2
\$\begingroup\$

I have a simple op-amp circuit (picture) designed to drive a capacitive load (~1nF) with a variable ac voltage up to 40Vpp and variable frequency up to 150kHz. The circuit takes signal from a function generator, amplifies it with a gain of 2, and passes it to the load. The op-amp used is an LTC6090, with rails connected to a (9V in) +24V, -24V DC-DC converter than can supply 21mA.

The circuit works well for the above requirements, but I need to limit its maximum output current to 2mA for safety reasons. The danger is that the user could bypass the capacitive load and get a shock.

I have considered a current sense amplifier interfaced to the output disable pin of the op-amp (difficult with bipolar signals), a 10k resistor in series with the load (limits the frequency by forming a low-pass filter with the load capacitance), a 10k resistor in series with each of the supply rails (spoils op-amp performance and adds noise) and I'm running out of ideas. I've also considered constant-current diodes, but I'm pretty unfamiliar with these and I'm unsure of their limitations in ac circuits (any ideas?!). Also, I can't seem to find any DC-DC converters at these voltages that have small enough maximum currents to limit it inherently.

Does anyone have any ideas for how to limit the current in a neat way, without impacting the voltage and frequency performance of the circuit? Thanks in advance! enter image description here

\$\endgroup\$
  • 1
    \$\begingroup\$ I'm not convinced your circuit could deliver 40Vp-p into that load if you limit the output current to 2mA. In any case, 40Vp-p is unlikely to hurt anyone. \$\endgroup\$ – Finbarr Oct 24 '17 at 10:42
  • 1
    \$\begingroup\$ According to this handy dandy calculator keisan.casio.com/exec/system/1258032632 at 150khz that load is around 2.125K... that means you need closer to 10mA to run that circuit up there.. \$\endgroup\$ – Trevor_G Oct 24 '17 at 13:48
  • \$\begingroup\$ Thanks both, you make great points :). Maybe I will need to accept a drop in the circuit performance. Say I drop the max frequency to 20kHz, do you have any neat ideas for limiting the output current? \$\endgroup\$ – Alex Oct 24 '17 at 16:13
  • \$\begingroup\$ Other than current limiter circuits in the +-24V lines.. no. No idea how they will affect linearity though. Though you could actually force the PSU up to its limit by putting in a dummy load at 17mA. \$\endgroup\$ – Trevor_G Oct 24 '17 at 21:07
  • 1
    \$\begingroup\$ Who's decided on the 2mA current limit requirement on the output and why? \$\endgroup\$ – Finbarr Oct 25 '17 at 0:04
1
\$\begingroup\$

I have found a very elegant (I think!) alternative to the current limiter circuit, in current-regulating diodes (CRDs).

Putting two CRDs, rated at x mA regulating current, back-to-back in series between the op-amp output and the load, works very well to limit the current to x mA, without impacting the ac performance. The op-amp of course struggles to drive the load when the frequency gets high, but this is unavoidable regardless of the current limiting method.

I used Semitec E-452, 4.5mA regulating diodes.

Highly recommended!

\$\endgroup\$
0
\$\begingroup\$

Your traditional current limiter should work maybe sort of. Values shown here are not overly meaningful.

schematic

simulate this circuit – Schematic created using CircuitLab

Alternatively you could dump more power from the supplies to bring them close to their current limits.

Issue with all of this is you need to size R1 and R4 to allow 2mA plus the op-amps ambient current which is also in the same ball park plus or minus a bit. As such, the 2mA limit will be foggy.

\$\endgroup\$
  • \$\begingroup\$ Thanks Trevor, that's really helpful. I think this current limiter circuit (with the appropriate values) could be a nice solution. And yes, limiting the current to less than the ambient current of the op-amp seems a bit counter-productive, so I would have to relax the constraints! \$\endgroup\$ – Alex Oct 25 '17 at 11:09
  • \$\begingroup\$ If you are really picky about the value you can make the small resistors 10-turn 1K pots. \$\endgroup\$ – Trevor_G Oct 25 '17 at 12:22

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.