HIP4080 based amp issues with DC offset (is it ?)

Hello everyone !

Long story in short - it's a cheap class D hip4080 amp using B31N20D for output mosfets. Rail voltage is almost 70v at 14.4v b+ input

Works fine with no load connected. Sinewave is good, no distortion.
Once load is being hooked up it goes into frenzy mode :D . I've tried with different dummy loads values and a subwoofer.

What have noticed is that if you turn on the amp and let it idle with no load connected for 10-15 seconds and connect a load afterwards it works just fine.

I've measured DC offset between the - and + speakers output. And here it is. Once the amp is being turned on there is the FULL rail voltage difference between the + and -, that 70v volts difference. Then It slowly sinks down to a ~ 0.7v after 15 seconds.

So if you connect a load after the first 15 seconds when the amp is being turned on and the DC offset has been stabilized - it works just fine.

If you turn on the amp with a load connected on - it draws a lot of current and there is ....well you know what happens when You have +70volts difference between + and - speaker terminals (i'm glad i'm using dummy load this time, not a subwoofer in the tests).

My question is - is it the HIP4080 or a leaky Mosfets. I don't see a muting transistor at the board (don't have schematics due to this being really cheap chinnese amp).

Both HIP4080 and the mosfets IRFB31N20D are not easy and cheap to find over here, so is there any test i can do to determine where the exact fault is ?

The amp has been carefully re-caped, resoldered and cleaned up.
 

Attachments

  • upload.jpg
    upload.jpg
    363.9 KB · Views: 199
That's why I posted this thread. This is the first amp with hip4080 using these type of mosfets. I've personally always been using irf3710z, but even the markings on the board says 31N20s. And yes, these are factory mounted. No previous repairs.
Picture attached.

Do you get the same offset (lower level) across terminals 6 and 7 of the 4080 when the amp initially powers up? - Yes I do. Just divided by 10x
 

Attachments

  • up2.jpg
    up2.jpg
    93.8 KB · Views: 199
Does the input of the op-amp used to drive pin 7 of the 4080 have a delayed rise to 6v?

- if we are talking about the NE5532 pin 2 and pin 3 both have 6 volts immediately after the amp has been turned on.

Is the relay engagement delayed as expected or does it engage immediately?

- works like it supposed to, the usual delay is present.

I've noticed a factory missing resistor hip4080 pin7 to ground (i think )
 

Attachments

  • up1.jpg
    up1.jpg
    135.8 KB · Views: 190
Does the amp drive DC to the load (high current draw) before the relay engages?

- No. Only after relay clicks

Is the large resistor near R59 within tolerance?

- Both reads ~ 2.2 ohms as they should be.

See if you can find the equivalent resistor (capacitor?) on the attached diagram.
-Could not find any, maybe this board is being used for other amps with additional features. But R18C in the diagram is 10k, mine is 1k. That's 10 times difference....
 
If you short the op-amp equivalent of IC4C terminal 1 to terminal 2, do you still get the rail voltage across the output? - Shorting it causes 18volts constant DC at the output, there is no more rail voltage at the speaker output terminals and no more sinking. Just constant 18volts DC offset.

After the DC drops, does the amp work perfectly driving a load up to full power?
- For now my power supply can deliver no more than 20A at 15volts. So far no problems, but that's like 300W of input power which is not much.
Maybe if I push the amp hard enough there will be issues. Next month i'm planning an upgrade for the power supply so I can test amps at home at full load.
 
NE5532 with pin 1 and 2 shorted, RCA unplugged.

pin 1-6.5v straight DC
pin 2-6.5v straight DC
pin 3-6.5v straight DC
pin 4-0
pin 5-7.1v noisy dc (really small small waves)
pin 6-7.1v noisy dc (really small small waves)
pin 7-6.9v (something at 65khz picture attached)
pin 8-12v

If it will help, i can make measurements with 40hz sinewave through the RCA's.
 

Attachments

  • 85044550_1129384833916949_7074056439274143744_n.jpg
    85044550_1129384833916949_7074056439274143744_n.jpg
    399.1 KB · Views: 70