1 channel went out

[guitarpsych]

I was having trouble with the fuse blowing now and then. I would get a hum gradually building and then the fuse would blow. Lately I would turn off the amp when I'd hear the hum. I have ordered filter caps and will replace them.

So now one of the channels has gone out. I tested the preamp and both channels work. I tested without the rectifier, fuse doesn't blow. I changed the all the tubes from one side to the other and the same channel is still not working. I looked inside and all the connections seem solid. Any idea what this might be?

[TomMcNally]

It could be many things, but the first thing I would check is
the cathode resistors that go to ground. If they open up,
that channel will die. In the original ST-70 they are 11.6
ohms, kind of a white ceramic looking thing. Get rid of
those on all of the tubes and put in some 10 ohm one watt,
Radio Shack should have them.

... tom

[guitarpsych]

I checked the cathode resistors I believe you are talking about. There are two of them, white, sort of look a bit like old capacitors... connected to pin 8 of two of the EL34s and goes to ground. They measure about 15 ohms. Do you think I should replace them?

[TomMcNally]

If they are reading 15 ohms, they aren't open, so the amp should work ... so something else must be wrong - maybe the driver tube. If you're going to rebuild the amp, you'll want to replace a lot of things - those resistors included.

[EWBrown]

If you have the original Dynaco "crispy" phenolic driver board, it probably has multiple problems, first of all to place under suspicion, are those large coupling caps, if one or more goes bad and gets leaky, it will inject a positive voltage onto the grid of the EL34(s) and make things really go bad, real fast....
Also check the bias voltage, if one of the two bias pots is scratchy or intermittent, the bias voltage could be jumping all over the place and causing some nasty problems.

I really recommend replacing the board, with a new design driver board, like Shannon's "red board".

If you want to preserve the original look and sound characteristics, there are "exact copy" new fiberglass boards withe the original Dynaco ST70 7199 circuit, but seeing as how good NOS 7199s are hard to find, and are costly, and the Sovteks, though reasonably cheap, aren't all that good, their quality is quite "variable" at best. Been there, done that, I ended up swapping them out for something a lot better...

/ed B in NH

[guitarpsych]

I checked the bias last night. The working side is 1.9. The dead side is 0. It would kind of jump around when I'd move the probe in the socket but settle at zero. All the tubes light up though. Is it possible the rectifier tube is bad, or would both channels go out if that were the case?

Replacement board: I wouldn't mind replacing the board that has the 7199s with something that takes 12ax7s and/or 12au7s. Is there a replacement board that does that?

[TomMcNally]

You might try rocking the bias pot a little on the bad side (with the amp off) then turn it on and see what happens. If the amp works at all, the rectifier is good. Did you try swapping the 7199's - if not, as Ed suggests, you probably have a bad capacitor.

DIYTube has a very nice ST-70 driver board available, and an easy to use Mouser.com "BOM" (Bill of Materials) you can order all of the parts except the tube sockets from - you can get them from Triode Electronics or Angela. There is plenty of support on here. Look at the links on the front page of the forum for "DIYTube Products"

[EWBrown]

If it was the power supply, both channels would be equally affected.
Bias at 1.9V is high, but not fatal, probably reflects the "dead" channel not taking its share of DC power. Typically, if the plate current is low, or zero, the grid bias has to be at the low "extreme" around -40VDC or more, in other words, too negative. if the current is high, then the grid bias is too "positive". A G1 grid, if it is left floating, or with no bias applied, or resting with positive voltage, will really "hog" current and try to go into the "Chernobyl" mode :o

/ed B in NH

[guitarpsych]

Interesting.
What's a good way to trace that problem back to the source, to find the earliest part in the circuit where things are not working?

[mesherm]

Check the voltage on pins 3 (anode) and 4 (screen grid). There should be high voltage on all those pins. If not then check the center lead and solder joint of the output transformer for that side where its connected to the can cap. You may have to power down and measure the resistance of the transfomer by measuring between the center lead and pins 3 and 4.

[Shannon Parks]

Download & print out this manual from Dynakitparts.com if you haven't already:
http://www.dynakitparts.com/store/manuals/Dyna-ST70.pdf

Go to page 9. Measure and write down the voltages from 'Voltage Test Points', the capacitor lug, the selenium rectifier, and PCB #3 & 18. Taking all these voltages will take about five minutes. Report back with them and we should have plenty to work with.

[guitarpsych]

Ok, I think this is everything you asked for:

Voltage Test Points
(for the bad channel)

EL-34
pin# and V
(pins 1 and 8 should be higher than 0. What could be the issue?)
------------------
1: 0
2: 6.75
3: 464
4: 464
5: -33.3
6: -33.3
7: 6.75
8: 0


GZ-34
pin# and V
--------------
1:
2: 473
3:
4: 388
5: 387
6:
7: 472
8:


7199
pin# and V
---------------
1: 280
2: 111
3: 39.5
4: 3.3
5: 3.3
6: 1.2
7: 130
8: 111

capacitor lug
A: 310
B: 399
C: 461
D: 472

selenium rectifier
bottom: 57 V AC
top: 64 V DC


printed circuit eyelets #3and #18: 398

[TomMcNally]

A side by side listing of the good vs bad side EL-34 voltages would be interesting. Since you have plate voltage, the transformer isn't open. The lack of anything on the cathode is odd ... is the ground point the 15.6 ohm resistor is tied to good ? If it's the tube socket, it could be suspect.

[guitarpsych]

The sockets look good. I went through with a meter touching from from front to back and 0 resistance on each pin.

I double checked the risistor you mentioned and even replaced it to be sure, but no change.

The bias reads 0 wherever I have the bias adjustment.

I've tried swapping the left channel tubes with the right channel tubes andit doesn't help (one side works, the other does not, the problem doesn't move with the tubes).

Can you think of anything else it might be? This is driving me crazy. I can't figure out what it could be.

[guitarpsych]

After a troubleshooting session with EricHayes we finally tracked it down to bad tube socket(s). Thanks everyone for helping out. I've got it working now.