st70 start up problems (smoking)

[kheper]

I have rebuilt my st70 with the diytube
driver board.

The speaker outputs are loaded with (2)
8ohm 10W wirewound resistors
connecting between "8" and "C" on
both channels.

When I fired the st70 up with the rectifier,
I heard a whirring noise and the wirewounds
began to smoke.

Could someone help me diagnose this?

[TomMcNally]

I would say you have the connections from the driver board to the EL-34's backward. They need to be connected in an X fashion ... not straight across. The amp is oscillating. Look at the schematic carefully, the area where it says "J1 EL-34 Left Rear" etc and make sure it is connected. It's crossed from the original board.

... tom

[kheper]

I think the smoke was due to an
open-ground problem. There
is no more smoke, yet the voltage
readings are rather high and the
tubes get very hot, but not red hot.


Measured the multi-cap:

A = 320
B = 375
C = 453
D = 464


Measured 5ar4:

2 = 463
4 = 379
6 = 366
8 = 463

[TomMcNally]

So I assume you were missing the ground on the low side of the output transformer, causing the feedback to act funny ? If your bias is set right, the voltages being a little high aren't going to matter, they are not that far out. How does it sound ? EL-34's do run pretty warm.

[kheper]

The connection to the star ground
was dodgy.

I am going to hook it up to the 'cheap
speakers' and hit it with a signal. Hold
on to your hats....

[kheper]

The cheap speakers were too cheap
to hear anything important.

It is hooked up now to decent speakers,
and it is sounding very clean.

Running the amp with an open-ground
fault is the same as running it with
no speakers. Should I worry about
anything here down the line? The
transformers did not get hot, only
the wirewounds.

Current music source:

xmms playing a ripped CD track: Tales
from Topographic Oceans --> USB
Sound Blaster --> Adcom 555 II
pre-amp --> ST70

[TomMcNally]

If the resistors got hot, they were definitely loading the output transformers. I don't think you damaged anything, it was just oscillating and putting out a lot of power and overheating the load resistors.

Glad it's working well - congrats ! I love the Adcom preamps, I have a couple of the 555 II's and a couple of 600's ... the features and remote control keep me from worrying about a tube preamp.

... tom

[kheper]

I have one concern: The bias voltages on
J6 and J21 go no lower than -41.7V.

How common is this low of a reading?

[TomMcNally]

I'm not sure I follow you ... If your voltage at the test points is correct (across the 10 ohm resistors) that's all that matters.

[kheper]

The voltage as advertised for coming off
J6 and J21 is -55V.

The right tubes can not get to 500mv.
They max out at 489mv, and that is
with opening the turn pot on the
board fully.

Shannon's board is very well tweaked.
The bass is precise and slamming,
not at all muddy.

[TomMcNally]

Have you replaced the selenium rectifier in the bias supply with a silicon diode, and replaced the capacitors ? You should check and make sure nothing has changed values if you're using original parts. You can mess with the values to get more voltage.

[kheper]

Yes. The only parts remaining in the
amp were the transformers and the
pots on the chassis. This was as
close to a total rebuild as is
possible.

[TomMcNally]

I think if you change the 10K to ground coming off the bias pots to 5K (simplest way it to parallel another 10K across it) you'll increase the voltage.

[kheper]

Thanks. That is a fine suggestion. When
I get some time, I'll look into that.

[Shannon Parks]

Good catch on the grounding issue. This happens to me occasionally as I get excited near the end of wiring it up and I forget the last bits. I think your amp will be fine - these tube amps can take a beating. Do check your LM234Z voltages at V2 & V3 pins 3 or 8. If they look in range, all is good. The solid state CCS is the most second most fragile part of the amp after the 5AR4.

If you wanted to be fancy, you could change the first 10k resistor (between the two cap negatives) to a ~15k. This would do the same thing as Tom's suggestion, but would also improve the ripple on your bias supply. But since the push-pull effectively negates any ripple on the bias, it really doesn't matter. Yellow_Light_Colorz_PDT_05

Now your amp is smokin', and not smoking. :o Spin some discs and enjoy.