Author Topic: High speed variable gain amplifier  (Read 1344 times)

0 Members and 1 Guest are viewing this topic.

Offline ricko_ukTopic starter

  • Super Contributor
  • ***
  • Posts: 1158
  • Country: gb
High speed variable gain amplifier
« on: February 26, 2020, 12:41:40 pm »
Hi,
I have to make a 250 KHz variable gain amplifier controlled by a micro.

The input signal can (in extreme cases) swing between -4V and +4V i.e. below GND. Initially I thought of the attached schematic using a digital pot. But that way the wiper (connected to the op-amp negative input) would go negative when the signal goes below ground. So it would go out of specs. And the rare digital pots that can go below ground are expensive.

Then I thought of using a DAC. But similar issue.

The variable high speed gain op-amps tend to be expensive... So what is the best low cost solution?

Many thanks :)



« Last Edit: February 26, 2020, 12:45:18 pm by ricko_uk »
 

Online SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15799
  • Country: fr
Re: High speed variable gain amplifier
« Reply #1 on: February 26, 2020, 02:48:11 pm »
A possible approach to lower cost is NOT to go with a dual supply at all.
To deal with a signal that can go below ground, you can shift it up (and attenuate it) so that it fits within 0/+5V, and then use a single +5V supply for both the digital pot and the opamp. Of course the opamp output will be shifted as well. What is it going to be used for?
 

Offline ricko_ukTopic starter

  • Super Contributor
  • ***
  • Posts: 1158
  • Country: gb
Re: High speed variable gain amplifier
« Reply #2 on: February 26, 2020, 05:28:59 pm »
Thank you Silicon Wizard,
it is used for detecting the amplitude of a sensor.

The sensor generates an AC signal proportional to the quantity measured. That amplitude, depending on the application, can vary between few tens of mV to 4.5V peak to peak - hence the reason for having a variable gain amplifier.

The output is then fed into a high speed peak detector to detect the amplitude and convert it into DC for a ADC.

I did think about shifting it but that would then introduce DC which could then reduce the ADC level to half (unless I add another stage - but because we have up to 42 channels it would impact the small board space available).

Any other suggestion? Or perhaps a refinement on your initial one?

Many thanks!! :)
 

Online SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15799
  • Country: fr
Re: High speed variable gain amplifier
« Reply #3 on: February 26, 2020, 05:43:12 pm »
I did think about shifting it but that would then introduce DC which could then reduce the ADC level to half (unless I add another stage - but because we have up to 42 channels it would impact the small board space available).

If you're going to feed the signal to an ADC, shifting the signal makes even more sense IMO (and is what we usually do)? As most ADCs out there can only deal with positive inputs anyway?
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7044
  • Country: nl
Re: High speed variable gain amplifier
« Reply #4 on: February 26, 2020, 06:11:49 pm »
The variable high speed gain op-amps tend to be expensive
A MCP6S91 is too expensive?
 

Offline ricko_ukTopic starter

  • Super Contributor
  • ***
  • Posts: 1158
  • Country: gb
Re: High speed variable gain amplifier
« Reply #5 on: February 26, 2020, 06:40:35 pm »
Thank you SiliconWizard and Marco,

Marco, that is a nice part! But it is a programmable gain amp not a variable gain. I need variable gain to adjust finely the signal range going into the peak detector and following that into a ADC.

SiliconWizard, you are correct, but the only issue is that as I was saying it would half the range into the DAC unless I add more electronics and the PCB area is limited (we would have to add 42 such stages).
« Last Edit: February 26, 2020, 11:43:44 pm by ricko_uk »
 

Offline Myke

  • Newbie
  • Posts: 2
  • Country: us
Re: High speed variable gain amplifier
« Reply #6 on: February 26, 2020, 09:51:46 pm »
You might want to take a look at an analog multiplier like the AD633. It's not that cheap but should be fast enough for what you want to do. It even has a built in summing stage so you can shift the level to a usable range for a ADC.

EDIT: Oh. If the inputted signal is in the 10s of mV and you want additional gain, you would probably need to follow it up with another amplifier.

What sort of sensor is this that's outputting something in the mV to 4.5V p-p (or is it 8V p-p?) at 250kHz? How fast do you want to adjust the gain and what does the spectrum look like?
Maybe a log amp is more what you're looking for. The LOG114 isn't that cheap but covers the range you're looking for. A series resistor would take care of the current input.
« Last Edit: February 26, 2020, 10:26:45 pm by Myke »
 
The following users thanked this post: ricko_uk

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17428
  • Country: us
  • DavidH
Re: High speed variable gain amplifier
« Reply #7 on: February 27, 2020, 07:29:04 pm »
Plenty of digital potentiometers support bipolar operation with a dedicated digital ground allowing use in AC applications.  There are also multiplying DACs which can operate with bipolar input and output signals in either 2 or 4 quadrant mode.
 
The following users thanked this post: ricko_uk

Online SiliconWizard

  • Super Contributor
  • ***
  • Posts: 15799
  • Country: fr
Re: High speed variable gain amplifier
« Reply #8 on: February 27, 2020, 07:35:50 pm »
Plenty of digital potentiometers support bipolar operation with a dedicated digital ground allowing use in AC applications.  There are also multiplying DACs which can operate with bipolar input and output signals in either 2 or 4 quadrant mode.

Of course, but the OP's point was mainly cost AFAIK. And indeed, when I looked that up a bit, single-supply/+5V min. digital pots are way cheaper. On average, you can get one channel for ~$0.50 (for low quantities), whereas dual-supply, +/-5V min. dig. pots. are usually more in the $3-$4 price range, so that's 6 times more expensive.

If you know of anything like that in the $0.50 price range (which I think was what the OP was after, don't hesitate to give some references.
That'll be the same issue with multiplying DACs IMO. Any IC that can handle over 5V is usually considered in the high-voltage area these days, and much more expensive.
 
The following users thanked this post: ricko_uk

Offline ricko_ukTopic starter

  • Super Contributor
  • ***
  • Posts: 1158
  • Country: gb
Re: High speed variable gain amplifier
« Reply #9 on: February 27, 2020, 08:12:52 pm »
Thank you David and SiliconWizard,
yes SiliconWizard is spot on. Cost is a priority because of the large number of channels each needing one.
But if you do know a part in that price range (~0.50 USD) please share it as it would be very useful!!
Thank you
 

Offline f5r5e5d

  • Frequent Contributor
  • **
  • Posts: 349
Re: High speed variable gain amplifier
« Reply #10 on: February 27, 2020, 09:30:36 pm »
still a few open questions about the signal that may affect the options

what AC frequency? - is it a range, external or generated by the DAQ circuitry?

how noisey is the signal, what's frequency range of the amplitude variation?

what is the required resolution and sampling rate of the final amplitude measurement
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf