotics manual states that 'S32=32' will *DISABLE* X2 modulation,
while you claim that using 'S32=32' on your modem will *ENABLE* X2 modulation.

Are you *SURE* that the C-L documentation is correct?

Most of the "switch-registers" in a modem have 'zero' values,
by default.  You need to set non-zero values to do something
which is "unusual", i.e., to *DISABLE* X2 is an unusual request,
so setting S32 to a non-zero value, namely binary '100000',
is an "unusual" thing to do.

If the C-L documentation is wrong, then using 'S32=32' will
cause your X2 modem to negotiate to a maximum-speed of 33600,
which is what you have observed.

P.S. I'm just under 22 hours early in this announcement,
(relative to: August 02 1998 - 08:42PM PDT)
but this problem has now been "open" for almost two months,
and the size of the HTML for this page is now 221328 bytes.
Happy second anniversary!  :-)

otta, the revision in cl documentation is very explicit in the addition of s32 register & to enable x2, s32=32. it's not yet in my active & stored profile, but response of at32? is always 0.

your comments have generated interest among other experts & have provided a lot of interactive discussions. the question already used up a lot of memory, so closing time is near.

others have made their position clear & i was hoping you can come up with something definite.

> the revision in cl documentation is very explicit
> in the addition of s32 register & to enable x2, s32=32.

Well, it may be "explicit", but it still may be WRONG!

> it's not yet in my active & stored profile,
> but response of at32? is always 0.

Is this after the "flash" to V.90 ?

For my US Robotics V.90 Sportster modem,
after using 'AT&F1', the value of S32 is '2'.

Setting S32 to '0' causes my modem to limit
its connection-speed to 14400, since use of the V.8 protocol
has been disabled by the '0' value.

Setting S32 to '66' (64+2) causes my modem to use X2,
rather than V.90, mode, at a speed of 48000.

Setting S32 to '98' (64+32+2) disables both X2 and V.90,
and causes my modem to connect at 28800.

Given that X2 technology is licensed from US Robotics,
it's curious that the "flipping" ('32' versus '0')
of the meaning of this bit has been done by the C-L firmware.

Use HyperTerminal to communicate with your modem.
Send 'AT&F1' to your modem, and then
send 'ATI3' and capture the output.
Send 'ATI4' and capture the output.
Send 'ATI5' and capture the output.

Include that "captured" output here.

Also, send 'AT&U17&N24',
i.e., floor connect-speed of 33333,
and ceiling connect-speed of 48000 bps.
Then, send 'ATDTnnnnnnnn', where "nnnnnnnn"
is the X2 telephone-number of your ISP.
When the modem "hangs-up",
send 'ATI6' and capture the output. Also,
send 'ATI7' and capture the output,
then "paste" the output here.

otta, is the s32 really important for x2?
1-not found in active & stored profile
2-writing any settings for s32 including 0 gives error message before & after flashing to v90

your latest comments to capture results of at commands was maybe based on USR modems. cirrus have different definitions (i.e.,ati6-country code.ati7-board firmware version etc.)

x2 in my case is really problematic. anyway, having been with the question since the very beginning, you are in a better position to give an expert opinion. please give your opinions/findings based on your comments & post it as an answer.

Finally, and yet no closer to an actual solution than a month ago, not to mention that many experts will still agree that your main trouble is in buying a substandard modem, and the next is expecting too much from it.
You were so sure that you were the only one to know the answer to your question, yet you wanted someone else to agree with you, and give you some quantified proof that you were correct in rejecting so many other replies. Thank you for giving someone the opportunity to put this question to rest.
In contrary to what you believe, experts are not here to educate you, just help you solve a problem.

1. perhaps. But if the CL chipset really has a problem, then fixing it in firmware may not be possible.

2. No. Ihave never heard of telephone company doing that. The tone is most likely generated by the host modem.

As for the connection speed, if you think about it, nobody connects at 56k. Typical speeds are 44k, 48k. If you get the theoretical maximum in the field, it does not look right.
Also if the CL chipset has a problem, you will not find it the datasheets. You would have to probe the implemented silicon version, and search for discrepancies from the datasheets.

joe_m, you really have a style. but i don't know if other experts agree with your opinions. you can speak for your self.

public, i read in last month news archive from 56k.com website that fcc is proposing to increase the 53k limit to 56k. anyway i wish to thank you for sharing your knowledgeable comments. i hope there are more like you here.

> is the s32 really important for x2?

Yes!! - at least for the firmware for a US Robotics modem.

Since Cirrus Logic licenses the X2 technology
from US Robotics, I expect that they would use
the same (most of the time!!) switch-settings,
so as to advertise that their modems will be
"compatible" with USR modems.  If you make something
too incompatible, nobody will buy it,
which seems to be a bad "business-decision".

> capture results of at commands was maybe based
> on USR modems. cirrus have different definitions
> (i.e.,ati6-country code. ati7-board firmware version etc.)

Anyway, please do it.  I can compare the output with
the data-sheets from C-L's web-site, with a "fresh" set of eyes.
Be sure to try 'ATI20' through 'ATI29',
in additional to the "usual" range ('ATI0' to 'ATI13')
because it is(!) a C-L modem.

Thanks.

In practical cases the spreed is limited by phone line quality, time of day , sun spots, etc.  and not really by FCC limits. Performance is better at night when there is less interference from other active noise sources.
Incidentally The discussion is closer to the probable cause. The tone is the server modem trying to renegotiate, and the CL modem has multiple defects preventing advertised operation.

*grinning* well i got a headache trying to read all the above in a hurry, but what i think the problem is that it is trying to connect with the wrong hardware handshaking method, so, go into your dialler setup and try selecting DSR if DTR is checked or DTR if DSR is checked, and if neither of those work try no hardware handshaking. You previous modem either used the default method, or some setup program changed it, and this modem does not use the same method in both 33 and 56k modes, or one method is faulty in either of those modes.
other issues...
are you very sure your telephone exchange can handle 56k?? there are still a lot about that are incapable of it.
use the shortest possible connection to your phone socket, keep it well away from other wiring.
clean all the plugs and sockets with an alcohol based cleaner on a q-tip (cotton bud) video or audio tape head cleaner is good. make sure all plugs connect firmly, if you have had your wall socket more than 5 years, you maybe ought to think about replacing it, it could be quite worn by now. all these things can affect the noise on a connection. You really have to get paranoid about the physical connections to get 56k to work, i had to to get 33k to work!

oh well, hope that is some help

RoadWarrior

Roadwarrier wrote:
> are you very sure your telephone exchange can handle 56k??

When your headache clears, you'll note that LYNDON1
has tried a US Robotics modem, and does get 56K.
However, his question pertains to the
(non-)performance of his Cirrus-Logic modem.

roadwarrior, can't blame you for not fully digesting the history of this question. idon't think i have a problem with dtr/dsr line signals. the hardware/software handshaking had already  been suggested by other experts already-no effect.

otta , here's the info you asked previously. hope you can make an expert opinion on this.

AT&FW3
OK
ati10
252
OK
ati11
254
OK
ati12
OK

ati14
CONNECT 115200/V23C/NONE/NONE/TX=1200/RX=1200

NO CARRIER
ati15
ERROR
ati20
OK
ati21
OK
ati22
Cirrus Logic, Inc.
OK
ati23
OK
ati24
OK
ati25
OK
ats32?
000

OK
ats32=0
ERROR
ats32=32
ERROR
ati6
US
OK
ati7
CIRRUS LOGIC, INC.  MODEM ENGINEERING  FIRMWARE DEPARTMENT
OK
at&v

ACTIVE PROFILE:
B1 E1 L2 M1 N1 T Q0 V1 W3 X4 Y0 &C1 &D2 &G0 &J0 &P0 &Q0 &S0 &U0 &Y0
%A013 %C1 %E1 %G1 \A3 \C0 \G0 \J0 \K5 \N3 \Q3 \T000 \X0 -C0 -J1 "H3 "O032
S00:000 S01:000 S02:043 S03:013 S04:010 S05:008 S06:002 S07:060 S08:002
S09:006 S10:014 S11:070 S12:050 S18:000 S25:005 S30:000 S33:010 S37:000

STORED PROFILE 0:
B1 E1 L2 M1 N1 T Q0 V1 W0 X4 Y0 &C1 &D2 &G0 &J0 &P0 &Q0 &S0 &U0
%A013 %C1 %E1 %G1 \A3 \C0 \G0 \J0 \K5 \N3 \Q3 \T000 \X0 -C0 -J1 "H3 "O032
S00:000 S02:043 S06:002 S07:060 S08:002 S09:006 S10:014
S11:070 S12:050 S18:000 S25:005 S30:000 S33:010 S37:000

TELEPHONE NUMBERS:
&Z0 =
&Z1 =
&Z2 =
&Z3 =

OK

*smiling* just wondering if you are in the UK, not sure if this suggestion has a bearing anywhere else, but, when you use your phone, is the phone line quite "loud"?  *smile* if so, it could be a problem with automatic gain control on your telephone line, which causes difficulties with a lot of modems dropping connection, because the data is quieter than the negotiation so when the agc adjusts on the exchange during the negotiation phase it hasn't enough "volume" to sustain the connection thereafter, the solution to this is to phone your telecom company and ask them to turn the automatic gain control off on your line at the exchange *smile* some modems are more sensitive to this than others, depends a lot on the line driver circuitry and the volumne and agc arrangements inside the modem itself. *Smile*

this is a commnon prob on British Telecom lines in the UK

Road Warrior.

public, you're the third expert to agree on certain defects of cl modem chipsets.

roadwarrior, pls review the history. you find i'm in the philippines were many telecom players are  entering with updated technologies.

*smiling*
it is the updated technologies that cause the problem AGC is an advanced exchange feature, found on latest digital exchanges.
*smile*
*shutting up now, thinking so far i am sounding stupid, grin*

Road Warrior.

roadwarrior, i don't know what the british telecom or UK equipment is. but if you review history you'll find i'm connected to EWSD switch (Siemens). Perhaps the agc can be set to manual. but why mess with it. w/ usr modem i can connect @ 48k. w/ cl modem under discussion, connection is sustained but @ 33.6k only.

making premise is not necessarily stupid. you have to start someplace and narrow the possibilities. :-)

*extract*
to enable x2, the S32 register should be 32. but reading is always s32=0. writing ats32=32 thru hyperterminal returns an error. i
        tried at+ms=vx2,1,300,0 as adviced by Cirrus Logic but no improvement either.
*snip*

just wondering, 32 may be the byte value, but maybe you have to poke it in in hex. ie 20, so maybe ats32=20 might work. *smile*

Ok, since USR invented X2 technology, the s32=32 command will force any X2 modem except the gateway telepath II to a 14.4 connection speed.  I've been working thru this rather long list, and it does seem that several points are quite valid.  1) that it certainly seems that your particular modem might have a manufacturing flaw.  2)Honestly I've worked for several isps and Cirrus Logic modems do not have nearly the best reputation in the tech support community.  I realize that you want to solve the issue, but sometimes there comes a point when it's good to let the fat lady sing...
Now then the above aside, you could try this string
at&fs202=32s10=60
it might keep the connection going long enough to do something. also, which version of DUN are you using?

jb

> the s32=32 command will force any X2 modem except
> the gateway telepath II to a 14.4 connection speed.

Use 'S32=34' to enable V.8 and disable X2,
or 'S32=98' to enable V.8 and disable X2 and disable V.90,
and you'll get speeds up to 33600 or 31200 or 28800,
instead of just 14400.

Correct me if I'm wrong, but Lyndon1 wants to connect at the 56k, so that s32=34 string isn't going to help him get that speed.  v.8 isn't the protocol he's trying to acheive...  also lyndon1, you could add s15=128, which will disable the v.42 but leave x2 up and running.  might make a difference

jb

PALANTYRI, after you had set switch S32,
you observed a maximum connection-rate of 14400.
Correct?
I gave you (and LYNDON1) the _CORRECT_ information
on how to obtain a maximum connection-rate in
the range of 26400/28800/31200/33600,
depending on the quality of the telephone-company
wires & the capabilities of the ISP's modems.

> ... Lyndon1 wants to connect at the 56k ...
LYNDON1 opened this question, over two months ago,
seeking a solution to that "want".
Please read the entire "history" of this question.

Otta, nothing personal but you just contradicted yourself.  you admit that the s32=34 allows a connection between 26400 and 33600, but then tell me that he wants to connect at 56k, something I already knew.  what's the point of using the s32=34 string if it's going to limit him to 33.6 or lower?  I HAVE read all two months of & I didn't tell him to use the s32 command, I was just explaining it's purpose for some of the previous posts

jb

> ... but you just contradicted yourself.

No, I didn't.  You posted a comment,
about being limited to 14400.
I replied to *YOUR* comment, full well knowing
that neither your comment nor my comment
would help LYNDON1 move closer to a solution.
You were "straying" from the solution to LYNDON1's problem,
and I attempted to "shepherd" your disgression.

>  you admit that the s32=34 allows a connection between 26400 and 33600,

Correct.

> but then tell me that he wants to connect at 56k,
> something I already knew.

I knew it too -- I've been trying to help since the start,
over two months ago.

> what's the point of using the s32=34 string
> if it's going to limit him to 33.6 or lower?

What's better for LYNDON1 -- a modem which connects
at X2 speeds, and then drops the connection,
or a modem which is "usable" at 33600 ?

> I HAVE read all two months of & I didn't tell him
> to use the s32 command, I was just explaining it's [sic]
> purpose for some of the previous posts

I was just explaining your partial explanation.

ok, I was just checking.  I'd have suggested the s32=34 string but I figured he wouldn't want it...  we'll see if any of this is useful to lyndon1...out of curiousity, the ph country extentions is what, the Phillipines?

jb

According to:
 http://rs.internic.net/cgi-bin/whois?ph-dom ...

Philippines (Republic of the) top-level domain (PH-DOM)
   Disini Software International Inc.
   P.O. Box 12054, Ortigas Center P.O.
   Pasig, Metro Manila 1600
   PHILIPPINES

   Domain Name: PH

   Record last updated on 29-Apr-93.
   Record created on 14-Sep-90.

Common problems with the X2, K56 and V.90 standards is that they are asymmetrical with respect to the upload/download speeds that are "allowed" over the PSTN.  X2, for example, conforms to the 33.6Kbps upload and 54Kbps download standard.  K56 conforms to the speeds differently.  V.90, however, is a comprimise between the differing technologies (X2 or K56) that is supposed to ensure interoperability.  Below is a quick table to see if you will be able to connect using X2 modulation.  FROM YOUR ABOVE COMMENTS, I ASSUME THAT YOUR MODEM IS NOT SUPPORTING V.90 BUT IS SUPPORTING X2.

* When dialing into the Total Control Chassis, you will only be able to connect using X2 modulation for high-speed.  Your ISP can upgrade the modems in the Total Control Chassis to the V.90 standard, but it costs about $650 per 24 modems and they are usually reluctant to shell out the money.

* When dialing into the Portmaster 3, you will be able to connect using X2 modulation, K56 modulation or V.90.  This is because the V.90 is a mixture of the two types of modulation techniques.

Now, for the important stuff...

Many times, users configure their DTE speeds for 115,200 (usually the highest setting that they can select).  This is the speed of communications between the modem and the computer NOT line speed.  The DTE speed is important because it sets up internal buffers for data overflow or starvation during serial communications (in your case a modem connection).  When using X2, or other high-speed connections, *AND* when using compression alogorithyms (like V.42 (4:1 compression, no error correction), V.42 bis (4:1 compression with error checking), etc.), your DTE buffers are actually expected to handle throughput speeds upwards of 224,000bps (56Kbps x 4 = 224,000bps).  Newer version of HyperTerminal and Dial-up Networking now allow users to select these higher DTE speeds.  This sets up the FIFO port transmit and receive buffers to work correctly with the higher-speed connections.  I will assume, however, that your workstation does not have this option, so we will attack this problem by forcing the modem to buffer data for the DTE connection.

First, we need to make sure that your modem is configured properly.  We will reset the modem to factory NVRAM settings using the command:  AT&F1&W <cr>
This writes the default hardware flow control template to the modem's NVRAM.
Next, issue the modem an ATZ <cr>.  This resets the modem loading default NVRAM settings.
Thirdly, send the modem the command:  AT&B0 <cr>.  This tells the modem to vary the serial (DTE) speed according to the connection speed.  The hardware flow control will buffer the incoming data.

Now, to test this, dial your ISP using the command ATDT ######### <cr>.

The X2 modulation produces a "special" sound when training.  We used to refer to it at USR as the "Submarine tone."  The training also includes what I can only describe as a "hiccup" or "missed beat" like if someone came along and bumped into a record player (if you're old enough to remember turntables).  This is perfectly normal.  X2 modulation also takes a little while longer to negotiate.  A lot of stuff is happening during the time the modems are exchanging squealing tones.

Once the connection is established, you should then see a login prompt.  Login, if you can, and leave the connection up for a while.  After the connection is torn down (either by you issuing the modem a +++ATH or by the remote side), please type ATI6 and post them here.  I will look the information over and tell you definitively whether the problem lies within your modem, your workstation configuration, your ISP's dial-in server or if what you are seeing is even a problem at all.  PLEASE TRY TO DIAL INTO THE USR TOTAL CONTROL CHASSIS AND NOT THE PORTMASTER, IF POSSIBLE.  WE NEED TO LIMIT THE NUMBER OF VARIABLES IN THIS EQUATION.

Look forward to hearing from you.

Regards,
turnkey

> Once the connection is established,
> you should then see a login prompt.

Alas, but that's not what LYNDON1 "sees".
Instead, a steady-pitch, indefinite-length "whine" is heard,
following the end of the "negotations",
but only with the Cirrus-Logic modem,
i.e., not with the US Robotics modem.

> please type ATI6 and post them here.

Good idea.

For a related "screeching-modem" problem, see:

https://www.experts-exchange.com/Q.10085963

I have experienced the same exact type of connectivity problems that you are experiencing on several different types of "56k x2" modems on different customer machines.   I would only be able to connect to the ISP if I lowered the connect speed.
I have used several Jaton modems, and I have been quite pleased with them.  After performing the flash rom upgrade to support v.90 they connected every time at the fastest speed possible on that line.
The Jaton flash utility is well written and I would not hesitate to flash your modem (if you haven't already).  It will not allow you to flash it if you downloaded the wrong version, so it is pretty well idiot proof.

With my past experience in dealing with this problem, and in working with Jaton modems I truly feel that installing the update will resolve this problem.

What model of Jaton modem do you have?  Is it the 56k Modulator?

Joel

Excuse any repeat advice...

Make sure that the speed setting in your MODEM properties in C/pannel is the same as in the properties of the DUN connection icon. Also make sure that you have a decent port speed in 'system/devive Mngr/ports'.

What you also should do is check that the other software settings
(like flowcontrol-hardware, error correction etc.) are reflected either by your Init string or are already set on your modem.

Also see that your Com port is set to H/W in C/Pannel

In the properties of DUN connection under connection/port settings, set receive buffer to LOW

Are you able to remove your PC to another location but on the same Telco switch? If so, try it (a friends house etc) If this works, it is your home line itself that has an error - This may be due to MANY factors.

I would sugest that you turn off error correction(both in C/p and as an AT command)

Are you able to test with another known working CL X2?

have fun

I used to use x2 technology modems until I had the same problem only to fing out my ISP had been upgraded to V.90. I went to my modem maufactuers web site for a quick fix.

kevcraig,
Did you happen to notice how long this question is, or how many comments or rejected answers have been posted to it?
I do not expect you to read through each and every posting, but at the very least read enough so you get some sort of idea what has been tried and what has not.
If you wish to post a comment, you are welcome to do so --  if not encouraged.  If you have the answer, I would strongly suggest that you read through the comments before posting it.
Your "Answer" is little more than a comment.  PLEASE DO NOT POST COMMENTS AS ANSWERS!

Thank You,
Joel R. Helgeson

hey lyndon1 are you still looking at this?

kevcraig, sorry but your proposed answer had been discussed before without improving results. perhaps the question remains why cl-md5650 modem cannot connect @ x2 modulation.

just got back from vacation...

my isp just upgraded their usr server w/ v.90 so i'm surprised i got 37,333 connection (slight improvement). however, w/ their portmaster 3 server, max. is still 33.6k. why?

the question of x2 connection remains unanswered & still more questions keep popping up on the cl modem. loading the jaton v.90 was a risk since my modem was packed under web excel.

> w/ their portmaster 3 server, max. is still 33.6k. why?

Your modem failed to negotiate an X.2 or V90 or K56flex
connection with their server, so the "fall-back" taken
was to connect using V34, i.e., 33600 or 31200 or 28800 or 26400 or 24000 or ...

otta, perhaps the cl modem is very selective w/ 56k connection. w/ several tries, i can only connect @ v.34 but not @ v.90 or x2 w/ the portmaster server. w/ usr server, i can connect @ v.90 but not @ x2.

eventhough the x2 question remains unanswered, i'm quite glad of the interest & valuable comments shown. for this, i'm thanking every experts for their contribution.

otta, being w/ the question from the very beginning, i prefer you close this whatever you like, observation, suggestion,etc.

if you decline, please recommend whom the points will be awarded to unless there are more comments.