I've attempted to make a high-voltage voltage-controlled current-source based on this design from John Caldwell, but for some reason, it fails with high loads. The design is essentially a Howland current pump that's been bootstrapped to increase the output voltage range.

When I attach a 10kohm or smaller load to the output, everything works perfectly. The current output mirrors the input voltage as long as the R10 rheostat is set reasonably. However, when I attached basically anything larger than 10k, the output suddenly pushes itself to one of the rails. It's a sudden change that happens. When slowly increasing the load with a rheostat, it suddenly goes from normal operation into the state where the output is at one of the rails.

I'm really confused as to why this is happening. I've tried adjusting the R10 feedback resistor when it goes into the bad state, but it doesn't change anything. I'm at the point where I'm not sure what I should be looking for when debugging this circuit. Could it be an issue with the bootstrapping? Is it an issue with how I'm controlling the gain with R10?

Let me know if there's any other information I should provide or there's any useful measurements I can take to get a better understanding of why this is happening.

Some possibly important part numbers are missing from the schematic:

OP1 = INA149 (difference amplifier)

OP2 = OPA140 Schematic

  • \$\begingroup\$ What voltages are VCC2 and VEE2? \$\endgroup\$ Apr 5, 2022 at 16:52
  • \$\begingroup\$ @KevinWhite Theoretical or measured? If measured, I'll need some time to get that. \$\endgroup\$ Apr 5, 2022 at 16:54
  • \$\begingroup\$ @KevinWhite In the bad state where the output goes to +40v(or -), VCC2 is 40v, VEE2 is ~12v. It doesn't seems to me like the issue is with the bootstrapping. \$\endgroup\$ Apr 7, 2022 at 11:57

1 Answer 1


Problems I see:

  1. R10 easily lets you set huge gains that will just saturate the output.

  2. The output to the load is relatively unprotected. Shorts to voltages far away from the output voltage will exceed absolute maximum ratings of both OP1 and OP2. They're fairly easy to destroy that way.

  3. The bootstrap is simplistic and loses headroom close to supply voltage extremes.

  4. The bootstrap loads the output asymmetrically, depending on operating point relative to the +/-48V supply range, and introduces distortion.

All those conspire to make it easy to drive the circuit into a stable operating point that is functionally useless, i.e. makes it "stuck". Never mind the unprotected nature of the load connection.

The input voltage is developed across the gain resistors R10+R106 (see below).

With R106=0 (i.e. shorted), R10's relative wiper position \$k\$ (from 0 to 1) relates to gain as follows: \$k = 1-1/G\$.

Then, R10 adjusts the voltage gain from 1 to infinity, and quite a bit of its adjustment range will just saturate the output. You should limit the gain for ease of experimentation. R106=100 will limit the voltage gain to 10.

For reference, the circuit without the bootstrap is:


simulate this circuit – Schematic created using CircuitLab

I would modify the bootstrap circuit to load the output less, and in a way that doesn't add distortion, as well as have maximum possible headroom.

  • \$\begingroup\$ So it seems like 1 & 2 aren't directly causing my problem here as long as I don't short stuff and set the gain conservatively. For 3, I'm okay with losing headroom at extremes, unless it's causing my issue. With regards to your 4th point and loading the output less, how would I go about doing that? Are there any resources on improving bootstrapping you could point me toward? Thanks for your help. \$\endgroup\$ Apr 5, 2022 at 19:20

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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