BrianGT 3875 kit troubleshooting

[ejn327]

A little background before I get into some questions.
I have a 2 channel lm3875 kit with 2 amps run in parallel for each channel (2 kits) and each one has a 18V 250 VA transformer.

When I turned it on I got some 60 hz buzz coming out of the speaker. I plan on putting a shield around the transformers and rectifiers so that should help cut down on the interference.

That is not the main problem though. The chips get real hot fairly quick, even with no load or input hooked up. I have them hooked up to some pretty heavy duty heat sinks (6 fins, 2"fit height, 3" wide and 12" long, 2 chips per sink).

The diodes in the rectifiers also get hot - could these be hooked up wrong? (I thought if they were backwards they would pop)

I let it cool down for a few minutes and was going to probe around to check some voltages from the rectifier and when I turned it on again the fuse blew. (it didn't blow turning it on 5 or so times and then it just went) I have both transformers wired into a 6 amp fuse.

I am using avel lindberg transformers and have the wires hooked up as follows
on the input (120 V) - blue and violet combines to L terminal
brown and grey combines to N terminal
and output (18V) - red - AC1 H, black AC1 N
yellow AC2 H, orange AC2 N

I have checked over the connections between the rectifiers and the amp boards and they are correct. Everything is grounded.
Does the Chassis Ground get tied back to anything other than the 3rd prong of the IEC plug?

[BrianGT]

Originally posted by ejn327
A little background before I get into some questions.
I have a 2 channel lm3875 kit with 2 amps run in parallel for each channel (2 kits) and each one has a 18V 250 VA transformer.

When I turned it on I got some 60 hz buzz coming out of the speaker. I plan on putting a shield around the transformers and rectifiers so that should help cut down on the interference.

That is not the main problem though. The chips get real hot fairly quick, even with no load or input hooked up. I have them hooked up to some pretty heavy duty heat sinks (6 fins, 2"fit height, 3" wide and 12" long, 2 chips per sink).

The diodes in the rectifiers also get hot - could these be hooked up wrong? (I thought if they were backwards they would pop)

I let it cool down for a few minutes and was going to probe around to check some voltages from the rectifier and when I turned it on again the fuse blew. (it didn't blow turning it on 5 or so times and then it just went) I have both transformers wired into a 6 amp fuse.

I am using avel lindberg transformers and have the wires hooked up as follows
on the input (120 V) - blue and violet combines to L terminal
brown and grey combines to N terminal
and output (18V) - red - AC1 H, black AC1 N
yellow AC2 H, orange AC2 N

I have checked over the connections between the rectifiers and the amp boards and they are correct. Everything is grounded.
Does the Chassis Ground get tied back to anything other than the 3rd prong of the IEC plug?

I would recommend disconnecting the power supply from the amplifier boards, and seeing if your power supply voltages are correct. If they are correct, I would test 1 channel at a time. If the channels are not working when running 1 at a time, then your LM3875 chips are probably bad, and I can mail you some new ones.

You certainly shouldn't be blowing a 6A fuse for the amp, so something is definately wrong. Try testing the power supply and each channel individually, and take it from there.

--
Brian

[ejn327]

I pulled out the power supplies and unhooked everything.
I am getting +27 and - 27 volts, so the power supplies are working.
When I hook up 1 amp it works and does not get hot. Its getting late but tomorrow I want to take the 1st windings from each transformer for one channel and the 2nd windings to the other channel. (red/black from each trafo to one channel and orange/yellow from each trafo to the other channel)

Could this be a cause to the heat, some sort of phase issue?

[BrianGT]

Originally posted by ejn327
I pulled out the power supplies and unhooked everything.
I am getting +27 and - 27 volts, so the power supplies are working.
When I hook up 1 amp it works and does not get hot. Its getting late but tomorrow I want to take the 1st windings from each transformer for one channel and the 2nd windings to the other channel. (red/black from each trafo to one channel and orange/yellow from each trafo to the other channel)

Could this be a cause to the heat, some sort of phase issue?

When you are running the amps, are you using any soft of resistor on the output? The datasheet suggests a 0.1ohm resistor in series with the output, before connecting the parallel channels together.

As for your question, can you clarify this? I don't understand.

--
Brian

[jleaman]

This actually sounds like the feed back resistor like i did.. i got 60v on the output asap when connecting power nothing blew up ut needed to change that resister to it's correct spot.. got a picture of your setup ie the chip and stuff ?

J'

[ejn327]

I had a 0.1 ohm 5W resistor at the output of each amp.

The feedback resistors are all fine.

After the rectifier the power is DC, but doesn't it still have a slight phase? Could taking different phases to 2 chips and then combining them cause problems?

The way I had it hooked up when it got real hot was one transformer for 2 amps in parallel. Tonight I want to try using the same windings from each transformer for 2 amps in parallel

[toque_101]

Is it possible you mounted the smothing caps in reverse polarity? I did that back when I made my first GC and the diodes got nice and hot... almost blew a capacitor doing that too

[ejn327]

the caps are all mounted correctly.
The problem with the heat and static noise seems to take place when I combine the amps in parallel (static noise only comes on one channel when in parallel, the other channel has no static).

With one amp running by itself it doesn't get very warm, but when it is combined it will run cool for a bit then heat up very fast. It also seems to heat up quicker when I have the source disconnected or paused.

Maybe I am not paralleling them correctly, but at the speaker level output I have a 0.1 ohm resistor on each amp and these are then combined togather (resistor in series with outout but outputs paralleled togather)

[BrianGT]

Originally posted by ejn327
the caps are all mounted correctly.
The problem with the heat and static noise seems to take place when I combine the amps in parallel (static noise only comes on one channel when in parallel, the other channel has no static).

With one amp running by itself it doesn't get very warm, but when it is combined it will run cool for a bit then heat up very fast. It also seems to heat up quicker when I have the source disconnected or paused.

Maybe I am not paralleling them correctly, but at the speaker level output I have a 0.1 ohm resistor on each amp and these are then combined togather (resistor in series with outout but outputs paralleled togather)

If both channels work perfectly fine seperately, then you might have a gain mismatch between the channels. If you want, I can mail you out some of the new resistors (the ones from thecurrent kits), and you can try matching them more closely.

--
Brian

[Peter Daniel]

Are you using separate rectifiers board for each parallel amp or they share the rectifiers? Also, check the DC offset of ea. amp when you run them separately.

Depending on the batch, I measured as high as 120mV of offset with 22k input impedance. In one case I even had 170mV.