Intermittent Thump Issue with with my new ST35 board

[dcgillespie]

Hifi-

One added thought I had about your problem over the weekend. Is there any chance that you have the primary wires of the output transformer mis-paired? That is, for a given output tube, is the screen lead of one side paired with the plate lead of the other? If so, that will always cause goofy problems, and could let the output stage oscillate all by itself. Just a thought.

Dave

[HiFiGuy]

I thought some of the leads on the output primary might be off, I checked the manual, and looked at pictures of ST35 on the web. Everything checks out in that respect. I have hooked up another tube amp I built as a daily listener. The ST35 is a higher quality amp but for now its sitting on my bench. I would like to do a custom adjustment of the feedback, I need to look up how to do that. I have a scope and a generator, that should be all I need. Now all I need is the time to mess with it. I was planning on taking over to my buddy's house. I have described the problem to him and he thinks he knows what to do, but he still wants to check stuff out. We will put our heads together and if we come up with a solution I will post it here. This is a weird problem I have not come across with any other tube amp.

[TomMcNally]

I've built about 6 ST-35's and haven't run into that problem.

I'm going to suggest what I suggest to everyone with an amp
that isn't acting right. Check each and every resistor and make
sure it's the proper value. It's very easy to put a 10K where
a 100K is supposed to go, or 10 ohms instead of 10K, etc.

We've helped fix dozens of amps on this forum and that seems
to be the most common problem.

I'd measure them with an ohm-meter.

... tom

[HiFiGuy]

I checked all components, they are correct.

[Brik]

Hi,

The symptom suggests an intermittently failing power supply component, an intermittent solder connection, a hairline crack in the PCB trace,
or infrasonic oscillation due to the output-stage-to-V/A-stage feedback through B+.

You might want to examine the PCB traces for hairline cracks, especially those of the ground and the B+.
You might also want to re-solder all the joints, wires, lugs, etc.

Then, the list of usual suspects:

1. Replace the fuse.
2. If there is an inrush current limiter, replace it.
3. If there are any "snubbers" left, take them out.
4. Replace the diodes. Both of them.
5. Replace all the power filter caps, one by one.
6. Replace all the power filter resistors, one by one.
7. If there is a choke, replace it.
8. And, finally, replace the power transformer.

After all this, it may just turn out to be a bad power cord or a flaky switch...
Or you may have to hire an exorcist for your speakers...

Good luck!

[Shannon Parks]

Here's my stabs:

1) The diodes may have been damaged with the snubber circuits. Replace them both. I can send you a pair free of charge if you PM me your address.

2) We can adjust your feedback to something like 50k on each pot since we are not sure of the tuning with the Anteks. This is actually a pretty decent ballpark SWAG. But since the breathing occurred with the feedback disconnected, I would say this is not the primary problem.

Shannon

[HiFiGuy]

I took the amp to my buddies house and he took a look at the amp. He noticed that my input jacks were mounted directly to the chassis. He suggested I make the input ground float. He also told me to ground the chassis to the AC ground. I have the amp in this set up now. As per the above suggestion I replaced the diodes. I also adjusted the neg. FB pot to 17K so I had more negative Feedback. If this does not work I will try the adjustment higher. I have it hooked up and I am listening to it now. I am going to put a couple of hours on it and see if the speakers are breathing. I will update shortly. Thanks for all your suggestions.

[HiFiGuy]

The speakers are not breathing now, the correct FB setting is between 17K and 20.3K, its just a matter of fine tuning the amp.

[HiFiGuy]

I popped in some reissue Genalex tubes and I noticed that a couple of the grid 2 windings were glowing like the cathode. I shut down the amp and put in some Russian military tubes. Do I need resistors between the board and the screen windings on the transformer? Thanks.

[Shannon Parks]

Did you need to re-bias?

Shannon

[HiFiGuy]

I have all the bias pots set at 390 ohms. I was going to try several quad sets that I have and see which one sounds best before calibrating the bias pots. I was looking at the original schematic, and Dynaco used a 95 ohm resistor shared by all the 6BQ5 cathodes. Were the originals known to have glowing G2? When the bias pots are set the grid will stop glowing? Thanks

[Ty_Bower]

I noticed that a couple of the grid 2 windings were glowing like the cathode. Do I need resistors between the board and the screen windings on the transformer?

I've seen the same thing myself with some Russian 6p14p tubes. I've also seen it happen in EL34 tubes too. I've even seen some small hints of it in a few beam power tubes (6L6, 6p3s-e). Usually, the beam tubes are much better behaved than the pentodes. I've tried (and failed) to cure the issue by adding a couple hundred ohms of resistance at the screens. I've seen it happen in ultralinear designs, and also in triode-strapped designs. I haven't dabbled much with true pentode amps.

It's annoying, and I'd like to know exactly what needs to be done to make it go away. The only cure I've found is to try a different set of tubes.

[dcgillespie]

There is really nothing that can be done to cure a screen with glowing hot spots, as in any properly designed circuit, it is a fault of the tube rather than the circuit. Beam power tubes are far less likely to have this fault since they are designed with aligned grids. Therefore, the screen is significantly shielded from the direct electron flow in a beam power tube. Such is not the case though in a true pentode tube, so any manufacturing deviations relating to the screen grid structure can cause excessive dissipation at any point on that grid in a pentode. As Ty said, when it occurs, it is most common with EL34s or EL84s, but can also happen in poorly manufactured 6L6 type tubes as well -- although it is much less frequent.

Dave

[HiFiGuy]

I took the amp to my buddies house and he adjusted the feedback using his scope and generator. The amp sounds great and no breathing!! I connected a 0.1uF capacitor from signal ground to chassis ground and this got rid of a lot of the noise at idle. Thanks for your help guys.