I have built an analog circuit PCB that I want to debug. It has to get a bipolar power +-5V and an input sinus signal. The signal on the CUT (circuit under test), has buffered inputs (but a common ground on an entire PCB).

Only issue is that I don't have a dual PSU at hand (only a single output one). Thus I got an isolated +-5V DC/DC converter, et decided to try with that.

Circuit basically works, but I've noticed that signal generator influences the power rails (basically generated sinus wave gets 'induced' directly into the power rail as noise with much lower amplitude).

So I've figured something is not right how the equipment is connected together. Namely the return path could cause this noise?

This is what I think is going on:


simulate this circuit – Schematic created using CircuitLab

The PSU and SignGen GND's are effectively connected to the mains Earth.

1) I was wondering If I could reduce/remove this artifact when I connect signal generator ground to the PSU output's 'Negative' terminal ?

2) Should I make some other connection to remove possible ground loop ?

3) How would I need to connect an oscilloscope to avoid making a second ground loop to mains earth again ?

  • \$\begingroup\$ How do you measure or see this noise? Are you use a scope or connect your board with a computer? \$\endgroup\$ – M KS Nov 24 '18 at 16:57
  • \$\begingroup\$ @MKS True notation, That's I guess related to my third point. But yes, oscilloscope. Ground lead is again tied to chassis. I've used this high frequency coil-like-spring-gnd-connection, cause having an usual Gnd crocodile wire, was able to pickup a lot of garbage. \$\endgroup\$ – crypton Nov 24 '18 at 20:19

The DPU01M-5 is a 1 watt unregulated voltage output device so this means that if the load current taken by your circuit is influenced by the signal at the input (quite a normal thing to happen I might add) then you will see signal artefacts on the power rails.

If you read the data sheet for the DPU01M-5 it has a load regulation value of 10% and, the smallprint says this: -

4.Load regulation is measured from 20% to 100% rated load.

So, if your circuit is drawing a DC current of (say) 60% of the rated current of the DPU01M-5 and there is an AC content to the supply current of +/-40%, then the output voltage will "waver" around by 10% and this is the likely cause of some of what you see.

| improve this answer | |
  • \$\begingroup\$ Hmm, I guess I have noticed that. When measuring (with an oscilloscope) right at the power pin of any of the op-amps, I see fluctuation at the pin (with a decoup. cap) which could be resulted from varying current consumption - when signal gen is OFF. Circuit btw, from power supply, is shown to consume around 40mA, need to keep PSU at 11.4V at this load to have ~5V rails. One of my trials was to add separate regulators after each +- outputs, but making a negative rail became a bit difficult (thus a separate topic in itself) - but could this help me more ? \$\endgroup\$ – crypton Nov 24 '18 at 20:25

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.