INA326 ECG design questions and problems

Xijun,

The OPA333 requires a minimum supply difference of 1.8V.  Also, it does not make sense to me that you would be able to input a voltage of 1V to a buffer whose positive supply is  only .553V as the amplifier would saturate before hitting your target output. Indeed, this is what I believe is happening; my suspicion is if you move your V+ supply up to 2.5V your results will improve for all cases except the Vin = 2.5V case.

Some other good amplifiers to consider that are low power are the OPA336 and the OPA348.

Matt

Hi Matt

Thank you very much for your quick reply again.

I am sorry that if I made you confused about the test procedures and results.

Below is the details of the test setup and results:

Setup:

===============================================

1.      Supply voltage: 3VDC

2.      Test circuit: see below (replace 5Kohm resistor by 10Mohm for measurement)

3.      Vin was connected to a DC power supply

Test results:

=======================================================

From the above test results, it did not seem that the op-amp was saturated because I could get expected voltage Vout, which was equal to Vin. However,

1. the input current at +IN input was expectedly high.

2. Voltage at -IN and +IN were not the same. (I though they should have very close)

I do not have solid Op-amp design experience. I would appreciate it if you could point out where the problems are or give me some suggestions.

By the way, I am going to test OPA336 once the samples arrive.

Thank you very much!

Xijun

Xijun,

Something does not seem right.  The voltages at the inverting an d non-inverting terminals should be the same.  I would remove everything (including the resistor) and see if you can get the OPA333 to behave the way that a buffer should.

Hi Matt

I agreed with you -- something does not seem right.

I have just done a quick test you suggested -- removing the non-inverting input resistor. Below is the test results: 

The results looked good except the input current at +IN input. I used a current meter to measure the DC input current at the non-inverting input. It was about ~0.1uA. This result matches my previous test result with 10Mohm input resistor.

Do you have any comments? I expect 100pA bias current.

Xijun

Xijun,

It is possible that the problem you were experiencing previously could have something to do with the shunt impedance of the volt meter if  you were measuring with respect to ground.  Keep in mind that the input bias current on an OPA such as the OPA333 which is an auto zero is a switched average of the voltage off the inputs; if you add anything to the circuit which perturbs this average it is possible that the bias current measurement can go awry.

My guess is that  you are not measuring bias current correctly with the meter you are using.  Typically bias current is measured using an integration capacitor over a specified time period.  You can figure out what capacitor you need for this using i = C*dV/dt, so if you expect 100pA, you can easily see a 100mV change in 10 seconds which equates to a 100pA/100mV/10 = 1nF measurement capacitor.  The capacitors that are best for this type of measurement are polypropylene or polystyrene as they have very low leakages associated with them.

Hi Matt

I really appreciate your quick response and I have to say 'thank you' again.

I am not sure if I completely understand what I explained. Just two more questions:

1. Is it okay to use a current meter, which is in series with +IN, for input bias current, measurement?

2. If I use OPA336, will I face the similar problem as OPA333 for measuring bias current?

Thanks!

Xijun

Hi Matt

It is Fluke 83. I set it to uA range. I think this can measure current down to 0.1uA.

If I put a volt meter in parallel to +IN, the bias current will go up from 0.1uA to 0.3uA. This means that the volt-meter take 2uA. 

Xijun

Xijun,

A fluke 83 is not sufficient in terms of accuracy or precision to measure pA of bias current.  I'm sure it will measure something, but not what you think.  Why are you trying to measure the bias current on the OPA333?  This is guaranteed in the data sheet...do you need it for some other purpose?

Hi Matt

I would like to use OPA333 to make activate electrodes, which need very high input impedance. I expect it is in the order of Gohm. I just wanted to verify the input impedance of the prototype I made to make sure everything works well. It does not mean that I do not trust the datasheet.

By the way, do you think OPA336 is better than OPA333 for my application because OPA336 takes 10pA bias current? What I exactly need is a single supple Op-amp with very high input impedance.

Anyway, thank you very much for help. I will stop here for this question.

Xijun

Xijun,

The OPA336 has an effectively higher input impedance than the OPA333.  There are other amplifiers such as the OPA340, OPA364, and OPA365 which could also work for your application but are higher in quiescent power.

Matt

Hi Matt

Many thanks! I will switch to OPA336.

Xijun

Hi Matt

Please refer to the following reference circuit you provided. Is the value of R1 Int  (316Kohm) for the integrator U1 important? It is unlikely to find resistors with the exact value. I am going to use 330Kohm. Is it okay? Thanks!

Xijun

Xijun,

The RC in the integrator feedback sets the high pass filter value.  You can set this RC to whatever you would like--330k should be just fine.

Matt

Hi Matt

Regarding question 2: 

"2. This question is about Driven Right Leg (DRL). I found that the DC level of DRL signal, see below, from INA326 is not exactly 1/2V+. This makes the DRL circuit to be saturated. Do you guys have any solution for improvement or avoid this."

Questions:

1. Can I use a capacitor with a large value, say 10uF, in between U4 and U5, to block the DC so that U4 will be be saturated? I have tried it and it looked good. Will this make the DRL circuit have no 60Hz cancellation?

2. How can I fine tune the the DRL circuit to make it has the best cancellation effect? I think I need to adjust the gain of U4 to achieve maximum cancellation.

Thank you very much

Xijun

sir plz attach the tina schematic of above post plz

thanks and regards

harmanpreet

SIR PLZ MAIL ME THE TINA FILE OR POST IT FOR THIS SCHEMATIC

SIR PLZ ATTACH THE TINA SCHEMATIC FILE THIS IMAGE

HELLO ,SIR PLZ  ATTACH THE TINA  SCHEMATIC FILE FOR THIS PROBLEM

See attached.  Eliminating the half supply pedestal and adding the 10M resistor to ground off the output of U4 helps with convergence.

Matt

See the following:

http://e2e.ti.com/support/amplifiers/precision_amplifiers/f/14/p/35437/436303.aspx#436303

Hi Xijun

                  I am working with the desgining of a non contact ECG sensor for my project. Do u have any recommandations to me based on your experiance? How did you deal with the desigining of active electrode? Thanks