1
\$\begingroup\$

OPA1611 without C15 (1 nF) between in- and output of op-amp, measured in "TP-PRZEWOD" OPA1611 without C15 (1 nF) between in- and output of op-amp, measured in "TP-PRZEWOD"

Schematic1 Schematic 1

Schematic2 - supply Schematic 2 - supply

OPA1611 with C15 (1 nF) between in- and output of op-amp, measured in "TP-PRZEWOD" OPA1611 with C15 (1 nF) between in- and output of op-amp, measured in "TP-PRZEWOD"

enter image description here PCB layout

enter image description here Output of op-amp

I have tested:

  • 4 different op-amps: TL071, OPA277, OPA1611, and uA741
  • different capacitors between in- and output of op-amp: (22pF, 33pF, 47pF, 100pF, 1nF, 10nF, and 100nF)
  • MOSFET driven directly by op-amp and via an NPN transistor; there are no differences between circuits.

Always oscillations, tested:

  • R2: 1 kΩ and 10 kΩ
  • R5: 47 Ω and 100 Ω

About the circuit: it's a tool for testing laboratory cables (banana plug).

X1 - a connector for the power supply (5 V, 16 A). Aimed DC/DC 5-12/12 V for op-amp supply.

C14 - 1000 μF

CSR-2.0-R010 - 10 mΩ shunt resistor

P-MOSFET - IXTH120P065T.

R_I - selectable resistor 10& ohm;, 50 Ω, 80 Ω, 100 Ω for 1 A, 5 A, 8 A, 10 A testing current.

PR1 - calibration 1 mA across T1.

Is it problem with the op-amp, layout, or something else?

enter image description here OPA1611 + 1 nF between in- and output of op-amp, measured in "TP-PRZEWOD"

enter image description here OPA1611 + 1 nF between in- and output of op-amp, measured in "TP-PRZEWOD"

\$\endgroup\$

3 Answers 3

Reset to default
1
\$\begingroup\$

You've tested various opamps and feedback caps, and that didn't solve the problem.

Next suspect would be inductance in the FET drain but you put a snubber so that should be fine. Well you could always reduce R34 to a few ohms, but frequency is too low for FET oscillation, so it's probably not that.

I see a huge split in the ground plane which is exactly in the wrong place...

enter image description here

Current will flow through the path marked as yellow which means it will upset the GND reference of everything on the bottom part of the board.

How to test this hypothesis:

Remove the Negative banana from the PCB plug on the top left and instead connect it to the GND wire of the power supply right on the +5V/GND power connector.

If the circuit no longer oscillates, then it's a layout problem.

\$\endgroup\$
4
  • 1
    \$\begingroup\$ why might they have split the ground plane in the first place? \$\endgroup\$
    – P2000
    Commented Nov 2, 2020 at 21:45
  • \$\begingroup\$ I hope OP answers this question \$\endgroup\$
    – bobflux
    Commented Nov 2, 2020 at 22:09
  • \$\begingroup\$ @bobflux , tested and situation is the same. Still oscillation. \$\endgroup\$
    – wrt503
    Commented Nov 3, 2020 at 8:42
  • \$\begingroup\$ @P2000 , my friend said, that it will be better than ground plane on all PCB without split. \$\endgroup\$
    – wrt503
    Commented Nov 3, 2020 at 8:42
1
\$\begingroup\$

It's a common problem when trying to design a constant current source using a MOSFET as the output device. It is usually fixed by increasing the value of C15 from 1 nF to something higher (maybe 10 nF): -

enter image description here

The problem is caused by the extra pole introduced when driving the MOSFET gate-source capacitance. The extra phase shift introduced by the MOSFET gate-source capacitance and the output resistance of T2 (emitter follower) is enough to degrade the phase margin of the op-amp to the point where it naturally oscillates when feedback is applied via R2 (the input resistor to the op-amp.

\$\endgroup\$
4
  • \$\begingroup\$ Thanks for reply. C15 I have tested from 10pF to 100nF. But without positive results. Also R2 I have tested as jumper (0805 0R), 1k, 10k, but also nothing help. T2 - się have tested without it. (Mosfet driven directly from opamp output, but situation is the same. All time oscillation. \$\endgroup\$
    – wrt503
    Commented Nov 2, 2020 at 20:25
  • \$\begingroup\$ What do9es the voltage look like at the top of R26 where it connects to the Fuse F1. \$\endgroup\$
    – Andy aka
    Commented Nov 2, 2020 at 21:01
  • \$\begingroup\$ how I can put photo from imgur? I cant find information how to do this... \$\endgroup\$
    – wrt503
    Commented Nov 3, 2020 at 7:59
  • \$\begingroup\$ imgur.com/cXoz0Ua - here it is. @Andy aka \$\endgroup\$
    – wrt503
    Commented Nov 3, 2020 at 8:45
0
\$\begingroup\$

Ok. First photo, you can see that laboratory power supply is turned on, but output is off. On osciloscope you can see oscilattion same as I have on output of opamp and on TP-przewod.

enter image description here

I have disassembly aimtec dc/dc converter and pluged linear power supply on 317/337. Its too much better than on aimtec.

Amplitude is now about 400uV, before on aimtec was 7mV!

Also now scale is 1mV/div.

[enter image description here enter image description here

\$\endgroup\$
1
  • \$\begingroup\$ On this video you can see what power supply can do - output is off. Probe of osciloscope is connected to TP-przewod: youtu.be/WpYM72BdQmo \$\endgroup\$
    – wrt503
    Commented Nov 3, 2020 at 11:07

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.