Choosing the most suitable 4-to-1 Multiplexer for my ECG application

Hello Guys,

I'm working on an ECG applcation, getting the signal from the heart, and processing it. The circuit I'm building is an analogue circuit, it has a single supply of +5 V and ground. Please see the picture to know the topology I'm trying to do screenshot  . I need you to help me find the most suitable multiplexer upon the following requirements.

Mandatory requirements:
1 - The multiplexer to take in 4 input analogue signals and output 1 analogue signal. (It's fine if the multiplexer is dual or even quad 4:1 multiplexer with common select signals, I would only use 1 circuit per IC).
2 - The multiplexer to have very low input noise (because it will take the signal from the heart and the signal is very small and sensitive, 2mV amplitude and 1 - 100 Hz frequency)
3 - The multiplexer to operate on single supply of +5V (or less) and ground (i.e. no -ve voltage supply to be needed).
4 - The multiplexer to be in DIP or PDIP package.
5 - The multiplexer select inputs will be input from a microcontroller.

Optional requirements (preferable)
1 - To have a SPICE simulation model
2 - To be cheaper than £10 (UK sterling pounds) because my project has a limited budget.

I usually order components from any of these websites, in uk.farnell.com, uk.rs-online.com, and http://www.digikey.co.uk/

Please be very elaborate in your answers and help me reach a specific answer.

adkryAsked:
Who is Participating?

[Product update] Infrastructure Analysis Tool is now available with Business Accounts.Learn More

x
I wear a lot of hats...

"The solutions and answers provided on Experts Exchange have been extremely helpful to me over the last few years. I wear a lot of hats - Developer, Database Administrator, Help Desk, etc., so I know a lot of things but not a lot about one thing. Experts Exchange gives me answers from people who do know a lot about one thing, in a easy to use platform." -Todd S.

☠ MASQ ☠Commented:
adkry, this isn't really meant as an answer, just an observation. You say you're looking for 2mV amplitude and 1 - 100 Hz frequency and so I assume you're really only trying to pick up the QRS component of the electrical activity in the heart.  Even so a heart rate of over 100 bpm isn't unusual & if your monitor couldn't cope with >100Hz it wouldn't practically be of much use.  I realize this is more about the exercise of designing the device but wondered if you might be painting yourself into a corner with those specifications?
0
Dave BaldwinFixer of ProblemsCommented:
Your biggest problem is going to be finding a MUX that has low enough input leakage to work in your circuit.  Leakage limits for the NXP 74HC4052 are up to +/-1uA which I believe is higher than the input current for the amplifier.
0
adkryAuthor Commented:
@Masueraid

I were only saying average specifications, the high pass filter is at 0.1 Hz, and the low pass filter is at 150 Hz. Please help me with the multiplexer

@DaveBaldwin

I tried to search for "input leakage" but couldn't understand it, can you briefly explain what is input leakage in a multiplexer and how it affects my circuits ?
You mentioned "which I believe is higher than the input current for the amplifier", first, I'm not gonna use that same exact amplifier in the picture, I might be using AD620, or INA122 or INA125 or others (I'm searching for a single-supply instrumentation amplifier, you can help me here http://www.experts-exchange.com/Hardware/Misc/Q_27465862.html#37333425). Anyway, so "Leakage limits for the NXP 74HC4052 are up to +/-1uA which I believe is higher than the input current for the amplifier" what does that mean or indicate ?

Thank you guys for your help. keep it up :D
0
10 Tips to Protect Your Business from Ransomware

Did you know that ransomware is the most widespread, destructive malware in the world today? It accounts for 39% of all security breaches, with ransomware gangsters projected to make $11.5B in profits from online extortion by 2019.

Dave BaldwinFixer of ProblemsCommented:
Any semiconductor has some current that will flow when you put a voltage across.  Even a diode reverse biased will have some small amount of current flow.  A 'low-leakage' diode will be in the picoamps, standard diodes in the nanoamps.  At the amplifier inputs, whatever leakage current there is in the input stage is swamped out by the normal input current.  If the 'off-state' leakage for the mux inputs is 1uA, then you could have 3uA from the 3 inputs that are off feeding into the 'on' input.  That will affect your measurement.
0
adkryAuthor Commented:
@DaveBaldwin

Ok so I understood what you meant by leakage current. I didn't understand "If the 'off-state' leakage for the mux inputs is 1uA, then you could have 3uA from the 3 inputs that are off feeding into the 'on' input." what 'on' input are you referring to ?

To make it brief, so what do you suggest ? Do you reckon I should simply put the multiplexer AFTER the instrumentation amplifier ? The signal quality is the most important priority so I don't want anything to affect the measurement. Or should I simply look for a multiplexer with very low input leakage current
0
Dave BaldwinFixer of ProblemsCommented:
The leakage current is for each input to the mux.  Each input is connected to the body.  The leakage current goes thru the body to the active input.

The way you have it wired up, you are converting all of your signals to common mode signals which means you get no output.  What are you trying to accomplish with the mux?

I wouldn't be trying to use a single supply circuit myself.  You have a problem with creating a signal 'ground' or common point when you do that.
0
adkryAuthor Commented:
ok, so first question, does the AD620 have a leakage current ? I didn't have a problem with connecting the heart signal directly to the AD620, will I have safety problems connecting the heart directly to the mux ?

I have 4 electrodes connected to the body, (1 of them is gonna be connnected to ground). I want to have the ability to input any combination of signals to the AD620 instrumentation amplifier. i.e. I have electrodes A,B,C and D. I want to input to the AD620 signals A - B, B - A, A - C, C - A, and so on.

why will I have a problem creating a ground ? my ground is gonna be 0 V, and everything will be connected normally, the components I use will be chosen so that they operate on single supply. Please explain.
0
gismo51Commented:
ok, so first question, does the AD620 have a leakage current ? I didn't have a problem with connecting the heart signal directly to the AD620, will I have safety problems connecting the heart directly to the mux ?

I have 4 electrodes connected to the body, (1 of them is gonna be connnected to ground). I want to have the ability to input any combination of signals to the AD620 instrumentation amplifier. i.e. I have electrodes A,B,C and D. I want to input to the AD620 signals A - B, B - A, A - C, C - A, and so on.

I would definitely connect each input through some sort of buffer amp so that all "processing" including switching/muxing is downstream of that.  If you want to sense the difference between each pair you could either reference each IA to the "ground" electrode (e.g. one that's attached low on the chest) and subtract them in a downstream IA fed by the output of two muxes that steer the desired (buffered) signals to that IA.  Or you could connect each input to two or more IAs, one for each pair of inputs you'd like to compare (as long as the number of inputs was reasonably low the latter would be my choice.
As to the safety issue, I think you 'd want to connect each lead through a current limiting resistance and include some means of limiting the voltage to the devices (IAs or muxes) in a fail safe manner.  For instance if you used IAs to buffer the inputs directly, a crowbar on the IA supplies set for +/-5v and a 100k resistor in series with each lead would limit the potential fault current to less than 50 na which should be pretty safe unless the probes are attached directly to the heart.


I have 4 electrodes connected to the body, (1 of them is gonna be connnected to ground). I want to have the ability to input any combination of signals to the AD620 instrumentation amplifier. i.e. I have electrodes A,B,C and D. I want to input to the AD620 signals A - B, B - A, A - C, C - A, and so on.

why will I have a problem creating a ground ? my ground is gonna be 0 V, and everything will be connected normally, the components I use will be chosen so that they operate on single supply. Please explain.

Many devices intended for single supply operation will work with signals at or above the negative supply (ground) but most will not tolerate inputs that are below the negative supply. With a single supply you must offset your inputs so that the result will remain between the supply rails (depending on the device the available signal range may be a few volts less than the supply). This presents some serious design issues when you're talking about inputs connected to a living body as well as some potential noise issues if the signal levels are low.  Capacitor isolation would remove the DC bias from the input leads but that wouldn't be fail safe and your application might need DC coupling.

A much better choice is to use balances supplies, at least for the first stage so that the signals can remain centered at ground potential and still be adequately removed from the supply voltages.  There are op amps with a "Norton" (current vs voltage input) front end that can work with inputs outside the supplies although those require precise input resistors to convert voltage into current which come with their own problems, especially in a high impedance design.
0
adkryAuthor Commented:
@gismo51

WOW, that was a great answer. I'm trying to read and understand but first I think you should have a look at my circuit.
circuit
What's the main advantage gained when connecting the electrodes to a buffer amp ? I don't prefer to input the heart signal directly into the IA, because I need to see many pairs of signals, hence I'm gonna have to build many IA's in my circuit.

You mentioned "you could reference each IA to the "ground" electrode and subtract them in a downstream IA fed by the output of two muxes" , so why and how do I reference each IA to the ground electrode ? why can't I just connect the electrode to the input of the multiplexer ?

so you reckon I should stick with douple supply ? I could use the LT1054 to get a regulated -5 V
0
Dave BaldwinFixer of ProblemsCommented:
You should be viewing the human body as a resistor network.  If you connect a circuit ground to the body, any other connections will try to pull the voltage to ground.  If you connect a single supply amp that is biased to mid supply (2.5v), the body resistance will try to pull that voltage towards ground.  If the input resistance of the amp is very high, then the lower resistance of the body will pull the input voltage down close to ground.
0
gismo51Commented:

What's the main advantage gained when connecting the electrodes to a buffer amp ? I don't prefer to input the heart signal directly into the IA, because I need to see many pairs of signals, hence I'm gonna have to build many IA's in my circuit.
You would need a separate IA for each pair of inputs but not for the reverse polarity pairs.  So for three inputs you only need three IAs (for A,B, and C: AB,BC,AB), four inputs would need six IAs (AB,AC,AD,BC,BD,CD), five inputs requires 10 IAs, etc.  At some point it get's costly to use a separate IA for each pair of inputs plus the leakage current for each IA begins to affect too many inputs.  In your original question it sounded like you only needed 3 inputs and in that case no extra IAs are needed.  If you want to use more than four inputs I'd probably go with an "ultra low leakage" mux like a Maxim MAA328 on the inputs.

You mentioned "you could reference each IA to the "ground" electrode and subtract them in a downstream IA fed by the output of two muxes" , so why and how do I reference each IA to the ground electrode ?
By that I meant you'd connect one of the IA's inputs (typically the minus) for each sensor lead to the common (ground) sensor lead that you mentioned.  You will need to have at least one sensor (attached to the subject) connected (through a current limiting resistor) to the circuit ground in order to keep the other inputs range centered on your internal ground.  Otherwise the inputs of the IAs or muxes could float out of range for the input devices.

why can't I just connect the electrode to the input of the multiplexer ?
You can but if the number of inputs is small you can eliminate the issues associated with muxing low level signals such as noise and offsets created by the mux).  Applying the muxing further downstream means you're switching higher level signals so the noise introduced is proportionately less.

so you reckon I should stick with douple supply ? I could use the LT1054 to get a regulated -5 V
I certainly would if at all feasible.  The circuit you posted illustrates the issue with single ended operation of your OP amps.  If you power them from a single supply you will have to bias the input to each amp so that it's centered at approximately 2 volts above the negative supply rail (the minus terminal of the 9V battery).  Otherwise the Op amps will rectify the signal minimums.  The input protection on the OP amps will eventually charge the series capacitor enough to provide this bias but there will be lots of distortion while that charging takes place plus a small amount of distortion to maintain it.
0
adkryAuthor Commented:
@DaveBaldwin
@gismo51

Thank you so much for your help guys, I really appreciate it, you made me rethink some design decisions. Man I got myself into so much trouble haha

1 - OK, I will only use 3 inputs (the 4th will be ground). I would use 3 AD620's. Here's the topology I'm gonna build, any comments on that ?
circuit
OK, I'm gonna use the LT1054 to get me -5 V. Seems I totally misunderstood when d-glitch told me to use a single positive power supply (http://www.experts-exchange.com/Hardware/Misc/Q_27465862.html#37200659), I don't think he meant use single supply for the op-amps, rather he might have meant generate the negative supply from the positive supply.

2 - Now that I've changed the topology as you can see in the picture, and decided to use dual +5 and -5 supply, what 4:1 or 8:1 multiplexer would be most suitable ? I've checked the MAX328 it's really good, but now that I've changed the topology, is it still the most suitable ?

3 - quick question, what's the difference between single input multiplexers and differential input multiplexers ?

Thank you so much for the explanations.  Please answer these last questions stating a specific multiplexer and I'll end the question.
0
Dave BaldwinFixer of ProblemsCommented:
1.  That looks better.

2. The Mux choice is much less critical now because the inputs are sourced from a low impedance and won't affect the signals.  Something inexpensive like the CD4051/2 will work.  One of the Intersil devices with lower on resistance will work even better,

3. A differential mux is two single muxes in the same package controlled by common control signals instead of separate controls.  The main advantage of having them in the same package is that non-linearities between the channels will usually track better.
0

Experts Exchange Solution brought to you by

Your issues matter to us.

Facing a tech roadblock? Get the help and guidance you need from experienced professionals who care. Ask your question anytime, anywhere, with no hassle.

Start your 7-day free trial
gismo51Commented:
DaveB's answers are spot on.  I would add some series resistance on the probe leads as I mentioned previously.  With batteries for supplies you don't have to worry about excess voltage so you don't need any crowbars but 5v still poses a risk to the patient when the electrodes make good electrical contact.  The resistance should be enough to limit any fault currents to less than 500na.  Those same resistors can protect the AD620 inputs from static discharge and other stray voltage. For that purpose you want to limit the current to less than about 10ma with an applied voltage equal to the local mains voltage (e.g. 120v or 200v).

The sensor leads should be shielded to minimize stray coupling and you might want/need to drive those shields with guard amp circuit as shown in the AD620 datasheet.  
0
adkryAuthor Commented:
DaveBaldwin, can't thank you enough man. I searched upon a low on resistance criteria and found the ADG608 to meet absolutely all my requirements, it even has a simulation model on multisim.
gismo51, thanks so much, you really helped me a lot, I will add the resistors you mentioned . The sensor leads will be totally isolated don't worry.

I finally found what I need. Thanks a lot
0
Dave BaldwinFixer of ProblemsCommented:
You're welcome.  Let us know how it works when you build it.
0
It's more than this solution.Get answers and train to solve all your tech problems - anytime, anywhere.Try it for free Edge Out The Competitionfor your dream job with proven skills and certifications.Get started today Stand Outas the employee with proven skills.Start learning today for free Move Your Career Forwardwith certification training in the latest technologies.Start your trial today
Hardware

From novice to tech pro — start learning today.