(I am quite new to electronics and pcb design.)
(necessary files are attached, tell me if you need something else )
-------------------------------------------------------------------------------------------------------------------------------------------
Hi everyone,


I am making a MOSFET array which will be used for EMG-signal acquisition but there are a few things i am uncertain about.

Useful EMG-signals range from 1-10mV and 8hz-250hz, these parameters can differ depending on the muscle group and location of measurement.

-------------------------------------------------------------------------------------------------------------------------------------------

The main principle of the circuit:

- 0) The circuit will be used to feed trough the EMG-signals of different muscle groups to a single EMG-sensor (amplification, rectification etc. circuit) after each other this way you only need to use one EMG-sensor for limbs or even the whole body.

- 1) When LI1-3, HI1-3 is set HIGH the gate of the corresponding MOSFET should also be set high and create an N-channel, thus making it possible for current to flow from drain to source.

- 2) Each muscle has two electrodes working as cathode and anode and a shared reference electrode for all muscles (when E1 works as cathode E2 works as anode, this is because of the ATP, ADP, phosphor levels varying between the electrodes, the dominant ionic charge at the electrodes is negative with the electrode closest to the motory cells responsible for the contraction or relaxation cycle having more negatively charged ions then the other).

- 3) The output from the mosfet array will then be directed to a highly sensitive opamp circuit (using ina106 and TL084's).

-------------------------------------------------------------------------------------------------------------------------------------------
LIa pin triggers the gate of the mosfet of the electrode on muscle 'a' positioned closest to the spine.
HIa pin triggers the gate of the mosfet of the electrode on muscle 'a' positioned furthest from the spine.

MMEa: stands for the electrode closest to the spine on muscle 'a'.
EMEa: stands for the electrode furthest from the spine on muscle 'a'.
-------------------------------------------------------------------------------------------------------------------------------------------

Desired output:

- 1) When all LI1-3, HI1-3 pins are set low the potential difference between the MOSFETs there source layers should at least be under 1mV (MOSFETs shouldn't conduct).

- 2) When LIa and HIa are set high the the potential difference between the source layers should only equal the potential difference between EMEa and MMEa (MOSFETs connected to LIa and HIa conduct).

-------------------------------------------------------------------------------------------------------------------------------------------

Questions:

- Does the attached circuit meet the above mentioned criteria?

- Is it possible that because of impurities or doping difference in the silicon layers of the MOSFET that when the source layers are connected with each other, a temporary potential difference larger then 1mV is created?

- When using a MOSFET driver for this kind of purpose (low current/voltage through N-channel) do i even need the High-side floating offset voltage pins because i can't imagine a voltage higher then that of the supply voltage of the MOSFET driver comming from the electrodes?

- If i don't need the High-side floating offset voltage pins do i even need to add the bootstrap capacitors which are currently in the circuit?

- I know the mosfet driver is kind of overkill but any better recommendations for smd MOSFET drivers or MOSFETS would be highly appreciated.

- Is there anything else wrong with my circuit or explanation or do you have any tips?


Thanks in advance

 

Woops, sorry there is a lot wrong with the circuit i forget about this thread.

 

Woops, sorry there is a lot wrong with the circuit i forget about this thread.

Sorry again, the circuit might still work forget my previous reply. Still need help.

 

I'll preface this by saying that I haven't designed an electromyogram preamplifier before, but signals in millivolt region and 10-250Hz frequencies are well within my remit - looks like a microphone preamp to me!

The LAST thing I'd want to put on the input of a preamplifier is a MOSFET. They are full of 1/f noise. I would amplify FIRST, and multiplex afterwards, when the signal is large enough not be degraded by the noise of the multiplexing circuitry.

What is your source impedance?

 

I'll preface this by saying that I haven't designed an electromyogram preamplifier before, but signals in millivolt region and 10-250Hz frequencies are well within my remit - looks like a microphone preamp to me!

The LAST thing I'd want to put on the input of a preamplifier is a MOSFET. They are full of 1/f noise. I would amplify FIRST, and multiplex afterwards, when the signal is large enough not be degraded by the noise of the multiplexing circuitry.

What is your source impedance?

Thank you for your response.

In ideal situations the impedance of the skin ranges from 10Ω to 100kΩ but with surface elelctrodes it is more between 1MΩ to 1.5MΩ for the distance of which the electrodes are sepperated (around 6-8cm). In the first amplification process there are also two 10kΩ resistors on the inverting and non-inverting pins of the ina106.

I hope this is what you meant with the source impedance.

 

I'll preface this by saying that I haven't designed an electromyogram preamplifier before, but signals in millivolt region and 10-250Hz frequencies are well within my remit - looks like a microphone preamp to me!

The LAST thing I'd want to put on the input of a preamplifier is a MOSFET. They are full of 1/f noise. I would amplify FIRST, and multiplex afterwards, when the signal is large enough not be degraded by the noise of the multiplexing circuitry.

What is your source impedance?

And also thanks for the tip of preamplifiing the signal before it is fed through the mosfets.

 

And also thanks for the tip of preamplifiing the signal before it is fed through the mosfets.

But do i even need the mosfets if i use multiplexers?

 

But do i even need the mosfets if i use multiplexers?

Perhaps not - something like a DG408/9 will do the job nicely. You will need more preamps, but a well-designed preamp is not expensive!

 

I'll preface this by saying that I haven't designed an electromyogram preamplifier before, but signals in millivolt region and 10-250Hz frequencies are well within my remit - looks like a microphone preamp to me!

The LAST thing I'd want to put on the input of a preamplifier is a MOSFET. They are full of 1/f noise. I would amplify FIRST, and multiplex afterwards, when the signal is large enough not be degraded by the noise of the multiplexing circuitry.

What is your source impedance?

(!!!Never mind, figured it out myself!!!)
I still have a small question about microphone preamps.

Do you perhaps know what parameter I need to search for to know the minimal voltage that the preamp can accurately amplify or is it just input ofset/noise voltage?
(!!!Never mind, figured it out myself!!!)

 

There's no minimum voltage, it all depends on the noise figure of your circuit. The noise depends on two things, the source impedance, and the device you use to amplify it.
I know the source impedance of a microphone (a few hundred ohms), but I haven't a clue about the source impedance of your electrode! You need to find that out - it makes a BIG difference if you have it wrong.
1/f noise will be your biggest problem.
see https://en.wikipedia.org/wiki/Flicker_noise and note the comment about MOSFETs!