St-70 dead channel

Hello all,

I'm a long time st-70 owner and first time poster, and I'm looking for some help in diagnosing a problem(s) with my amp.

Aprox. 6 years ago, I replaced a bad driver board with one of those sold by Triode Electronics. I also opted to run it in triode mode. All was well for a year or so, and then I put the amp in storage and didn't use it until last year. When I took it out of storage and put it back into use, I noticed that it didn't bias correctly -- I couldn't get it *lower* than about 1.95V when it was warmed up. Also, it would eat a rectifier tube every few months or so (which I would know had happened because it would start to blow fuses). Yesterday, I replaced a blown rectifier and after maybe an hour of use, the left channel of the amp went dead.

I've done basic troubleshooting (switching tubes, cables, etc) and that yielded nothing. I opened it up and tested voltages at 3 & 4 of the EL34s to check for a possible bad output transformer, and they were in the 530 - 540 range (which is high from what I am reading).

Okay, that's the extent of my knowledge. Undoubtedly, I should have explored the bias problem and the blowing rectifiers before something failed. Any help would be appreciated!

Check the 15.6 ohm resistor going from pins 1 and 8 of the EL34 tube to ground. It may be open. Also I second what Bob mentioned above. Your selenium rectifier is probably going bad and you don't have enough negative voltage to bias the tubes.

Sal

Thank you both! I'll upload a photo this afternoon when I am working on it.

I replaced the selenium rectifier when I did the driver board upgrade. I haven't switched the caps in the bias system though.

So could something with the wonky bias be causing the dead channel? Of course what I'm really worried about is that the output transformer is bad on that side.

The good thing is you have a good working channel so you can compare measurements between the left and right sides. Check the 15.6 ohm resistor, if that resistor opens, there will be no current flowing through the tubes and the channel will be dead.

Sal

I was sure hoping the resistor was bad, but I just checked and it's testing right at its rated resistance. Is there anything else which can fail besides the transformer which would totally shut the channel down?

Joe

Yeah, I switched the left set in for the right -- for both the output and driver tubes. Still dead, unfortunately.

OK, your next step to rule out the transformer is to set you meter to measure at least 500 volts DC and measure from chassis ground to pin 3 and from ground to pin 4 of the output tubes. Let us know what you measure. Be careful, there is lethal voltages there.

Did you swap the middle tube on the printed circuit board with the tube in the channel that works?

Sal

Huh, I just realized that the only tube I can't swap out is the middle 12AU7 -- the triode board had EF86s on either side and a 12AU7 in the middle. I just looked at the tube, and it's an RCA made in the US. This, I believe, means it's totally ancient. I don't have a spare 12AU7 on hand to swap in. Can I use one of the EF86s just to test?

Sal, voltages at 3/4 were 530-540.

Nope the EF86 tube is totally different than a 12AU7. If you have a 12AT7, 12AX7, 12BH7, 5963 or a 5751 tube and many other dual triodes with the same pinout, you can try them out in a pinch.

As far as your voltage measurements, looks like your audio transformer primary side is good. The screen and the plate of your output tubes is getting voltage from the transformer. Disconnect your speakers and measure from ground to the other speaker taps, compare them with the other audio transformer. I am not familiar with the Triode board to give you assistance on that.

Sal

Okay, I'm using the wonders of Amazon prime to get a 12AU7 here tomorrow to swap in and rule that out.

The Sony "receiver" I'm using as an amp while trying to get my ST-70 running again is exceptionally good motivation to fix my amp!

Joe

You should download the schematic for the board. It is basically the same as the original 7199 board just using more commonly available tubes. I have the 6AU6 version. The voltages on the driver tubes will be different for each board (6AU6, EF86, 7199) but they should be similar from side to side on each board. Take readings from each EF86 and both sides of the 12AU7 and compare them. Post the results.

Tony

My 12AU7 came in yesterday but unfortunately swapping it in did nothing -- I've now switched out all tubes. So I emailed Triode and asked for a voltage check list which they sent. Below are the voltages they specify followed by my measurements. There is clearly something not right, but I have no idea what. My measurements done with no tubes in the amp besides the rectifier (as directed by the Triode spec sheet)

Thanks for any ideas!

Joe

Triode's Spec:

EF86
1 Screen 55 - ­80V
2 Shield 0V
3 Cathode 1.2 –­ 1.4V
4 Filament N/A
5 Filament N/A
6 Plate (Anode) 80 –­ 110V
7 Shield 0V
8 Suppressor 0V
9 Grid 0V

12AU7
1 Plate (Anode) #1 235 – 265V
2 Grid #1 80 ­ 110V
3 Cathode #1 90 ­ 120V
4 Filament N/A
5 Filament N/A
6 Plate (Anode) #2 235 ­– 265V
7 Grid #2 80 ­ 110V
8 Cathode #2 90 ­– 120V
9 Filament CT N/A

My measurements

Right EF86
1. 159
2. 0
3. 1
4. 0
5. 0
6. 331
7. 0
8. 0
9. 0

Left EF86 (the dead channel)
1. 0
2. 0
3. 1
4. 0
5. 0
6. 335
7. 0
8. 0
9. 0

12AU7
1. 471
2. 336
3. 0
4. 0
5. 0
6. 473
7. 336
8. 0
9. 0

Thinking about this more and looking at schematics, I see two problems:

1. The voltages are way high. Is this a bias problem? Bad power transformer?
2. There is no voltage at EF86's grid on the bad channel. Is this enough to kill the whole channel?

I didn't have time to diagnose #2 further, but maybe the grid coupling capacitor went bad?

I found this procedure which I use for testing a Dynaco 70 output transformers.

http://www.curcioaudio.com/Stereo%2070%20Restore%20Addendumsl-R0.pdf

Look down the first page and find section B

So I solved the dead channel issue: it was an open screen bias resistor. Now to order parts to replace the bias circuit (which is still way high).

Fry's to the rescue for bias caps and resistors. Now biased to a much more reasonable 1.56v. Thanks to everyone for the help.