Always, always, always replace the bias selenium rectifier with a silicone diode!!

I am selling a Dynaco ST-70 so I recapped the printed circuit board, set the bias and listened to it for a few hours. As usual, the amp sounded great and the bias held to what I set it at.

I walk away for a while with the amp on, came back to it and decided to check the bias... hmmmm over 4 volts on both sides.. Yikes! Grabbed another meter, same thing, I took a look at my variable isolated AC supply and the ST-70 was pulling over 3 amps... nominal 1.6 amps. Immediately I turned off the amp. I must have caught it just as the selenium failed as the tubes did not have a chance to turn cherry red. Had I not checked it, for sure something would have failed.

This morning I checked, the selenium rectifier was open. Replaced that with a 1N4004 diode and also replaced the 2 50uf caps with 100uf... bias nice and normal again.

Moral of the story, take that selenium rectifier out, it will fail and most likely take the power transformer out with it. It is also good idea to have a AC amp meter connected as you can take a quick look to see if the amp is pulling more current than it should.



Regards,
Sal

My ST-70 is stock and seems to be functioning normally. It draws around 1.70 amps.

Can you tell me the amp draw that is normal?

Also, how do you confirm/check the function of the selenium rectifier? I apologize it this info is posted or in the manual.

Thanks!

Depending where you set your bias and input line voltage, I would say 1.7 amps would be the maximum the ST-70 should pull at idle.

To check the selenium rectifier, unplug the amplifier, wait a few minutes so all the voltages will bleed off then put your meter on R X 1 scale and measure the selenium rectifier, you should get meter movement one way and no movement when you reverse your leads. To get the ideal readings, the wire should get disconnected at one side as you may be reading the resistance of the bias circuit.

If you can bias your tubes, then I would say your rectifier is still good, but a time bomb waiting to happen.

Thank you. Seems once amp is warmed up and running the amp draw is 1.60-1.61. (This is per a plug-in Kill A Watt P3 energy monitor.)

Bias voltage, however, has dropped since powering up about an hour ago. Should this reading be taken after warming up for an hour or so? While playing music, or not?

Readings have been: 1.589, 1.546 , 1.533, 1.510-1.520, and now have leveled off at 1.497-1.503. Should I reset bias at this point of warmed-up-ness?

Also wondering about the temps generated by the amp. Top of power transformer 105degF. Seeing surface temps top of output tubes around 200 degF. Seems hot, but I'm new to tube audio operation....

Your welcome Bob,
In the dozens of ST-70's I restored, I never had to parallel the resistor as I was able to set the bias with plenty of room either side on the bias potentiometer with the tubes that were installed in the amp. If the potentiometer was at one extreme, I would parallel the resistor, so far I have been lucky.. :-)

The temps you are seeing is normal, the original power transformer will get very hot as you can see... I mean feel. As far as the bias goes, the 1.56 volt will set the tubes to run at 50ma each at the plates if the tubes on the left or right side are matched. I would leave the bias set where it is, or set it even lower to 1.4-1.45 volts, the output tubes should last longer and there will be a less of a strain on the power transformer and you won't hear the difference in sound.

Regards,
Sal

PS: The ST-70 I have on my bench bias drops also after warm up. I wouldn't worry about it... I would worry if it rises.

Bob Latino wrote: only minor thing that could happen is that in some amps after a change from the stock selenium rectifier to a diode the bias range will change somewhat. You won't get your "1.56 volts DC" bias voltage near the center of the bias control anymore. The standard thing to do if this does happen is to alter the values of the two 10K resistors on the 7 lug terminal strip. The usual cure is to add another 10K resistors in parallel with the 10K resistor between lugs 1 and 2 (to give a 5K effective resistance) and a larger 20K or so resistor in parallel with the 10K resistor that is now between lugs 3 and 4. A 20K resitor here will give you an effective 6.66K resistance between lugs 3 and 4. This should now give you the correct bias somewhere near the center of the bias control. The reason for the necessity of the resistor alterations is due to the greater forward flow of current by a diode as opposed to a selenium rectifier.

Bob

OK, I replaced the selenium stack with a 1N4007 diode. I had read "somewhere" that I needed to INCREASE the resistor between the diode and the B- lead to 18k, but leave the resistor from the B+ lead to Gnd @ 10k. That doesn't work as I cannot get lower then 2.5V (with the 1.56 R). So if I understand Bob right, the first R (between diode and B-) should be ~5k, then the 10k pot, then a 6-7k R to ground. Is that correct?
This drops the total R significantly which would result in a higher I through the whole string, but we are really only concerned with the voltage divider aspect and this change makes the pot a larger percent of the total R giving a wider range of adjustment, correct? I will try these changes (a bit harder since a made PC board so I have to replace R rather than just add in parellel.)

So my bag of 1W resistors is not that extensive. I did find a 6.6k, so I replaced the 18k (the one between the diode and B-) with 6.6k and left the 10k from B+ to gnd. WORKED!! It was a bit "fiddly" but I was able to get 1.56 +/- .005 on both sets. I suspect that smaller R would make it less "fiddly" but for now, I am happy. I let it burn in for ~1/2 hr. with no significant change in bias. Other voltages checked out OK too, so now it is cooling off. Then I'll button it up, give it a brief try with "throw away" speakers and hopefully be listening to beautiful tunes by tomorrow. My ~50 year old amp (I bought it used in 1968) is about to be reborn!

Hmm... If I remember correctly, the guy I bought my amp from said he paralleled diodes across the selenium rectifier. Is it OK if it is not actually removed?

frank wrote:Hmm... If I remember correctly, the guy I bought my amp from said he paralleled diodes across the selenium rectifier. Is it OK if it is not actually removed?

I have never had a problem doing that and I've done hundreds in stock Dynacos. Selenium rectifiers fail high-resistance or open-circuit. Besides, it makes a good terminal strip!

GP49 wrote:

frank wrote:Hmm... If I remember correctly, the guy I bought my amp from said he paralleled diodes across the selenium rectifier. Is it OK if it is not actually removed?

I have never had a problem doing that and I've done hundreds in stock Dynacos.  Selenium rectifiers fail high-resistance or open-circuit.  Besides, it makes a good terminal strip!

Thanks!

you've got your choice of spending 10 cents for a diode, or $100 for a new set of tubes . . . .