0
\$\begingroup\$

I'm using two potentiometer in amplifier design for controlling gain and offset. the potentiometer that used in offset control works perfect. but that one used in amplifier resistor have a big problem. When it joined to circuit, output signal become too noisy.

enter image description here

My circuit design in this section as follows: enter image description here

I changed the potentiometer in feedback path, but result is same. Also I cut the offset path and test gain section separate, but output don't changed.

can I help me or explain this problem?

thank you,

|improve this question
\$\endgroup\$
  • \$\begingroup\$ P_ofs1 does not match P_ofs. And it won't control offset either. \$\endgroup\$ – Andy aka Feb 26 '17 at 14:44
  • \$\begingroup\$ that's right. I'm using a buffer between P_ofs1 and P_ofs. that not shown in the picture. However, problem not related to the offset section. \$\endgroup\$ – Saeed Vali Feb 26 '17 at 14:52
  • \$\begingroup\$ Wideband video op amps are very sensitive to capacitance in the feedback path. Your digipot has 15pF capacitance which may be causing the circuit to oscillate, and it only has 1MHz bandwidth - not a good choice for a 0-20MHz amplifier! \$\endgroup\$ – Bruce Abbott Feb 26 '17 at 16:26
  • \$\begingroup\$ @BruceAbbott ok, that's right. but what is the solution for such design? \$\endgroup\$ – Saeed Vali Feb 26 '17 at 21:03
  • 1
    \$\begingroup\$ Use the digipot as an attenuator, and run the op-amp at fixed gain. Your digipots only work with positive voltages between 0-Vcc, so you need to bias the signal going into the pot. If the signal is AC then couple it via capacitors so the digipot bias doesn't affect op-amp offset. Op amp offset adjustment may require negative voltage, which the digipot cannot produce. However you can use resistors to combine the pot's positive output with a fixed negative offset - just make sure the digipot output doesn't go negative. \$\endgroup\$ – Bruce Abbott Feb 27 '17 at 20:07
2
\$\begingroup\$

The digital pot can have significant capacitance from its analogue pins to both ground and the internal digital clocking input. This means two things: -

  • high frequency gain is high and op amp noise gets massively amplified
  • when clocking in data, you get noise injections to the inverting input.

With your non-effective offset control, you will get very little noise injected because the digipot is in series with a resistor, R28.

|improve this answer
\$\endgroup\$
  • \$\begingroup\$ Thank you for your answer. What is the solution? by this fact, any digital pot can not be used in common amplifier. \$\endgroup\$ – Saeed Vali Feb 26 '17 at 21:01
  • 1
    \$\begingroup\$ I use digipots in amplifiers. I use them as pots with the A connection on the op amp output and the B connection to analogue ground. The wiper goes to the inverting input. Gain can be controlled from unity to as high as 16. Try ang get gains higher than this and you are on slippy ground both in terms of temperature stability and noise. Once the gain value has been programmed do not keep any digital line toggling or you will get noise. Choose a digipot with great potentiometer tempco and don't rely on the wiper connection to pass more than a few microamps. \$\endgroup\$ – Andy aka Feb 26 '17 at 22:13
0
\$\begingroup\$

I'm just researching a similar problem where a digital potentiometer is introducing a lot of noise into an audio circuit. This article from Analog Devices (who make digipots) says the SPI clock introduces noise.

https://www.analog.com/en/education/education-library/faqs/faq_if_i_use_the_digital_potentiometer_in_audio.html

|improve this answer
\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.