OPA211: Build a high precision voltage controlled current source using a difference amplifier

Benyuan Liu

Part Number: OPA211
Other Parts Discussed in Thread: INA592Dear TI team, We want to build a high precision VCVS using a combination of a difference amplifier and a high precision op-amp. The frequency range is 1KHz to 1MHz and the current is fixed to 0.8mA. Currently, we are building a composite enhanced Howland circuit from the following links: www.analog.com/.../diff-amp-heart-of-precision-current-source.html We would like to use INA592 and OPA211 to build this circuit, and we would like to know is this pair the best combination whose performance might surpass the ADI's solution? With my best regards. Benyuan

over 4 years ago

0 Michael Steffes over 4 years ago

Guru 47535 points

I think you mean the INA592 - not the INA529

www.ti.com/.../description

0 Michael Steffes over 4 years ago in reply to Michael Steffes

Guru 47535 points

So I looked into that ADI reference, the op amp selection is mainly about I/O range to handle the Iout *Rload , if the Iout is 0.8mA what is the load range

You say 1MHz max, but the INA592 is only a 2MHz device, might get bandlimited and show errors at higher F..I could not find a faster device, I would maybe be looking at the 20Mhz AD8273 then pick a TI op amp based on I/O swing mainly, but also maybe at least a 10Mhz device.

Your low output current requirement means no external transistor is needed

0 Benyuan Liu over 4 years ago in reply to Michael Steffes

Intellectual 595 points

Thanks for your reply. I though a newer device might have a better performance :) Especially that the CMRR of INA592 is superior to AD827x series. I thought maybe INA592 has better gain resistors match.

If I were using INA592 and OPA211 to build the Howland current source, what the conservative bandwidth of the signal might be? I am familiar with LTSpice but not TINA right now,

0 Michael Steffes over 4 years ago in reply to Benyuan Liu

Guru 47535 points

Can't really run a sim without your load, min/max R or whatever it is.

0 Benyuan Liu over 4 years ago in reply to Michael Steffes

Intellectual 595 points

I will try to use LTSpice or TINA to do the simulation. Thanks!

0 Michael Steffes over 4 years ago in reply to Benyuan Liu

Guru 47535 points

Here you go, For +/-18V supplies, the OPA211 looks like a great choice,

Increasing R1 extends the bandwidth for this gain of 0.5/R1 solution.

At 10kohm, it was introducing a pole that caused phase margin issues.

At 100ohm R1, this is 113kHz BW

INA592 with OPA211.TSC

0 Michael Steffes over 4 years ago in reply to Michael Steffes

Guru 47535 points

After some effort, I imported the AD8273 model into TINA and have the running with the OPA211. With R1 =500ohm the forward path is very flat through 1Mhz. But the current source is rolling off on output impedance to where it is down to about 10kohm at 1Mhz from a very high DC value.

If you can in fact operating on a TINA file, let me know and I will insert into these,

0 Benyuan Liu over 4 years ago in reply to Michael Steffes

Intellectual 595 points

Sorry, I cannot get your TINA file working. Maybe I am missing some libs. After playing TINA for some time, I got it working using the following methods: 1, download TI-Spice model and demos for INA592, OPA211 respectively, 2, open the demos, copy the component OPA211 into the INA592 demo, 3, reconnect the lines, I found that the flat passband is approximately ~25 KHz, which is pretty low, maybe I should use the difference amplifier from ADI and try its combination with OPA211. Thanks! BTW: I could not use the rich format editing and do not know how to upload the tsc file and the AC results. TINA is a little bit hard to play with (I do not know how to add labels as I do in ngspice and ltspice), and the wiring function is hard to ``follow''. Anyway, thank you!

0 Michael Steffes over 4 years ago in reply to Benyuan Liu

Guru 47535 points

Well TINA is pretty easy once you get used to it - use "cntl -T" to open a text window to document in the circuit.

I have to save files in earlier versions, I had saved the INA592 version in V9. Yes, the bandwidth is low with that but I got i up to about 600kHz by increasing that R1 up to about 10k. Using the AD8273 and R1= 100ohm is about 9Mhz BW. So that actually might be too fast, the file is attached as it is actually quite a job to import the spice model into TINA. This one is saved as a V7 file

AD8274 with OPA211 current source.TSC

0 Thomas Kuehl over 4 years ago in reply to Benyuan Liu

TI__Guru* 93105 points

Hello Byliu,

It has been a while since we have heard from you. Hopefully, you have the information you need to realize the precision voltage-controlled current source. If possible, please close this e2e inquiry.

Thanks to Michael Steffes for all his useful contributions on the subject.

Regards, Thomas
Precision Amplifiers

0 Benyuan Liu over 4 years ago in reply to Thomas Kuehl

Intellectual 595 points

Thanks. Do I need to close this e2e inquiry manually?

0 Michael Steffes over 4 years ago in reply to Benyuan Liu

Guru 47535 points

Yes, at the bottom you should see a button saying "resolved"

Amplifiers

Amplifiers forum

OPA211: Build a high precision voltage controlled current source using a difference amplifier