strange amp problem!!!

[moonchild]

hi there
i have a 2 channel car amp(110WrmsX2) that i've been using in bridged mode as a subwoofer amp for a long time now with no trouble at all but suddenly it started acting strange

it turns on fine and sounds perfect at low levels but when i raise the volume it cuts and goes into protection and then it tries to come back again and then cuts..... and so on unless i reduce the volume then it plays normally !!!

it overheats with the load connected even with no input! but if i disconnect the sub it doesn't heat at all .

i took it apart and removed the heat sink and power it up for 5 minutes with no load and no transistors(power or drivers) got even warm!!!! so i removed all power transistors and mosfets and even the power supply's diodes and measured them with my multimeter with no sign of failure for any of them,and the pcb looks clean with no sign of overheating or cracks,and there are no pots at all to set the bias .

i measured the power supply which seems to be ok at +/- 42 volts.

i noticed that one of the channels output is -8 volts dc with no input or load connected (that's a huge offset)!! the other channel is fine.
what could be the cause of this problem??

 

[AndrewT]

Hi,
-8Vdc offset is enormous. Sounds like something has broken, it's just the smoke hasn't escaped yet.

-8vdc through a 4ohm speaker will be pulling 2A through the negative transistor of the output stage as well. Both the amp and speaker will overheat. Disconnecting the load and running it open circuit stops the current flow and apparently cures the overheating.

 

[Tekko]

Its most likely one of the driver transistors or one of the predriver transistors that have started to give up.

 

[widowmaker]

Probably dry or leak electrolythic caps in feedback...

 

[moonchild]

thnx for all the replies.

i've changed the output power transistors with no change at all, is it possible for one of the driver transistors to do this while still giving a good reading on the multimeter , i thought it could only be good or bad but no (in-between) states!!

do i have to take all drivers and predrivers out to test??

i'm also confused because the amp works at low power level (20watts) and then it cuts when bushed harder , i've fixed many amps before with blown power stages but i never faced that situation

widomaker , how could i spot the dry or leaky elctrolytic ??

 

[Tekko]

Sometimes bad caps have grown thicker, or leaked onto the board, but this can be hard to see as they often have glue around them on the board.

 

[moonchild]

well i've looked alot at the pcb and i can't find any sign of oil leaking from a cap , and by the way the caps arn't glued to the board...

 

[moonchild]

c'mon guys i need ur help , i have another amp but this one means alot to me

 

[anatech]

Hi Moonchild,
Look for cracked solder joints near the heavier components on the board. A capacitor may leak a little directly underneath the cap. You may not see it. Look in the voltage amp section (you didn't have high current draw with the load disconnected).

-Chris

 

[Eva]

These failures are quite common and are usually due to cold solder joints that end up losing contact due to car vibrations. Don't bother too much with the output stage, the problem is probably in the input stage or the VAS.

 

[moonchild]

hi , happy new year everybody

after some testing i found that the bass and emitter of all power transistors have -8 volts on them while the collectors are on Vcc or Vdd according to polarity of the transistor, this is only on one channel the other one seems fine..

any help????

 

[AndrewT]

Hi,
I am not too good at fault finding, but I think that all the emitters at -8V indicates that all the output devices are carrying the same currents. However the bases should be either +0.7V or -0.7V away from their respective emitter voltages. Confusing?

There may be a NFB loop trying to correct this error. Go back to post 2? and check what he said.

 

[Tekko]

I had a similar problem on a guitar amp, one output transistor conducted enough current to put the speaker out at 27v, all transistors were functioning, the problem was broken pads on the driver transistors.

Possibly you have a similar problem, as its only 8v the drivers are probobly fine. Try bending the transistors back and forth and see if anything moves on the solder side of the board.

 

[anatech]

Hi moonchild,
What is important is how much voltage is across the emitter base junction. This will tell us if the current amplifier stage is working or not. Otherwise the fault is earlier in the circuit.

-Chris

 

[poobah]

Moonchild,

You have one good channel right? Start at the input and work toward the output or vice-versa. Work one stage at a time and compare the bad channel to the good one.

You should be able to pin down which section your problem is in...

 

[anatech]

When working in the output stage, reference your meter to the speaker output. This will give you more meaningful measurements.

-Chris

 

[moonchild]

thnx guys for the help , i'll do some more testing today and i'll keep u updated.

 

[moonchild]

i'm back with some more testing results

the supply voltage during this test was +/- 38.5 volts.
none of the output transistors even got warm during the test although they arn't mounted to the heat sink!!!

i took ur advice anatech and referenced my multimeter's negative to the speaker output terminal ( emitters ) during this test , and here r the results:

for the NPN transistor pair:

Base 0.4 volts
Collector 47 volts

for the PNP transistor pair:

Base 0.47 volts
Collector 30.7 volts

then i connected the multimeter's groung to the supply ground and repeated the test with the following results:

the NPN pair:

Base -7.6 volts
Emitter -8.1 volts

the PNP pair:

Base -8.6 volts
Emitter -8.1 volts

waiting for ur replys

 

[poobah]

Moonchild,

We will hit dead-end pretty quick... can you post or link to your schematic? Kinda hard to work blind. And also the make and model of your amp... somebody here might have a schem already -

Thanks,

 

[anatech]

Hi moonchild,
Right now the outputs aren't conducting, or just barely. Mount the outputs to a heatsink or you may lose them when something goes wrong.

Poobah is quite right. We need to make and model, ideally a scan of the schematic to give you some useful answers.

-Chris