Trying to measure very small ripple voltages, but can't.

Audioguru again

Joined Oct 21, 2019
6,672
The value of C40 is way too low and is allowing the output of C40 to have a high impedance at hum frequencies allowing wiring to pickup 60Hz mains hum. If the hum is at 120Hz then the filtering of the bridge rectifiers will affect it.
 

ci139

Joined Jul 11, 2016
1,898
If the hum is at 120Hz then the filtering of the bridge rectifiers will affect it.
affect it through . . . which wires ? op amp input? output? power rails??
_________________
there is an "interesting routing of the GND line p3 p4 near the p5 - ? the GND to the op amp takes a long route to the opposite direction - making it kind of an antenna -- if i interpret your schematic correct AND . . . if it anyhow resembles to the actual trace routing of the PCB ? or you just draw it that way for your own convenience . . .
. . . the "signal returns" should ?? group together near the GND-s of the 3-pin regulators
. . . options for Your "Line Out" routing depend a bit what's on the other side
 
Last edited:

Audioguru again

Joined Oct 21, 2019
6,672
Maybe the main filter capacitor following the rectifier bridge is low value or defective and is allowing a 120Hz level that is so high that the 12V regulator drops out of regulating at 120Hz.
Maybe the high level of 120Hz is feeding into nearby low level audio wiring.

It is important to know if the hum is 60Hz or 120Hz. Which is it?
 

Thread Starter

Dynaman

Joined Jan 17, 2008
94
Capacitor ESR is also an issue in filter circuits.
Try using a very low ESR capacitor(s). If that doesnt help, then the problem is with something else.

Did you mention how you are noticing the problem if you cant measure the ripple?
What made you think the ripple was a problem in the first place?
Subbed on board PS for external bench supply.
 

Thread Starter

Dynaman

Joined Jan 17, 2008
94
Maybe the main filter capacitor following the rectifier bridge is low value or defective and is allowing a 120Hz level that is so high that the 12V regulator drops out of regulating at 120Hz.
Maybe the high level of 120Hz is feeding into nearby low level audio wiring.

It is important to know if the hum is 60Hz or 120Hz. Which is it?
Both but 120 is stronger looking at the analyzer.
 

Audioguru again

Joined Oct 21, 2019
6,672
Ripple at the output of the power supply is 120Hz because it is full-wave rectified.
60Hz is electricity in your home "sprayed" into sensitive high impedance audio inputs by the wiring all over the house, or by a ground loop.
 

MrAl

Joined Jun 17, 2014
11,389
Subbed on board PS for external bench supply.
Well no, i meant how did you know there was ripple not how you corrected it.
How did you measure the original ripple or know it was there before you subbed in the bench supply. What made you want to sub the bench supply.
 

ci139

Joined Jul 11, 2016
1,898
manage the prev. posts first

a random side-note / remark :: 60/120Hz line noise - - - is - - - a - - - bug . . . . . . . . . it of course depends on what is the output capability your transformer + how much you use from it . . . to "get rid" of it - you need to introduce capacities about 10-s of mF (10'000 xN µF) there are a lot of possible filters to manage with lesser . . . . . . but if the power demand of your output increases everything "filter" becomes insufficient -- leading to requirement to introduce buck/boost , fly-back (followed by LDO)
 
If you can listen to the hum, you should be able to measure it on your scope:

-Use a 1X probe
-AC couple your scope, and set it at the highest sensitivity, I gather it is 5 mV/div
-Set your scope to "line" synchronization, and set the timebase at 2 or 5 msec/div.
-Set the scope averaging to 4 or better 16 times.

If it is line-related hum riding on the DC, you will clearly see it.

Of course, the ground probe in your scope is an excellent antenna, picks up a lot of stray fields. Use the shortest possible ground probe length. There are many hacks, using paperclips and other things, to do that.
The following video shows the effects of ground leads. Granted he is talking about high frequency, but good grounding techniques still apply. The paperclip wire hack appears from 6:45

 
Last edited:

Thread Starter

Dynaman

Joined Jan 17, 2008
94
Well no, i meant how did you know there was ripple not how you corrected it.
How did you measure the original ripple or know it was there before you subbed in the bench supply. What made you want to sub the bench supply.
Oh sorry.. connected the amplifier to a spectrum analyzer to try to find where the audible hum was coming from.
 

MrAl

Joined Jun 17, 2014
11,389
Oh sorry.. connected the amplifier to a spectrum analyzer to try to find where the audible hum was coming from.
Ok then if you can discover the sensitivity of the spectrum analyzer then you should be able to build a little op amp amplifier to amplify the AC component and measure it with a meter or scope.
To make it simple, you could probably start with a gain of 10, then 50, then 100, until you start to see the AC component. You can use capacitor coupling so you dont have to worry about the DC offset affecting the meter reading, which may measure part of the DC too.
 

Thread Starter

Dynaman

Joined Jan 17, 2008
94
Your GND goes to TF tap
Your Full Period Rectifier goes immediately to "2.2mF" ** . . .
. . . and to "3-terminals"

** -- there might be some resistance - either before and/or in series - to protect the rectifier diodes and
possibly the components "down the flow" + *** . . . or combined filters

In addition to that the electrolytic cap.-s require ceramics for High frequency range improvement...
...there ought to be *** a small resistor before the 3-terminals and
another .1u ceramic just at it's input to improve the transient response

nothing serious - just the potential noise sourcses . . .
Can you explain the small resistor purpose?
 

ci139

Joined Jul 11, 2016
1,898
there was an article about it (stability) -- but i do not remember in which exact context so i could look it up
but the apx. idea behind is that if you enable power to oscillations they tend to occur the small resistor even up to 100Ω cuts the energy off (works as a damper)
there should be smaller capacitors in parallel with the electrolytic ones . . . (misty) . . . but it worked for custom power supplies made from discretes
well something https://www.rs-online.com/designspark/emi-filters-for-low-power-dc-dc-converters , https://www.powersystemsdesign.com/...ments-and-filter-interactions-part-ii/18/5475 , https://cds.cern.ch/record/2038629/files/265-289-Kunzi.pdf

other
https://www.electronicdesign.com/po...e/21808839/3-ways-to-reduce-powersupply-noise
 
Last edited:

DarthVolta

Joined Jan 27, 2015
521
Greetings AAC'rs: I have an audio preamplifier that turns out needing a very quiet power supply in order for it to perform as quiet as I would like. Attached is the bipolar power supply for it. When I had added a second set of filter caps to the power supply, the noise drops about 17 db improving the preamplifier performance significantly. My problem is, using my scope, I can't see the ripple on the output of the regulators either way. Any suggestions on measuring minute amounts of ripple on the output of the regulators? Any help appreciated.
I had the same problem on my stereo...until I cranked it up..for ages I thought what was that loud to me, should be visible....but it was hidden in a few mV of noise...until i turned it up a lot more.
 

ci139

Joined Jul 11, 2016
1,898
i reworked the PC speakers power supply . . . the 9V.rms~AC transformer one - but the added inductors and caps +more (as infact the "rectifier" was a 3 stage one and became in total as complex as the PC speaker design ... if not more ...) took so much space i couldn't fit it to the original casing nor to the wall adapter nor the speakers one . . . it worked (better) but i dropped it just because i didn't found any suitable casing for the "improved adapter"

the point is that there will be no noise with TF+rectifier when they're working way below their MAX ratings = the problems arise when you put the TF to it's limits so it cannot provide the "continuous power" . . . there are quiet switched regulator topologies though (performing well near max ratings)
 
Last edited:

Audioguru again

Joined Oct 21, 2019
6,672
I do not think you said what is the input signal source. If it is an electret microphone then the resistor that powers it MUST NOT be connected directly to the power supply voltage, but must be fed from an RC filter.

Does the hum disappear when the input signal source is removed and the input is grounded?
 
Top