Troubleshooting a Dynaco ST-35

[Captouch]

Originally Posted by ArtG

Remove all the output transformer connections on both the primary and secondary side, Measure all the resistances on both sides of each transformer. On the secondary side (to the speakers) measure from the ground connection to the 8 ohm tap and the 16 ohm tap. Also measure each of the resistances to the transformer frame. Repeat this on the primary side. Measure from the B+ (center) feed wire to each plate wire and to each screen grid wire. Also measure from the B+ feed to the transformer frame. Although the secondary resistance will be low, in no case should any of these readings be "0", which would indicate a short. This will either find or eliminate the output transformers as the issue. If they are original Dynaco transformers, the resistances on the primary side from the B+ feed from one side (plate or screen grid lead of one tube) to the other will not be identical. That is normal, and is due to the way that they are wound.

Do you have access to a variac? It's is possible to run the amplifier with the output tubes removed, to determine which stage is drawing excess current, however, the B+ voltage will go high, depending on your AC line voltage.

I'll try testing the output transformers per your guidance. Thanks.

No, no variac here, though I may be able to go to a friend's and use his.

[petertub]

Or simply disconnect the center ( red) wire from each output transformer, if no smoke reconnect one transformer at a time. All done with tubes removed.
If one of the transformers causes smoke you have found the culprit.


Luckily new transformers are available from dynakitparts.com

[Captouch]

Originally Posted by petertub

Or simply disconnect the center ( red) wire from each output transformer, if no smoke reconnect one transformer at a time. All done with tubes removed.
If one of the transformers causes smoke you have found the culprit.


Luckily new transformers are available from dynakitparts.com

OK, that makes sense.

But is there no definitive way to comparatively measure both transformers connected (comparing measurements vs each other) and tell which one is bad that way?

It's just that it's very tight to solder all those connections to the EFB board through the old can cap chassis hole, and I anticipate having to remove some GND connections to safely get to the output transformer connections, remove those, reapply the ground connections, look for smoke. If no smoke, remove the GNDs again, reconnect one O/P red wire, reconnect the GNDs, check for smoke. If no smoke, remove GNDs, reconnect other O/P red wire, reconnect GNDs, look for smoke again. Or I just attempt to dive in there with the soldering iron and hope I don't burn any wires in the process. Dexterous, unfortunately I'm not.

It's a pain to the point that I'm contemplating just cutting both O/P red wires in the middle to avoid all that other desoldering/soldering. Then splice back together and heat shrink as needed.

[petertub]

Originally Posted by Captouch

OK, that makes sense.

But is there no definitive way to comparatively measure both transformers connected (comparing measurements vs each other) and tell which one is bad that way?

It's just that it's very tight to solder all those connections to the EFB board through the old can cap chassis hole, and I anticipate having to remove some GND connections to safely get to the output transformer connections, remove those, reapply the ground connections, look for smoke. If no smoke, remove the GNDs again, reconnect one O/P red wire, reconnect the GNDs, check for smoke. If no smoke, remove GNDs, reconnect other O/P red wire, reconnect GNDs, look for smoke again. Or I just attempt to dive in there with the soldering iron and hope I don't burn any wires in the process. Dexterous, unfortunately I'm not.

It's a pain to the point that I'm contemplating just cutting both O/P red wires in the middle to avoid all that other desoldering/soldering. Then splice back together and heat shrink as needed.

Sounds like a plan. Cut one and test, 50% chance that you identified a bad transformer ( if the problem is here)

If tubes are out there is only the can cap and transformers that can be a trace to ground, that is if all cabling issues are cleared.

[Captouch]

OK, good news, seems to be fixed.

Before I decided to cut one of the O/P transformer red wires, I was poking around measuring to see what else might be shorted to ground and I found that one of the output transformer leads to the output tubes seemed shorted to ground.

So thought I had a bad output transformer.

I cut the lead that was shorted and then measured the transformer side of the cut wire to GND, but it wasn't shorted. Then I measured the tube side of the lead, and no short, so I was confused. Then I tried one of the PC-13 pads that was connected to the same tube pin that the believed shorted transformer lead was connected to. It was shorted.

Anyway, to cut to the chase, I found that it was one of the screw/locking washer/nuts that held the PC-13 board to the chassis didn't have an insulating washer and the locking washer was cutting into the adjacent pad, shorting that pad to GND. I was gifted this ST-35, so I don't know the history of when these PC-13 boards were installed and by who.

When I removed that screw/washer/nut, no more short. I checked the other PC-13 board, equivalent position, and that had an insulating washer.

So mystery solved. I just left that screw out for now since I don't have the right washers here, patched together the O/P transformer lead I had cut, double checked everything or shorts again (none), and then powered up. No smoke.

Then put the tubes in, powered up, and biased. All seems good.

The only thing I saw that wasn't sure about was that the bias was fluctuating rather rhythmically between 0.26-0.27v. Then I measured the various B+ voltages at the various B+ caps, and those voltages were also fluctuating a bit within a 10v range.

Is this typical with the diode/SS rectification scheme that's used for EFB? Or should it be rock solid at a voltage?

[ArtG]

I'm glad that you found the issue! Yes, those PC-13 boards are quite close to the chassis.

The fluctuation in B+ voltage may be due to fluctuating AC line voltage. 1 volt difference in the AC line will cause a 3 volt B+ fluctuation, assuming that you have 120VAC, not 230 V. I once lived in a location where the AC would go from 119 to 125 and back, over a few minutes! Of course, the power supply voltage will affect the bias current that you are measuring.

Also, you should be making your measurements with the inputs shorted, that is the center pin of the RCA connector attached temporarily to chassis ground (both channels).

[Captouch]

Originally Posted by ArtG

I'm glad that you found the issue! Yes, those PC-13 boards are quite close to the chassis.

The fluctuation in B+ voltage may be due to fluctuating AC line voltage. 1 volt difference in the AC line will cause a 3 volt B+ fluctuation, assuming that you have 120VAC, not 230 V. I once lived in a location where the AC would go from 119 to 125 and back, over a few minutes! Of course, the power supply voltage will affect the bias current that you are measuring.

Also, you should be making your measurements with the inputs shorted, that is the center pin of the RCA connector attached temporarily to chassis ground (both channels).

Thanks for your (and everyone's) help and suggestions.

Yeah, I'm in U.S., so 120VAC. I wasn't monitoring the outlet voltage - maybe next time I'll do that.

And I didn't know that the measurements should be taken with inputs shorted. I do have shorting plugs from one of my receiver's phono inputs, so will use those next time.

Always learning a lot from the great people here.

Thanks again!

[petertub]

Nice to hear that it was solved. And this fault could have been really expensive if you not have been methodical.

Trouble shooting is an art where symptoms should be tested against theory made according to observations and tested with possible actions/measurements.


Good luck and enjoy the music !

[Captouch]

Originally Posted by petertub

Nice to hear that it was solved. And this fault could have been really expensive if you not have been methodical.

Trouble shooting is an art where symptoms should be tested against theory made according to observations and tested with possible actions/measurements.


Good luck and enjoy the music !

Thank you Peter. Your suggestions were extremely helpful - I appreciate your taking the time!

[petertub]

Another observation is that the problem was caused by something outside the schematics. There was no faulty component, there was no broken wire or board strip.

If the amp had been totally disassembled, the components one by one examined, and then
assembled the fault would have mysteriously disappeared with no trace.