|
Vintage Test Gear and Workshop Equipment For discussions about vintage test gear and workshop equipment such as coil winders. |
|
Thread Tools |
16th Jun 2022, 5:50 pm | #1 |
Pentode
Join Date: Sep 2014
Location: Stuttgart, Germany
Posts: 176
|
R&S NU002 power supply hum
Hi all,
I have this R&S NU002 transmitter that probably served as driver for a larger RF amplifier. It has a power supply attached that gives me a bit of a headache, producing hum seemingly out of nothing. Attached is a pic of the power supply and a PDF with the circuit diagram. Its using an LM2575 step-down regulator to produce +12 and -12V for the transmitter. The hum is on the -12V line only, whereas the +12V line is clean although it has a substantially higher current draw. I connected the power supply to a current sink to make measurements. In the attached screenshot you can see the hum on the output line (pin 2 of choke L21, white curve), with a lot of spikes around that correspond to the 50kHz switching action of the regulator (the red band on top is the switching voltage at pin 2 of the regulator). The green curve is the input voltage to the regulator at pin 1. The hum is clearly synchonized with line frequency, but I cannot find any reason for the regulator to let it pass. The feedback voltage at pin 4 does not carry the hum! With a line regulator I would suspect the large capacitor, but the step-down regulator in this circuit should not let it pass. I checked the capacitors anyway, the readings correspond to those made on the +12V part. The waveforms on the +12V part are identical except for the output line. Any ideas where this hum could come from? cheers Martin |
17th Jun 2022, 10:21 am | #2 |
Octode
Join Date: Dec 2008
Location: Wincanton, Somerset, UK.
Posts: 1,757
|
Re: R&S NU002 power supply hum
Try disconnecting the ac feed to the +12V bridge rectifier, and see if the problem persists.
John |
18th Jun 2022, 10:11 am | #3 |
Pentode
Join Date: Sep 2014
Location: Stuttgart, Germany
Posts: 176
|
Re: R&S NU002 power supply hum
Hi,
I forgot to mention the hum is only present at currents above 100mA. John, cutting the AC feed would require desoldering the rectifier. Do you suspect currents in the +12V supply inducing the hum? The only common point between both supplies is the ground-plane. When testing the -12V supply the output of the +12V supply was open. Its rectifier did not heat up. cheers Martin |
18th Jun 2022, 11:42 am | #4 |
Octode
Join Date: Dec 2008
Location: Wincanton, Somerset, UK.
Posts: 1,757
|
Re: R&S NU002 power supply hum
OK. But you're looking for some kind of second-order effect. Are those blips on the -12V about 10mV p-p? Do they change with load current? (above 100mA). Are they present on pins 3 & 5? I mentioned the rectifier as the blips look like the kind of current pulses associated with the main smoothing caps charging. Hence possible induced cross-talk.
John |
18th Jun 2022, 12:21 pm | #5 |
Nonode
Join Date: Oct 2008
Location: Warsaw, Poland and Cambridge, UK
Posts: 2,669
|
Re: R&S NU002 power supply hum
I'm with John on this. Those dips in the output correspond precisely with the times that the input capacitor is charging. I'd be looking for a dodgy ground connection somewhere: I think the path that ground current takes to the input capacitor is somehow shared with the ground you're measuring from, but not with the ground reference that the regulator is using for its feedback, hence the measured pulses at the output.
Chris
__________________
What's going on in the workshop? http://martin-jones.com/ |
19th Jun 2022, 11:48 am | #6 |
Pentode
Join Date: Sep 2014
Location: Stuttgart, Germany
Posts: 176
|
Re: R&S NU002 power supply hum
I'm trying to add some quantitative stuff... with +30° in the lab. This morning I used the small scope ;-) with all ground leads fixed at "GND2" location.
Contrary to what I said before, the hum appears as soon as current is drawn from the circuit: approx. 5mVpp @ 100mA, 20mVpp @800mA (rated power). I checked the supply with current draw alternating between 200 and 400mA at 1kHz rate, see the pics (first synced with switching frequency, second synced with line). Looks OK for me. Then I checked the other pins of the regulator:
I made a test where I supplied DC into C13 of the +12V supply, with the intention to null the current on the AC line. No change in hum intensity at all. Not very conclusive right now... could it be a design flaw of R&S for such an equipment? The whole electronics look very well designed. When I got the unit off an internet aucution, one of the bridge rectifiers was faulty - actually a shorted diode. The transmitter still worked, but the transformer and rectifier were running veeeery hot. cheers Martin |
20th Jun 2022, 1:26 pm | #7 |
Octode
Join Date: Dec 2008
Location: Wincanton, Somerset, UK.
Posts: 1,757
|
Re: R&S NU002 power supply hum
OK, so you measure considerably less "hum" on pins 3 & 5 than on the actual output? Two possibilities:
- the pulse are being picked up by the "loop" formed by your oscilloscope probe / earth lead. - Inductor L21 is picking up a stray magnetic field in a similar scenario. As you have less "hum" measured on the input to L21 it's pretty certain that the regulator is not to blame. John |
20th Jun 2022, 4:04 pm | #8 |
Pentode
Join Date: Sep 2014
Location: Stuttgart, Germany
Posts: 176
|
Re: R&S NU002 power supply hum
Hi John,
thanks for your input. I can see the hum on the power lines inside the transmitter, a good deal away from the power supply. With the ground lead attached to the chassis close to where I measured. So I think the hum is effectively there, and not a measurement artefact. Inductor L21 is mounted on the inner side, close to the transformer, whereas L11 (for the +12v supply) is on the outer side. That could account for it, but I wonder if the waveform shouldn't be smoother in that case? Where and how should I make further measurements to get meaningful results? I have a differential plugin for my scope, would that be useful? cheers Martin |
20th Jun 2022, 5:04 pm | #9 |
Octode
Join Date: Dec 2008
Location: Wincanton, Somerset, UK.
Posts: 1,757
|
Re: R&S NU002 power supply hum
Do you think that the "cooking" which the mains transformer suffered might have given rise to a shorted turn or similar? That might account for a change in leakage flux. Long-shot.
With a differential probe I suppose look on each side of L21. Try interposing a piece of steel plate or similar. John |
20th Jun 2022, 8:42 pm | #10 | |
Pentode
Join Date: Sep 2014
Location: Stuttgart, Germany
Posts: 176
|
Re: R&S NU002 power supply hum
Quote:
cheers Martin |
|
21st Jun 2022, 8:13 am | #11 |
Pentode
Join Date: Sep 2014
Location: Stuttgart, Germany
Posts: 176
|
Re: R&S NU002 power supply hum
I checked. If there was a shorted turn or something else that produces current the transformer should heat up noticably even without charge.
The unit ran without charge for 3 hours, and the transformer does not heat up at all. At least not noticeable to the touch... cheers Martin |
2nd Jul 2022, 2:48 pm | #12 |
Pentode
Join Date: Sep 2014
Location: Stuttgart, Germany
Posts: 176
|
Re: R&S NU002 power supply hum
Hi all,
still trying to hunt down the hum on the -12V supply. I made measurement, this time taking car of where the ground leads are attached to (up to now they were attached to the chassis). Attached is a PDF with the results. Both supplies were charged with 500mA. In each supply, when I attach the ground lead on pin 1 of the rectifier I get -15mV "humps" at pin 1 of the corresponding regulator. With the ground lead attached on the negative side (pin 4) of the rectifier the hum at C12/C22 is 25mV in the +12V supply and 10mV in the -12V supply. Voltages at pins 3 and 5 of the regulator correspond. I noticed there is a drop of "hum voltage" across C12/C22 in the -12V supply only. It can be measured directly (ground lead on one side of the capacitor) or indirectly by measuring between rectifier and R21, where I get 15mV instead of 10. The voltage drop at C22 is actually what I get at the output. What I checked so far:
Thanks for any further ideas... cheers Martin |