Nasty Sounding Tech 21 Trademark 60

[Ted]

Tech 21 Trademark 60

It came to me with a melted power transformer, and the LM3876 power amp detached from its heatsink.

Replaced transformer with something about the right size and reattached power amp.

It sounds like a fuzzbox on both channels.

Tried an external speaker. Same.

It has a good +/- 30V on the power amp. and good +/-15V on the op amps.

Tried an external pre-amp into the FX return. Still sounds same.

Tried replacing LM 3876. No change.

Any thoughts?

[J M Fahey]

Serch and post the schematic so we can be more specific.

Start by measuring voltage on all LM3876 pins and post them here, all with no speaker attached.

[glebert]

Serch and post the schematic so we can be more specific.

Start by measuring voltage on all LM3876 pins and post them here, all with no speaker attached.

Tech21 is VERY protective of their schematics. They do not give them out. They will however talk you through stuff. I was working on a TM60 and ended up talking to guy who designed it for like an hour. I think his name was John.

[J M Fahey]

Cool

That said, LM3876 fit there can´t go too far from the circuit shown in the datasheet application note so the latter should be a good reference.

But we never heard about pinout voltages asked for, we´ll have to wait for that.

[Gingertube]

See the datasheet
http://www.farnell.com/datasheets/2001459.pdf

I would be looking at the gain set resistors R11 = 20K and Ri = 1k and that 22uF cap (check for a shorted cap too). Most likely very similar values on the amp.

Cheers,
Ian

[Ted]

Sorry for delay. Finally catching up with some of the stuff on the shelf:

Resistor values are 47K and 2K2 which are both about double the data sheet values so should give the right gain. Cap is 2.2uF which gives a rolloff below 31Hz which seems reasonable. All test good.

So. Voltages on LM3876 are as follows:

1: 31.3
2: 0
3: 0
4: -31.3
5: 31.3
6: 0
7: 0
8: -2.38
9: 0
10: 0
11: 0


These all make sense according to the datasheet.

But. Very visible crossover distortion on output of amp

Input on Pin 10 looks clean enough. (This is with no load)

[Ted]

Fixed it. You need to sink 0.5mA from the Mute input for the LM3876 to work. This is pulled down to -31V with a 30K resistor. Which should give about 0.9mA. Someone out their on the internet noted that the LM3886 (which is a very similar device) will generate crossover distortion if you don't allow a good margin over the 0.5mA. Changed the resistor for 22K. Distortion gone. This means its now pulling 1.2mA. Which is way more than the data sheet says it needs. But within limits. I don't like this. But maybe the original transformer put out a volt or two more which was enough to pull the mute down harder. Anyway, its more than the data sheet says it needs but still within limits, so I'm calling that fixed.

[g1]

Weird. You said another chip gave the same result? Are they using anything other than 8 ohm load? Did you use the formula in the datasheet to calculate Rm ?

[Ted]

I didn't look too closely at the original device - it sounded terrible so I just replaced it. LM3876s are hard to get hold of so I bought it from a random ebay dude. (There may be a clue here). The resistor on the board already was already sinking around twice the current the manufacturer suggested. (I pulled it and checked value out of circuit). There is an inductor in the (8 ohm) loudspeaker path. A better solution would probably to have used an LM3886, which is pin compatible and available from suppliers who don't operate out of sheds.