[Okta Webinar] Learn how to a build a cloud-first strategyRegister Now

x
  • Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 547
  • Last Modified:

Hot motherboard power tranistor-

Greetings:  I just replaced a power tranistor that is mounted on a motherboard.  (N45H2A).  It was replaced previously by the distributor of this motherboard.  It ran for about 3 years and how it is out again.  After replacing the tranistor I find it is running pretty hot.  Can keep my finger on the heat sing for only about 3 seconds.  I haven't been able to find (cross reference) this transitor with another one that will handle more current/power.  A larger heat sink would be extremely difficult to install.

My question is, is it normal for such a transistor on a motherboard to run hot?  I hope someone can come up with a solution to this problem.  I don't trust hot transistors.

Karl


0
KarlAibling
Asked:
KarlAibling
  • 3
  • 3
1 Solution
 
jhanceCommented:
>My question is, is it normal for such
>a transistor on a motherboard to run
>hot?  I hope someone can come up with
>a solution to this problem.  I don't
>trust hot transistors.

Hot is a _VERY_ relative term.  Most power transistors are rated to run at an operating temp of at least 70C, some as much as 85C.  Above that, the semiconductor effect degrades so much that it's difficult for a conventionally designed transistor to operate.  If you can touch it for 2-3 seconds, it's NOWHERE NEAR even 70C.  70C is VERY HOT to the touch and will leave you with a blister (trust me, I know this from experience - OUCH!)

Two things:

1) Are you sure this is a power transistor and not a regulator?  The number you gave N45H2A is not a transistor type number I'm familiar with.  Usually these are 2NXXXX or 2SCXXXX. It's possible it's a "house marked" model where the number is not an industry number.

2) Since this MB is 3 years old, and you've replaced this component.  It seems likely that you'll get another 3 years out of it.  Do you really care if this thing is working in 3 more years?  It's already obsolete, isn't it?
0
 
KarlAiblingAuthor Commented:
jhance - Thank you for your answer.  Yes, it is a PNP transistor, cross referenced to an NTE378, if you want to check the characteristics. It is rated at 10A and 50 watts.  I haven't checked the pin voltages because I really don't know how it is wired up.  According to previous experts, it is used for a voltage regulator.  (I expected to see a VR when I pulled the heat sink off)

I checked one of the two "transistors" in my old 200MHz computer.  One of them was running warm; I guess around 120F.  I deflected air from the CPU fan to this transistor.  However, I can't do that with this layout.  

Point well taken on the obsolecence of the computer.  The problem is, I simply can't stand to discard anything that is running good.  It is a head thing.

  I am leaving the question open for a couple days just in case someone else, or you, comes up with a solution other than "fly it into the ground."  :-)

Thanks again.

Karl
0
 
joedCommented:
A transistor running too hot to touch is  not going to last. There is something drawing too much current thru this transistor. This is why it failed in the first place. You need to look for another problem on you motherboard.
Check for other components that may be hot. Pull any cards like modem and sound card.
Rereading your post I see that it was repaired previously. It could be that this part was under designed for the application. You could try installing a cpu type fan to assist in keeping it cool.
0
 The Evil-ution of Network Security Threats

What are the hacks that forever changed the security industry? To answer that question, we created an exciting new eBook that takes you on a trip through hacking history. It explores the top hacks from the 80s to 2010s, why they mattered, and how the security industry responded.

 
jhanceCommented:
joed,

>A transistor running too hot to touch
>is  not going to last.

That's baloney and a gross generalization!  Are you an engineer?  If so, you ought to know better...

Check a manufacturer's data sheet for one of these.  There will be no degradation from the expected MTBF for a component that is operated within it's rated specs.


Karl,

A 50W power transistor would need a fairly large heatsink to be able to dissipate it's full 50W.  The small bent aluminum heatsinks are only good for about 10W or so.  A 50W heatsink is going to be pretty big.

What kind of motherboard is this?  A Pentium MB (or later) usually has a power supply onboard for making the CPU voltage (typically 2-4 V) from the +5V supply.  Since a CPU can take 20W or so it would not be unusual to require 10A or so.  This would make the worst case dissipation in the regulator 10A x (5V - 2V) = 30W.  Enough to get it REALLY hot if the heatsink wasn't up to snuff.

As hard as it is to believe, MB makers don't always use parts that are correctly rated.  They cut corners.  A smaller heatsink could certainly have been included.  I don't think you're going to solve this problem by changing to a higher rated transistor.  The problem is that the heatsink is too small.  A higher rated transistor will still get just as hot since the temp depends on the power dissipation and NOT on the rating of the device.  A 500W transistor will still dissipate 30W in this application and still get too hot.

To fix it, either get a bigger heatsink, or make the existing one work better.  Blow a fan on it or bolt an extension on to the top of it.
0
 
KarlAiblingAuthor Commented:
jhance:  I'm inclined to accept your answer, but will still leave it open for one more day.

An update on this transistor:  I checked the voltage on all three legs which are:  Power Supply Ground to the Emitter= +5.05V
                                   "                "       to the Base    = +4.39V
                                   "                "     to the Collector= +3.46V

As you can see, the power supply output voltage is +5.05V with the normally lower Base voltage and an output voltage of 3.46 which is the CPU voltage.  (CPU calls for 3.45V)  

The motherboard is a 133MHz system, with, I think an AMD 133MHz CPU.

Karl
 
0
 
jhanceCommented:
Well your Vce is 1.59V.  According to the AMD K5/133 datasheet, that beast sucks down 44mA per MHz.  So at 133MHz, that would be 5.85A.  At a drop of 1.59V that means your transistor is dissipating 9.3W.  A TO-220 package can't do this w/o a heat sink.  Something like this:

http://www.aavid.com/atp/datashts/datashts.html

is rated to do up to 11W.  According to their graphs, 9.3W should make the operating temp about 60C.  Not too hot as most TO-220 transistors are rated to at least 70-85C.  60C is VERY HOT to the touch but not all that hot for reliability purposes.  If your heat sink is smaller (look at some of the other AAVID models on their web site and see if you can find a match, you can use the graph to estimate the operating temp.

In spite of what joed thinks, this is not rocket science nor is it black magic, it's just basic engineering calculations.  Even a 2-year Associates in Electronics Technology should be able to follow this....
0
 
KarlAiblingAuthor Commented:
jhance:  Thanks again for the info and the fine site.  I've decided to run it as is.  It would be extremely difficult to install a larger heat sink.  (It does have a small extruded heat sink on it)  It just isn't big enough.  I still might get ambicious and jury rig another heat sink on it.

Thanks for your help jhance.

Karl


0

Featured Post

[Video] Create a Disruption-Free Workspace

Open offices have their challenges. And Sometimes, it's even hard to work at work. It's time to reclaim your office and create a disruption-free workspace. With the MB 660, you can:

-Increase Concentration
-Improve well-being
-Boost Productivity

  • 3
  • 3
Tackle projects and never again get stuck behind a technical roadblock.
Join Now