Need big help with repair of class D sub amp. have schematic

fastline

Member
2010-10-21 5:49 pm
I need to repair a sub of mine that suddenly developed issues. First started with a popping/crackling sound with no output. I took amp apart and testing all major components with no findings so I reinstalled. Amp worked but only tested at low volume for a minute or two. So I added the rest of the bolts to the amp and was going to do further testing. Turned amp back on and since then it only makes a VERY loud hum in the 60-120hz range as soon as power is applied.

I removed all 6 of the filter caps and they test at 950uf (rated at 820uf). I charged them with my diode tester and they hold a small charge, etc. The just seem good to me... I diode checked the bridge and it tests good but not sure if I should apply power to it for a proper test?

The circled component is what I found the other night that was certainly failed either from a poor solder or heat. The OEM identifies this issue as the popping I would hear but should not cause the hum. I am wondering if the resistor problem could have hurt something else? I am certainly missing something here.

I did remove a couple mosfets for bench testing and they are fine so far. I usually look for dead shorts on mosfets but being told this noise is likely in the PS and not the transistor side. Not really sure.

Any experienced help would be appreciated...
 

Attachments

  • FSR amp mod 2.jpg
    FSR amp mod 2.jpg
    490.8 KB · Views: 261
you mark the R30A that paralled to R30B... It supplied input of REG U5 that use to regulate voltage to U8 (gate driver) at 15volts... It's seems that input is also switched... so it would be better if you have an osilloscope to look at the voltage rail... is it real flat or not... and also the switching wave... that's the quick and exact way to locate the hum...
I guess your supply rail not "clear" enough that cause hum...
If switching device (power or driver) have a problem then your fuse will blown out... so far there's no problem with that... just hum...
If you don't have an osilloscope... I suggest you to check or change all the voltage rail cap (on reg supply) one by one (in this schematic 33u cap)...
Hope it helps...
 

fastline

Member
2010-10-21 5:49 pm
In this hum, I have only switched it on for approx 1 sec at a time because I am concerned of further board or driver damage. I have not powered it on the bench as of yet.

Velodyne tells me if I do not have the J7 pins connected to the input board, the amp would hum very loud but I cannot see why....

What you are saying is to check for ripple on the DC regulated buss going to U8? I have certainly checked and checked on the main filter caps and I really think they are just fine but have not given much looking at the smaller caps.

On my board there is no R30B. The A and B were used on a newer revision, I only have 1 resistor there.

ALSO - I DID find a potential big issue yesterday. There is a diode that is jumpered to flow current to either pin 14 or 7 of the 74HC14 chip at U9. This power comes from one side of R31. I cannot find those pins at all on the schematic but looking at the chip specs, 14 and 7 are either the power supply for the chip or the ground. Both should be rather important!! This diode is bridged in as an "oops" in design I think and one side had come loose and was not conducting well. This probably would have stopped function of U9.

I really thought I had the smoking gun here some I soldered everything back in "again" only to find that it still hums.... grrrr.

Should I build a current limiting circuit to say 1A to test out this circuit? I wonder if I juice it up and try to run it, it will probably blow a fuse pretty quick. I am kind of wondering if the failure at R30 or the disconnection at U9 caused something else to go out.
 

fastline

Member
2010-10-21 5:49 pm
UPDATE:

I just removed every single smaller electrolytic cap on that board (about 8) and all test perfect. I have yet to find any component that is actually bad on the board and still think I have a contact or trace issue somewhere. Just not sure about anything on it just yet.