strange low frequency switching noice from diy sg3525 smps

[Herschel]

So I made a powersupply for an audio amp using sg3525ic and an ir2110 mosfetdriver
The smps works fine, but the problem is, when i connect an amplifier+speaker to the smps, i could hear strange low frequency switching noice from the speaker output:
 here is a video showing the issue:   https://youtube.com/shorts/UIPuuqmjZPY?feature=share

so to check weather its the noice from the amplifier, I disconnected the amplifier and connected an led to the transformer output(before the full bridge rectifier) and the LED isn't constantly on, its blinking and
 here is a video showing that:    https://youtube.com/shorts/7hwPGjBLxPw?feature=share

this thing only happens when i connect the output to the feedback circuit

[Circlotron]

Despite what looks like a common mode choke after the diode bridge, you should really have a filter choke between the bridge and the filter capacitors. One choke on each rail. Without a choke, the DC output voltage of the inverter is mostly dependent on the peak voltage of the inverter output, and the PWM for voltage regulation can't work properly. When you connect the feedback the loop is unstable because a large change in pulse width does not produce a corresponding change in DC output voltage.

[Herschel]

Despite what looks like a common mode choke after the diode bridge, you should really have a filter choke between the bridge and the filter capacitors. One choke on each rail. Without a choke, the DC output voltage of the inverter is mostly dependent on the peak voltage of the inverter output, and the PWM for voltage regulation can't work properly. When you connect the feedback the loop is unstable because a large change in pulse width does not produce a corresponding change in DC output voltage.

Actually i didn't used a common mode choke, I used one choke for each rail. Also the chokes I'm using have iron powder core. Are iron powder cores suitable for this purpose?

[Herschel]

the DC output voltage of the inverter is mostly dependent on the peak voltage of the inverter output, and the PWM for voltage regulation can't work properly. When you connect the feedback the loop is unstable because a large change in pulse width does not produce a corresponding change in DC output voltage.

I used one choke for each rail, chokes with iron powder core... can this be prevented by adding a capacitor before ther chokes or, is iron powder cores suitable for this purpose?

[Circlotron]

Powdered iron is fine. Looks like you have used 21 material. What is the switching frequency?
Definitely no capacitor before the chokes.

[Herschel]

Powdered iron is fine. Looks like you have used 21 material. What is the switching frequency?
Definitely no capacitor before the chokes.

switching frequency: 60khz
does the right value of inductor matters? mine has a really low inductance, lower than the specified 20uH...

I did used 470uf capacitors before the inductors, later I chaged them to 0.1uf and, the low frequency ticking sound now changed to a high frequency with a cloudy noise
anyway I removed those 0.1uf capacitors and still the same noise... 

I really doubt on the inductance value, with a really low No. of turns, may be i should try rewinding them...

[Herschel]

So I made inductors of 22uH and replaced it with the old ones and seems like the noise have reduced a bit... but not fully. maybe i should try adding a common mode choke at the output of the powersupply?

[Phil1977]

Hearable switching noise often is a symptom of oscillations in the control loop. Did you check that? E.g. by hooking a scope to the optocoupler?

[T3sl4co1l]

I would be inclined to suspect EMI.  Try permutations of 'Y' capacitors. In particular, the one to bridge '-' is probably wrong, but one from there to output may be helpful.  Additional filtering (another CMC?), shielding (stuff everything into a box and filter all wires that pass through it) and testing are likely to prove fruitful.

Tim

[Herschel]

Hearable switching noise often is a symptom of oscillations in the control loop. Did you check that? E.g. by hooking a scope to the optocoupler?

maybe I need to take my smps to my college, They have osciloscopes and i don't own one... So I use LEDs instead to note fading or blinking

[Herschel]

I would be inclined to suspect EMI.  Try permutations of 'Y' capacitors. In particular, the one to bridge '-' is probably wrong, but one from there to output may be helpful.  Additional filtering (another CMC?), shielding (stuff everything into a box and filter all wires that pass through it) and testing are likely to prove fruitful.

Tim

I tried using the Y capacitors, and it did improved the noise under one "condition":

I was testing my smps with a 400 watt class D amplifier board, which is locally assembled one, I found out that the noise is created by that class D amplifier when I connect it to my smps. Does class D amplifers oscillates to create noise?, but after replacing that with a class AB amplifier, which I made, the noise just vanished! (still audiable when I hold my ears really close, covering the tweeters which I'm happy with)

So the *condition, only applies for the class ABs not Ds. And yes, the Y caps helped a lot

Is it a good idea to use a CMC with .1uf mylar filter caps only for the classD amp does it improves the performance and reduce noise?

[ftg]

Could you post a picture of the full SMPS and your test setup?
The rectifier and electrolytic capacitor board looked like it was a separate unit at the end of cables.
Even on 60kHz the wires will have some inductances and currents will flow in funny ways.

[T3sl4co1l]

Yes... a class D amp is itself an SMPS, to which all the same filtering considerations also apply.

In particular, the speaker output may not be well filtered, which puts EMI onto that lead, and by reaction, inversely onto the board itself, and thus up connecting cables to other equipment (and through all these cables, as antennas, into the air).

Filtering options are a bit more limited, because you need to know if the output is filtered at all for one (some class D amps have direct drive to the speaker, depending on coil inductance for SMPS filtering; this requires short wires and usually a ground connection to the speaker frame, and must only be done in self-powered speakers, not over any lead length), and thus whether to provide filtering specific to the outputs; and because the audio input itself cannot be filtered arbitrarily low, because of the audio itself; a common-mode choke can be useful here, filtering the input ground with respect to the shield/enclosure (EMI reference plane) while leaving the signal conductor alone (other than by effect of the CMC, and a little RFI filtering say 1nF + ferrite beads).

Tim

[Herschel]

Yes... a class D amp is itself an SMPS, to which all the same filtering considerations also apply.

In particular, the speaker output may not be well filtered, which puts EMI onto that lead, and by reaction, inversely onto the board itself, and thus up connecting cables to other equipment (and through all these cables, as antennas, into the air).

Filtering options are a bit more limited, because you need to know if the output is filtered at all for one (some class D amps have direct drive to the speaker, depending on coil inductance for SMPS filtering; this requires short wires and usually a ground connection to the speaker frame, and must only be done in self-powered speakers, not over any lead length), and thus whether to provide filtering specific to the outputs; and because the audio input itself cannot be filtered arbitrarily low, because of the audio itself; a common-mode choke can be useful here, filtering the input ground with respect to the shield/enclosure (EMI reference plane) while leaving the signal conductor alone (other than by effect of the CMC, and a little RFI filtering say 1nF + ferrite beads).

Tim

the class D amp I'm using have an inductor + capacitor at its output, This is just a simple IR2153 class D amp


this is the most similar circuit I could find, also the maker of this amp have added some more components to it at the audio input side... and yes, ICs for powersupplies/inverters are used here

[T3sl4co1l]

Ok, output filter choke, that's a start.

Looks to be single-side layout, which means no ground plane, and all the stray inductances wreak havoc on the EMI response in the 10s of MHz where those traces cease being good conductors but stray inductance takes over.

Tim

[Herschel]

Ok, output filter choke, that's a start.

Looks to be single-side layout, which means no ground plane, and all the stray inductances wreak havoc on the EMI response in the 10s of MHz where those traces cease being good conductors but stray inductance takes over.

Tim

Even though, there are some design flaws in the amp, these amps are getting crazy good local reviews in youtube, and those reviews are true. The product seller tested the amp before my eyes, and it sounds great with a really faint humming why? they where using transformers instead of smps. So, really low noise

anyway I'm using this class D amp just to drive the subwoofer. maybe I should try adding CMC+ filter caps before the class D amp, so the noise may not affect other class AB amps connected to that smps

[Phil1977]

I´d propose to check your DIY-SMPS for control loop oscillations anyhow. They easily occur, they often are easy to prevent by adding some dampening into the control loop. But if they occur, they do not only produce noise, they stress all the power components of the SMPS in an extreme way.

It´s just a good advice: If you can check your SMPS with a scope then do it. Check that the control loop does not oscillate between its extremes. Small oscillations in the control loop show that it´s doing its job, but if e.g. the current through the optocoupler LED oscillates between 0 and some saturated value, then something is definitely wrong.

[T3sl4co1l]

Ok, output filter choke, that's a start.

Looks to be single-side layout, which means no ground plane, and all the stray inductances wreak havoc on the EMI response in the 10s of MHz where those traces cease being good conductors but stray inductance takes over.

Tim

Even though, there are some design flaws in the amp, these amps are getting crazy good local reviews in youtube, and those reviews are true. The product seller tested the amp before my eyes, and it sounds great with a really faint humming why? they where using transformers instead of smps. So, really low noise

anyway I'm using this class D amp just to drive the subwoofer. maybe I should try adding CMC+ filter caps before the class D amp, so the noise may not affect other class AB amps connected to that smps

This doesn't mean very much...

Understand that correct operation in one very limited set of functions (in this case, audio amplification) says nothing about operation in a completely different scope (RF emissions).

It's pretty easy to make a class D amp sound alright. That's everything below 20kHz.

It's not even that hard to make one measure well.  Notice these are already two different things, because "sounds good" is already a wide margin above what a typical amplifier might measure as: the gain/flatness might be off by many dB, the distortion might be several percent, the noise and hum background might be noticeable in one setting but not another; and most of all, listeners are notoriously fickle in their perception, it takes an expert listener to properly assess an amplifier (in whatever way a listener can assess one accurately at all, which again, still isn't fantastic compared to electrical measurements).

We're talking about anything from 100kHz to 100MHz, very different ball game -- and absolutely nothing you can hear, unless it gets stuffed sideways into an amplifier, rectified and detected, in which case hum, squealing, beeping, clicking and so on are all possibilities.

Tim

[Herschel]

I think, I've found the problem...    It was the optocoupler!

the original circuit was using a 4n25 opto ic, the feedback voltage going into the opto ic was stable when I tested it with an LED,

But when I tested the output from the optocoupler, the led was blinking, also the blnking speed+ frequency of that weird sound changes/increases when I connect the led to the opto ic

So i decided to replace the 4n25 with a pc817 opto ic I had, and that low speed ticking sound just vanished! still has a faint, normal smps high frequency with 50Hz hum

what did that opto ic just had to do to make that sound?

anyway, i hope there wasn't any 100khz/100mhz noise as my contraption with a 1volt incandescent light and the 474j capacitor didn't glow while testing

so what about the classD amp? it just simply amplified the noise in the smps output powerlines as it had a builtin opamp for gain, as opamps aren't happy with noisy powerlines...

[T3sl4co1l]

Compensation problem, then?

Again, completely in the dark without an oscilloscope -- you're seriously handicapped without one, we can only guess.

Tim

[Herschel]

Compensation problem, then?

Again, completely in the dark without an oscilloscope -- you're seriously handicapped without one, we can only guess.

Tim

I know but anyway I decided to continue with what I have🥴, and later I would recheck the whole circuit with an oscilloscope after college reopens....