Rectifier tube 5AR4 sparking of Dynaco ST-70 amp

[kevinkr]

Everyone has covered it very well, so I can only rehash.. LOL If this is a vintage ST-70 and not a recent reproduction I would go through it very carefully and replace all of the coupling capacitors and check resistors to make sure they are in tolerance. (Replacing resistors is not a bad idea)

If there is a selenium rectifier rather than a silicon diode in the bias supply it should be replaced, a 1N4007 would be fine.

Total coincidence that this happened after the rectifier was replaced and diodes added. You should replace that rectifier once you have the amplifier sorted out. I have never seen a 5AR4 red plate so this one was grossly overloaded.

 

[Aaron M]

Before you install a new matched pair of EL34s, check the coupling caps for leakage. The original "Black Beauty" caps become very leaky with age. Even a simple Ohmmeter can find that.

The original black beauty caps have been replaced by the previous owner.

I made some bias adjustment the bias setting it at 1.5V, now the amp is working flawlessly.

Will go on study the relevant knowledge and do further mods.

Thanks

 

[Aaron M]

I started adding additional diodes to Stereo 70's a long time ago. Probably about the time I came on this site and called them junk. I first added them to my Quicksilver mono block amps
Another thing I do after I add the additional bias pots is to slightly lower the bias. You will hear absolutely no difference after doing this and the tubes will last a lot longer.

You are right, I set the bias at 1.5V.

Thanks

 

[Aaron M]

Everyone has covered it very well, so I can only rehash
If there is a selenium rectifier rather than a silicon diode in the bias supply it should be replaced, a 1N4007 would be fine.

The original selenium rectifier has been replaced to silicon by the previous owner.

Thanks for your advice.

 

[kevinkr]

Good that the selenium rectifier has been replaced.

A meaningful reduction in dissipation would be to reduce measured bias to somewhere around 1.2 - 1.3V

 

[Aaron M]

Good that the selenium rectifier has been replaced.

A meaningful reduction in dissipation would be to reduce measured bias to somewhere around 1.2 - 1.3V

The previous owner told me he set it around 1.35V, I am wandering if this impact the sound.

 

[rayma]

The previous owner told me he set it around 1.35V, I am wandering if this impact the sound.

It's actually more important that the two tubes in the pair have exactly the same DC bias current,
to minimize the DC in the output transformer. The bass will be better, the closer they are to being
the same. For this you need a sense resistor and bias pot for each output tube. This is a little OT,
but I seldom see this mentioned.

 

[kevinkr]

1.35V measured reflects an approximate 14% reduction in current and will significantly improve output tube life, in my experience changes of this magnitude are not generally audible.

Edit: To Rayma's point I assumed these tubes are well matched, if they are not the simplest thing is probably to replace them with a matched set. Alternately a separate cathode resistor of say 10 ohms for each tube is a good idea and would be part of the changes required if adding the ability to adjust the bias of each tube independently. Separate resistors would allow you to verify whether they are matched or not.

 

[Aaron M]

I cannot confirm EL34 tubes are well matched or not, as what the owner said and did, they were matched, he marked each tube as their position when I picked amp up. But two 7199 are different brands, Left one is GE, Right is Amperex.
Anyway I am planing to buy another whole sets of tubes, any suggestions for brands of these tube?
Now the bias is 1.36V and it's steady after playing around half hr, I didn't hear different of sound comparing to at 1.5V.

Thanks

 

[petertub]

It's actually more important that the two tubes in the pair have exactly the same DC bias current,
to minimize the DC in the output transformer. The bass will be better, the closer they are to being
the same. For this you need a sense resistor and bias pot for each output tube. This is a little OT,
but I seldom see this mentioned.

This is what matched tubes are here for! And it should be tubes that are
matched at approximatly the same voltage as they meet in the amp.

 

[petertub]

I cannot confirm EL34 tubes are well matched or not, as what the owner said and did, they were matched, he marked each tube as their position when I picked amp up. But two 7199 are different brands, Left one is GE, Right is Amperex.
Anyway I am planing to buy another whole sets of tubes, any suggestions for brands of these tube?
Now the bias is 1.36V and it's steady after playing around half hr, I didn't hear different of sound comparing to at 1.5V.

Thanks

You don't need new 7199, the fact that they are from different brands does not
matter. The day you need to replace them you could either :
- get a new board that uses other tubes or
- get adapters that allows you to use the 6U8 /ECF82 that is available at a much
lower cost then 7199

The EL34 (or simular) that you but should be a matched quad, this allows you
to mix them in any combination, and up to two of them may be lost during use
and you always have one matched pair. 5ar4 ; get an additional spare.

 

[6A3sUMMER]

Yes, 4 bias pots is a good idea.
I expected that anybody who installed 4 individual cathode resistors, and then measured the balance between a pair of EL34s might come to that conclusion (so often the two currents do not match).

And going from 50mA per EL34 to 45mA will not only extend the life of the output tubes, the very hot power transformer will run cooler too.

 

[Aaron M]

Hello Petertub, thanks for your advice, I am going to buy some tubes online.