Prosthetic arm amplifier and filter design

hisham.i

Hello...

Am undergraduate EE, and am working on my final year project which is the development of prosthetic arm which will work on signal from muscles taken from surface myoelectric sensors.

First of all i will explain about circuits i want to build, and i wish i will have comments which will help me to complete my project successfully.

My first goal is to take the EMG signals from muscles and make an amplification and filter circuit. So i did some research and designed this circuit, using instrumentation amplifier which is AD620, and TLC2274 for filtering circuits and second stage amplifier.

I attached the circuit in the PDF file. I have some questions concerning the circuit i attached:

1- I put the V+, and V- pins in AD620 to +-5 respectively, and the reference pin to 0v, and i made the gain of this stage to 5. So am not sure if the reference pin should be connected to 0v or to any other voltage?!

2-After i have designed the low pass(450HZ cut off frequency) and high pass filter(20HZ cut off frequency) using butterworth and sallen key, i added a second stage amplifier with gain 200, so i will have a total gain 5*200=1000. Also i make an offset of 2.5 v at this stage in order to allow this signal to enter the Analog to Digital Converter which am planing to use(ADS8325).. Is there any problem in my desing and do i have to take any other things into consideration for this circuit?

After this circuit works am planning to enter the data to a DSP, but still have low information about what DSP i have to use for such application, Please any help concerning the DSP type and DSP tutorials.

Thank you very much.
Hisham

Attached Files

amplifier and filter.pdf (36.7 KB, 72 views)

schip666!

Didn't you just ask this in: http://physicsforums.com/showthread.php?t=456551 ?
I think the fully differential circuit in the AD620 data sheet, as posted by dlgoff in that other thread, is what you want to use. The Ref you ask about is the "body ground" common point attached to the victim someplace distant from the myo-signal you are interested in.

You may still have issues with 60Hz (or 50 if you are not US) noise if your bandwidth is 20-450Hz.

But I don't see how you get a 2.5v offset in the last stage that has a 200 gain. Because it's a DC coupled amp, the + pin labeled U1:A(+IP) would have to be _very_ accurately set to 2.5/200 volts. You might be able to make it an AC coupled amp with an effective DC gain of 1 by putting a capacitor in series with the R1 input. Then pin 3 can be used to set the offset directly.

hisham.i

Yes i asked about this issue, but i did some simulation using proteus i decided to change the gain in the first stage in order if there is any dc offset not to be amplified.

So i send my new circuit if there is any new considerations i have to take also.

The source of the 2.5 v is not shown in the diagram but i did a buffer cicuit with a voltage divider in order to get a buffered 2.5v.
I want to add this offset in order to be able to enter this signal in the range of the ADC.

schip666!

The output stage's gain will be applied to both the signal and your bias input. That's why I said the +bias needs to be 2.5/200v, not 2.5v, and that will be hard to do. If you can AC couple the last stage then it will have an effective DC gain of 1 (or, hmm, maybe 2 at the + input...I always confuse myself with this...). If you are simulating this, try it out and see...

yungman

The way you draw the circuit is confusing, it is not the convensional way of drawing. Make sure you put bypass cap on the power supply pin of every IC.

The way you do the offset on the last stage is not a good idea. Using your circuit, connect a 2M resistor from pin 2 of U1A to -5V to do it as a summing and you get an offset of +2.5V and the output of U1A. Tie pin 3 to ground.

Learn to think through the circuit with your mind instead on relying on simulation every time. This is a very very simple circuit. Take the time draw out the frequency response will give you tremendous insight. Using simulation right at the begining will rob you from your experience. Use simulation verify your finding.

Try not to use resistors that is too high value. You will be create parasitic poles from stray capacitance or input capacitance. Also it will pick up noise from surrounding. I would not use 1M resistor if I can help it. This is all from real life experience from years of designing analog circuits like this.

hisham.i

Yes you are right schipp, the input bais will also be amplified by th gain of the last stage amplifier, so what i did is the following:
1- i amplified the signal from the low pass filter by a gain of 200, then i added a new Op Amp with unity gain to make the offset..i didn't find another way to add the offset while i want to amplify the signal by 200 gain.
Yugman thanks for your notes, but i didn't understand what you mean in these words :
Please if you can draw a schematic to know what you mean, but what i want to do is to amplify the signal by 200 and the add an offset of 2.5 v.

hisham.i

I attached a circuit design from Texas Instrument for ECG monitor amplifier and filter circuit, the last stage of the circuit Op AMP A3 it is used to make a gain of 200 and to shift the voltage 2.5v in order to allow the signal to be in the range of the ADC.
But the last stage will add 2.5 volts to the signal then multiply it by the gain 200 right?

Do i have to make another OP AMP to make the addition of 2.5 after the signal amplified by 200, or this circuit works as it is demonstrated in the daigram?

Attached Thumbnails