Suspected high ripple on SMPS 0.628v at 12v

[hafrse]

Hello,

I have a problem with an analyzer which shows spurs from CW 50hz (highest amplitude) , harmonics at 150hz,350hz,etc... suspected the power supply and checked one of the voltages -12V and there is a ripple of about 0.63 volts, the same ripple is on other output voltages. Is this normal ? problems in the mains rectifier electrolytes (180uf 450V) ?
attached a picture of the power supply and trace of the ripple on the -12v.

Thnaks for any input

[bktemp]

That's no ripple but switching noise. It is probably normal if you didn't use proper rf probing techniques (the ground clip of a scope probe is only useable for <10MHz, above that you should use a ground spring or a proper rf connection).
If there is no 50/100Hz ripple on the voltage rails, I doubt this is the source of your spurs.

[Refrigerator]

As bktemp said, that's just switching noise, you might want to check the output caps if you want to start troubleshooting.

[hafrse]

Correct, the problem was a measuring error, no problems with caps or the supply voltages.
However, I moved the smps from the chassi about 10 cm and all the spurs on the analyzer dissapeared!
It seems to be a radiated electromagnetic field from the smps beacause of its age and run time (over 60K hrs) .
 Now I need to do some shielding, any hint?
Thanks

[Refrigerator]

You might want to check if the chassis is grounded properly.
For shielding you can use many things, tin foil should work ok.

[hafrse]

Hello,

Only got 20% better with mechanical shielding, still 50Hz noise visible, picture shows noise at 50Hz from carrier wave.
Could it be the mains  450v caps that are bad or have high ESR?
Thanks

[SeanB]

Put the screw back by the mains inlet connector on the PSU, that is part of the input filtering, and if it is missing you will have radiation from the PSU. that will be from the current flowing through the side panels to get to the mains earth pin in the mains filter, which otherwise travels in a long loop up the board and down on the case.  You also need to have the top cover back on, it is part of the shielding, and the screws must all be in.

[hafrse]

Put the screw back by the mains inlet connector on the PSU, that is part of the input filtering, and if it is missing you will have radiation from the PSU. that will be from the current flowing through the side panels to get to the mains earth pin in the mains filter, which otherwise travels in a long loop up the board and down on the case.  You also need to have the top cover back on, it is part of the shielding, and the screws must all be in.

Thanks for your respond, I did that before , It get little better , but still 50Hz noise visible with its harmonics, I will try to replace the large 180uf 450V caps (2 of them) on the filtering side, I can feel upward curved deformation on the top of them with my finger (not visible in the picture of the first post in this thread) , the instrument had over 60K hrs on its neck. Moving the power supply outside the box all 50hz noise with its harmonics are virtually 0.
Any other things I can try?

Thanks

[SeanB]

If they are bulging replace, will have some effect but also check ESR of the final output caps then, and replace in any case those after the filter inductors on all the rails. Just because they are Nichicon does not mean no failure, with 60k hours probably all, from the smallest to the largest, are in need of replacement.

[hafrse]

If they are bulging replace, will have some effect but also check ESR of the final output caps then, and replace in any case those after the filter inductors on all the rails. Just because they are Nichicon does not mean no failure, with 60k hours probably all, from the smallest to the largest, are in need of replacement.

Agree, I will start with the large Nichicon  ones since the are bulging and   noise is fundemnntal 50Hz, will also check the ESR on the final output caps,  I have now ordered 2 x  220uF 500V 105'C VICHAY LONG LIFE 193, original are at 180uf 450, hope that is ok.
I will report the result, thanks for your help.

[bktemp]

Moving the power supply outside the box all 50hz noise with its harmonics are virtually 0.

What happens if you leave the power supply in the box, but add some insulation material around it to avoid any 50Hz currents flowing through its metal case into the analyser?
If the electrolytic capacitors are the actual problem, it wouldn't make a difference if you put the power supply outside the box or leave it inside.
My guess would be a bad connection to GND somewhere near the analogue frontend, injecting the 50Hz current into the signal path.
It could also be some bad emc filter capacitor near the mains input.
Since you see 50Hz, it can't be anything after the bridge rectifier, because rectifying a sinewave doubles its frequency. Therefore bad capacitors show up as 100Hz + harmonics.

[hafrse]

Moving the power supply outside the box all 50hz noise with its harmonics are virtually 0.

What happens if you leave the power supply in the box, but add some insulation material around it to avoid any 50Hz currents flowing through its metal case into the analyser?
If the electrolytic capacitors are the actual problem, it wouldn't make a difference if you put the power supply outside the box or leave it inside.
My guess would be a bad connection to GND somewhere near the analogue frontend, injecting the 50Hz current into the signal path.
It could also be some bad emc filter capacitor near the mains input.
Since you see 50Hz, it can't be anything after the bridge rectifier, because rectifying a sinewave doubles its frequency. Therefore bad capacitors show up as 100Hz + harmonics.

When I move back the power supply  slowly nearer an nearer , I can see the noise becoming more and more, so it is radiated 50Hz

 you are right, it is 50 Hz and the problem should be before the rectifier, there we have  2 large polyester caps together with 2 coils (input filter ?) , also, could it be the mains socket filter box on the bottom middle side of the power supply ? (please see picture is in the first post)  .

The large 450v caps should be replaced anyway since they are bulging / over 60k hrs  for a preventive maintenance reason.
Thanks

[bktemp]

When I move back the power supply  slowly nearer an nearer , I can see the noise becoming more and more, so it is radiated 50Hz

 you are right, it is 50 Hz and the problem should be before the rectifier, there we have  2 large polyester caps together with 2 coils (input filter ?) , also, could it be the mains socket filter box on the bottom middle side of the power supply ? (please see picture is in the first post)  .

That is interesting, but is still does not rule out a problem at the analyser itself (like an oxidized grounding/shielding connection somewhere).
If the field is so strong, what happens if you put a scope probe near the power supply? Or maybe using a small inductor as a magnetic field sensor. Is there a hotspot somewhere? A SMPS shouln't emit a large magnetic field at mains frequency.
It could be a bad bridge rectifier (turned into a half-wave rectifier) + bad mains cap, but most likely the noise comes from the section near the mains input.
To rule out a magnetic field generated by damaged common mode chokes you could bypass them temporarily (short them to avoid any current flowing through the windings) and see if the noise changes.

[hafrse]

When I move back the power supply  slowly nearer an nearer , I can see the noise becoming more and more, so it is radiated 50Hz

 you are right, it is 50 Hz and the problem should be before the rectifier, there we have  2 large polyester caps together with 2 coils (input filter ?) , also, could it be the mains socket filter box on the bottom middle side of the power supply ? (please see picture is in the first post)  .

That is interesting, but is still does not rule out a problem at the analyser itself (like an oxidized grounding/shielding connection somewhere).
If the field is so strong, what happens if you put a scope probe near the power supply? Or maybe using a small inductor as a magnetic field sensor. Is there a hotspot somewhere? A SMPS shouln't emit a large magnetic field at mains frequency.
It could be a bad bridge rectifier (turned into a half-wave rectifier) + bad mains cap, but most likely the noise comes from the section near the mains input.
To rule out a magnetic field generated by damaged common mode chokes you could bypass them temporarily (short them to avoid any current flowing through the windings) and see if the noise changes.

Short the 2 large chokes on the right side in the picture ? will also try to put a scope probe near the power supply.

[bktemp]

Yes, but short the correct pins (pins on left side and both pins on the right side), othwise you will short out the mains input if you short pins from the left side to pins on the right side of the common mode choke.
I have no idea if it will make any difference, but something in that area must be emitting the 50Hz noise and there aren't many components.
It could also be somthing like the varistors failing and starting to conduct at the peak of every mains cycle, generating all kinds of harmonics.
Radiated 50Hz noise from a SMPS is rather unusual.