1
\$\begingroup\$

I'm building a 3.3V 400mA power supply around an ultra fast buck reg (3 MHz or 1.6 MHz, LM2734Z/X.) According to simulations, the ripple will be <2mVp-p. The noise floor of my scope is about 500µVp-p on 5mV/div (the minimum setting), so how do I measure such a small signal?

I'm thinking I'm going to need an amplified probe of some kind, but I suspect these will be very expensive.

I was thinking of fine tuning the trigger and using the averaging function to remove the noise from the signal, but would this be adequate for power supply ripple? Scope is a HP 54501A, 100 MHz -3dB b/w, 10 MS/a digitising, quite an old scope but only cost £150.

\$\endgroup\$
  • 1
    \$\begingroup\$ I guess I have to ask, if your noise ripple is less then 2mVp-p, why do you need to see it on a scope? Seems to me like this is plenty small and there isn't much reason to show it. \$\endgroup\$ – Kellenjb Nov 2 '10 at 1:12
  • \$\begingroup\$ Guess I didn't read all of the way through Andrew's post first. I agree with him. \$\endgroup\$ – Kellenjb Nov 2 '10 at 1:14
  • \$\begingroup\$ you sure move fast on projects. It was just a couple days ago that I recommended going with a buck. Must be nice to jump on projects this fast. \$\endgroup\$ – Kellenjb Nov 2 '10 at 1:18
  • \$\begingroup\$ are you expecting the ripple to be at a frequency in the MHz? \$\endgroup\$ – Kortuk Nov 2 '10 at 2:03
  • \$\begingroup\$ I've been working on the same buck for a few weeks now. \$\endgroup\$ – Thomas O Nov 2 '10 at 7:52
1
\$\begingroup\$

Just an idea... Measure with two probes and use the add function for the two channels :)

\$\endgroup\$
  • 1
    \$\begingroup\$ This will probably get lower resolution though, due to the need to put both channels at 1V/div. \$\endgroup\$ – Thomas O Nov 2 '10 at 13:05
2
\$\begingroup\$

I have a similar scope. Not a bad little unit but I have to take the front panel off and clean all the buttons. :-)

You've already figured out that accumulate mode will give you a good idea of the "envelope" of the ripple over time. In order to get a clearer picture than that you'll need either a better scope or amplified probes, but I'm not sure why you'd need to worry so much about that ripple. Is this merely academic or are you trying to do something unusual?

\$\endgroup\$
  • \$\begingroup\$ I just want to see this ripple. \$\endgroup\$ – Thomas O Nov 2 '10 at 7:53
  • \$\begingroup\$ My 54501A almost died on me. It kept power cycling a few months back and wouldn't work properly (if it worked, the trace was always maxed out.) Then I got a can of compressed air and blew it out. Working fine ever since. \$\endgroup\$ – Thomas O Nov 2 '10 at 7:54
  • \$\begingroup\$ Nice. Mine needs a good cleaning, I have to play with the buttons to get any to register. I have a Tek 7854 too which is more just for cool factor than anything else, but since it's one of those old chassis scopes it can do some pretty amazing stuff with the right modules. \$\endgroup\$ – akohlsmith Nov 2 '10 at 13:22
  • \$\begingroup\$ I would check the connection to the keyboard PCB is secure. The connector is very long (about 1.5ft) and travels across the mainboard to the CPU. \$\endgroup\$ – Thomas O Nov 2 '10 at 15:58

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.