power acoustik one channel runs hot

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
power acoustik saphire 1200 watt x 2 amp.....I am having a problem with my amp, i checked every mosfet transistor out and i didnt find any shorts, to put it short one side of my amp gets hot and i believe thats why there is a slight hum in that channel, what would cause a hum and those mosfet chips to get hot...tho other channel works like a charm and doesnt get hot jsut slightly warm...any help i would appreciate it..thanks
1200 "car watts", right? with an actual real MOSFET output stage (and not just a MOSFET power supply with bipolar output, and a huge "MOSFET AMP" badge on the outside?)

If one channel is running hotter than the other (and I also assume this would be evident while it is idling) but the amp still sounds OK, I would check the bias. You can check the idle current by measuring accross the emiter/source resistors (they should be from each power MOSFET's source to the output. large ~0.1x ohm resistors ). measure the voltage in he good channel and the bad one. it should be higher in the bad channel. If this is true, I would start looking at the VBE multiplier and then possibly the drivers. the VBE should be a small transistor that is hopefully in thermal contact with the heatsink. Should be connected to the driver transistors. check the voltage from the collector to the emitter on both. if the voltage is too high, the bias would be too high and the amp would draw too much current at idle and get hot. on some amps, there is a pot to adjust the bias. Bias typically goes sky high if these open up.

If you have access to a scope, I would check for oscilations, too. annother way to check for oscilations would be to check to see if the resistor in the output zobel is getting hot. it should be a resistor of at least a few watts power handling at a few ohms, it will go from the speaker output to ground with a small cap in series with it. In the amps I have seen, these were closer to the actual speaker connections than by the output transistors.
the usual problem with these would be a shorted output device (due to the devices being overworked from being turned up too loud, or cummy design, etc...) you can remove the bad MOSFETS and test them. there should be very high resistances. If you meaure a couple of ohms or less between any terminals it's shorted. You can also check the good MOSFETs from the othe channel to make sure of your readings. That's the nice thing about a stereo amplifier - you can compare with the other working channel!!! What might have happened was that one MOSFET (0r one side - let's say the upper bank) shorted, and then the feedback loop made the lower half conduct like crazy to keep the output at 0V. I ahve notices that some car audio amplifiers have odd or crappy overcurrent protection that might not shutdown the amp in all cases. However, if the failure persisted, the bottom mosfets would also be at risk and fail, too....

One way to check if there are no other problems besides output devices would be to remove the output devices and stick in some ~100 ohm resistors where the output devices were, from gate to source or from base to emiter (if bipolar). This will allow the driver transistors to "power" the output. you should be able to see a signal on the output, if there isn't a load connected. If you see output, and there isn't any significant DC on there, the preamp and driver stages should be working fine. If it gets this far, you shou check the bias. measure the voltage from gate to gate (or base to base) of the "output transistors" and see if it's less than a few volts at idle. It should be close to that of the good channel.

You want to check this and not just replace shorted transistors since there might be a fault elsewhere in the driver or bias stages, etc. that would fry the new parts!

How are you testing the amplifier. I wouldn't do this in the car! I have a small 12 volt marine battery in my garag....I mean lab! You could also use a spare computer power supply. be sure to leave the fuse in the amplifier. A meter would be helpful to see the amp's current draw if you can spare one. You might consider a smaller value fuse to protect the rest of the circuitry (or any new parts you bought/pulled/found/stole/borrowed!). You could also limit the current into the amplifier with something, perhaps a car headlight or something in series with the car amp's 12 volt input.

Since you got the amp for free, this is agreat learning opertunity, and a chance at a free/really cheap amplifier if you succeed. You won't be out anything if you screw up (provided you take basic safety precautions, remember to keep a *fuse* in the amp, and not a paperclip!). If you are unable to repair the failed section, you can either remove the failed MOSFETS and use it as a single channel amp, or cut out both output sections and use the power supply for something like a GainClone or something.

You might consider posting on the SolidState forum, since the output stage in a car amp is identical or very similar to home, pro audio, and musical instrument amplifiers. The only differences in the two kinds of amps are the power supply and the packaging.
i run it off of a car battery with a 30 amp charger connected to it i set the charger at 5 amps when i am not putting audio through whatever i am powering....if i bench test an amp i usually put it at the full 30 to keep adequate power..of course when i am looking for noise in an amp i connect no charger to listen for hum.. by the way i thought i was looking for a 0 ohm between the terminals for the shorted one...at least thats wha i have read..
zero ohms is definietely shorted, but I've seen ones that would read point something or a few ohms. These are the ones that don't act like a piece of wire between the dud terminals, but get really hot when power is applied. I had a MOSFET PWM speed controller in a cheap cordless drill do that to me. It's still blown either way...
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.