lost channel testing 1kHz sine wave

Infidel

New Member
So idiot here was wondering how his amp would read on a multimeter - found a 1 kHz sine wave generator online, had a speaker hooked to the left channel and a multimeter on the right channel.
4 ohm speaker.
Adcom GFA-5300.
distortion alert went on for a second, then thermal protect light.
now the left channel only makes a little distorted noise, right channel is normal. right side of amplifier chassis warm, left side cold.

What type of rebuild am I looking at?
Parts most likely to have failed?
 
So idiot here was wondering how his amp would read on a multimeter - found a 1 kHz sine wave generator online, had a speaker hooked to the left channel and a multimeter on the right channel.
4 ohm speaker.
Adcom GFA-5300.
distortion alert went on for a second, then thermal protect light.
now the left channel only makes a little distorted noise, right channel is normal. right side of amplifier chassis warm, left side cold.

What type of rebuild am I looking at?
Parts most likely to have failed?

injecting 1K signals is kosher....what size was it? I like 150mv...unless on phono then 1/50th of that peak!

What were you expecting to see on the DMM? what was it set to? It should have been set to DC volts or maybe AC volts....

was 4ohm speakers ok? some amps like onkyo R1s for example seriously get the knickers wadded up anything less than 8....
 
It was briefly at full volume from a macbook, with the website volume at 50%.
I had ear protection on.
The GFA-5300 has factory rating for 4 ohm load, but warns thermal protection may kick in - don't know if I actually fried something or if a nights rest will reset. Regardless, I do listen to some stuff where the distortion lights flicker - it is loud, but sounds great - but never had it go to thermal protect mode. The thermal protect light is no longer on, but it may not have worked in time.
What is likely to fry?
Something as simple as a recap of affordable parts?

speaker is ok...
 
Any fuses?

WTF were you trying to do, pardon my French?

BTW, multimeter must be set to AC if you're trying to measure the working output voltage.
 
Well I cannot see form here what signal levels you used.
Experience says if it goes 'dark' signal wise on a channel and it is not a fuse - its far more than a capacitor.

Are you capable of doing any testing, looking up the service manual and things like that - its ok if you cannot - but that means an internet diagnosis/repair is not going to succeed.
 
Full volume on your computer is still practically full volume. Lowering the volume in software is not the correct way to do it. You still get the same output signal, but "compressed". It also thus adds distortion, and lower sound quality. Audibly so, even.

You may well have blown output transistors due to overloading the amp.
 
I hope you weren't measuring current on the meter, that's a series connection, dead short. You sure the meter was on the (still) good channel?
 
The amp was still obviously fried on the left channel today so I opened it up for the first time.
The fuse at M001 is out.
The output wire coming off the left channel board to the + output is discolored - like near burned (yellowish) - I'm guessing it used to be white like the one to the right channel.
I had the meter to measure volts . Right channel is still fine.
At the very least the power supply caps are going to need to be replaced, they are bulged a bit and getting close enough to see the numbers my nose started to burn a little. The four of those will run $28 total unless I upgrade.
The rest of the caps do not appear to be damaged on physical inspection.
Not sure if the Mosfet's were Fairchild (not available) or Vishay. Might be fine, hope so.
I haven't rebuilt an amp yet but have a few to practice on.
I don't have the last two pages to the service manual, which are kind of important.
 
What is the appearance of the blown fuse? Is the element simply melted, or is some of the fuse tube blackened inside, maybe even with bits if metal stuck to the glass?
 
I'm curious as to what your original goal was. Measuring output voltage of an amp is going to show you input volts times gain, basically. Just more volts. Pretty difficult to get any other information out of a meter measuring volts.

Most folks will use a scope to compare input vs. output waveform.

Good luck on your fixes.

Chip
 
reading between the lines the amp appears well made ..
maybe compare good channel to bad with a dmm ..power removed of course ..
good place to start is the output devices .
 
What is the appearance of the blown fuse? Is the element simply melted, or is some of the fuse tube blackened inside, maybe even with bits if metal stuck to the glass?

Hard to tell, it's a tiny fuse. No black, no metal stuck to glass. Maybe 1mm missing from the filament.

For now I have removed the stinky leaking power supply caps. Nichicon 10,000uf 63v, 4 of them, 2 were a bit bulged and leaking. May be looking for upgrade.

The person that assembled the amp didn't heat shrink and solder the connections from the PSU to the left and right channel boards. That made it easier to flip the board over to remove the caps, no telling if it had any part to do with the failure.

I'm going around the board comparing resistors from one channel to the other, but don't have a layout diagram to work with. So far they are all within expected range of each other.

Haven't looked into how to test the output mosfet's
 
Given your description of the fuse I'd probably just replace it and give it a try. If the caps are leaking then they need to be replaced but, personally, I wouldn't tear the whole damn thing apart without first trying another fuse.

Get the leaking caps replaced, try the fuse. If it blows again then dig deeper. Don't make it into more trouble than there may be.
 
Last edited:
Pull the good fuse from the right channel and test it in the left channel. If it blows, something else went, if it holds, the fuse gave up its life for your testing fun.
 
good point - no need to go to the store to buy a fuse yet, and they are sold in 4 packs...

is there any way to bridge / parallel a pair of GFA-5300's? I'm guessing there is a way, but not straightforward as in needing to have some resistors to keep the amps from killing one another. I know I'm not going to need the volume "louder" - I just don't want to get to clipping and that seems to come from running out of headroom.
I'll be recapping one, so may be recapping two so they match. (ePray purchase perhaps for the second).
One person looked at the schematic and thinks a pair of the IRFP240's may be shorted but no explanation for the burnt speaker output wire - 16 gauge, 3 inches long - that's a lot of current.
Wonder if the Israeli ebay seller of mosfets is reliable? Not that I'm going there yet, just wondering about my possibilities before deciding on which brand of caps will be best. Relatively new to this but spent the last few months building a tube preamp and power supply.
But yes, for now I'll try the fuse after I get replacement power supply caps.
Wish I could change the title of thread.
 
Parallel isn't of much benefit unless you're wanting to increase power available to real low impedance loads.

Bridging will increase available power more effectively at more typical 4-8ohm loads.

If the amps do not have built in bridging capability, you'd need to find some way to provide a normal signal to one channel and an inverted signal to the other channel. There use to be adapters for doing that but they don't seem to be widely commercially available. There are plans around though for building bridging adapters. As well, the amps need to have or be capable of common ground operation for bridging.
 
Last edited:
Back
Top Bottom