Topics

PS503A noisy output, suggestions?


Jared Cabot
 

I'm currently restoring and testing my PS503A 500 series module power supply, and I have found the output is a little noisy compared to what it says in the specs.

I've removed and tested all the capacitors after slowly running them up to their full rated voltages a few times on my bench power supply and none are leaky (actually the leakage is way down in the single digit and very low double digit microamps), and when tested on my HP 4276A LCZ meter, they all test well withing spec, even at or close to their rated voltage (I used the DC bias function to compare readings at 0V and 40V), so I'm pretty sure the capacitors are looking all ok from my limited knowledge.

However, I'm reading around 11-13ish mA P-P ripple on the output (before 'reforming' the capacitors it was around 31mV P-P), where the manual says it should be less than 3mV P-P.
https://i.imgur.com/DMh71qJ.png

I am using a TM5006 mainframe though (I have no other mainframe with a high power slot), could my mainframe having a switchmode power supply have an effect on the ripple compared to a 500 series mainframe with a plain transformer?

Any ideas what I can do to get this ripple down to spec?


John Gord
 

Jared,
Yes, the switching supply is likely to give more noise.
Part of the problem may be ground noise between your oscilloscope and the PS503A. Try winding your probe or cable a few turns through a large ferrite core (perhaps a snap-on core). Both the signal and ground paths go through the core; this acts as a common mode choke and can help reduce ground-difference high frequency noise. (It won't do much for power line frequency noise.)
--John Gord

On Mon, Nov 9, 2020 at 05:23 PM, Jared Cabot wrote:


I'm currently restoring and testing my PS503A 500 series module power supply,
and I have found the output is a little noisy compared to what it says in the
specs.

I've removed and tested all the capacitors after slowly running them up to
their full rated voltages a few times on my bench power supply and none are
leaky (actually the leakage is way down in the single digit and very low
double digit microamps), and when tested on my HP 4276A LCZ meter, they all
test well withing spec, even at or close to their rated voltage (I used the DC
bias function to compare readings at 0V and 40V), so I'm pretty sure the
capacitors are looking all ok from my limited knowledge.

However, I'm reading around 11-13ish mA P-P ripple on the output (before
'reforming' the capacitors it was around 31mV P-P), where the manual says it
should be less than 3mV P-P.
https://i.imgur.com/DMh71qJ.png

I am using a TM5006 mainframe though (I have no other mainframe with a high
power slot), could my mainframe having a switchmode power supply have an
effect on the ripple compared to a 500 series mainframe with a plain
transformer?

Any ideas what I can do to get this ripple down to spec?


Jared Cabot
 

Ah yes, I actually saw that ferrite ring trick mentioned in another TM500 module manual, and it just so happens that I just saved a good sized ferrite ring 'just incase' only a couple days ago. Who ever said hoarding junk never came in useful!

I'll give that trick a go and see what happens.

Also, I'll keep an eye out for a 4, 5 or 6 bay TM500 mainframe too.......


Thanks!
Jared

On Tue, Nov 10, 2020 at 10:47 AM, John Gord wrote:


Jared,
Yes, the switching supply is likely to give more noise.
Part of the problem may be ground noise between your oscilloscope and the
PS503A. Try winding your probe or cable a few turns through a large ferrite
core (perhaps a snap-on core). Both the signal and ground paths go through
the core; this acts as a common mode choke and can help reduce
ground-difference high frequency noise. (It won't do much for power line
frequency noise.)
--John Gord