irs2092 in parallel on one +-65V source

Hi all
I'd like to use 4 irs2092 in parallel on one +-65V (1500W) source, like I saw it in several youtube videos
(all components from aliexpress)

While with only one irs2092 everything works fine, problems arise by adding further irs2092 in parallel.

Problem:
impulses at loudspeakers with 2Hz to about 10Hz randomly (rally noisy loud spikes)

so there is a kind of interaction of these amps , ..no idea what is the reason

I already tried:
  • changing the audio input signal (different Bluetooth receiver & dsp boards) -> no impact
  • different combination of irs2092 (to check if any of these is particularly damaged) -> any mix gives similar result (->no single defect of one amp)
  • R13 is 101 (not sure if this plays a role as described in other posts)

I only have basic electronic and soldering skills, and would be thankful for your support.
best regards
 

Attachments

  • 20240307_203423[1].jpg
    20240307_203423[1].jpg
    473.1 KB · Views: 104
  • 20240305_222145[1].jpg
    20240305_222145[1].jpg
    331.8 KB · Views: 94
thanks russc for the idea
I tried it (see photo)
amp 3 directly connected to source
amp4 with switch in-between

when I heard the music on speaker amp 3 i switched on the amp4 (connected to subwoofer 8ohm)

unfortunately same noisy (maybe at bit higher frequency than if i have connect 3 or 4 amps simultaneous)

in general it seems that the amp produces very high noise at low volume levels and have extreme switching impulse on the speaker during on/off
- the quality seems to be bad

any further idea ?
Did anyone use this amp for home cinema ? or is it scrap in terms of noise levels?
Hope I get this things running acceptable as I already bought 8 irs2092 and two of the 65V source.
 

Attachments

  • switch.jpg
    switch.jpg
    199.8 KB · Views: 46
Its years since I used the irs2092.
I seem to remember an application note about stereo systems where the 2029 clocks have to be out of phase or problems can occur.
The clocks can be tweaked to be a little different.

Dont connect the reset pins together on the 2092's.

I found with large smoothing caps the amp boomed on power down.
IR told me to add a reset control circuit that tripped when 12 volt supply went below about 10 volts
This isnt simple as the reset pin must be driven through an opto coupler and a resistor or 2092 blows up.
I used a small PIC with A2D to measure voltage and driver opto coupler accordingly.
 
new test showed that:
when I switch on/off one more amp sometimes the 65v source quits working (can be started after some while again)
both existing sources (same type) show same behavior

unfortunately I don't have a different +-65V trafo to test if all trouble comes from the source type