diytube.com

I have an odd situation that I've never experience before. While trying to set the bias on my ST-70 (which uses Shannon's driver board) I was able to read and set the bias on 3 of the 4 EL34s. However, on one of them I got no reading whatsoever (in other words, the reading on the meter did not change at all when I connected the leads to the amp). There did not seem to be any ill effects when the amp was in operation: no tubes glowing red, no problem with the sound, the other tube in that channel did not need to rebiased to 'compensate' for the other tube, the odd tube glowed and seemed to function normally). I expected that I must have a loose connection inside the amp to the bias check point. This turned out NOT to be the case. So, on a whim, I substituted another EL34 that I had lying around and, presto...I was able to read and set the bias for that tube as normal.

Is this a somewhat common occurrence? I'm going to guess that perhaps there is a missing connection inside the tube that causes this phenomenon but does not render the tube inoperable...?

Any thoughts? I'm just curious...

Thanks,
Joe

Did you check your cathode resistor connection? You may have an intermittent connection that disconnects the resistor that you are reading the tube current across. If this is the case, the connection was established again when you plugged in the new tube. When the cathode resistor is disconnected the tube is not harmed, but it is off completely and no current flows through it. At low volumes, it is possible that you won't hear anything too terrible, but at higher levels this ought to sound pretty bad.

Thanks...yes, I did check the cathode resistor connection.

I don't see any way the tube could work and have a missing connection of this sort. It is possible that the tube is not working at all - with a failed anode or cathode connection. A failed screen grid connection would also effectively shut down the tube. If no current is flowing through the tube, you won't see any reading at the bias set points. I haven't encountered a failure of this sort, but it seems possible. I don't recommend this experiment, but I expect that channel with a bad tube sounds pretty bad at high volume when half of the waveform starts to clip badly.

s this a somewhat common occurrence? I'm going to guess that perhaps there is a missing connection inside the tube that causes this phenomenon but does not render the tube inoperable...?

Any thoughts? I'm just curious...

Thanks,
Joe

An intermittent socket/pin connection wouldn't show as a bad solder connection. So placing a different tube in the socket may not reveal the problem again, if that tube's pin makes a better connection. If the tube is suspect, you should run it in a different socket and check for the biasing voltage a number of times. Some light pin corrosion, maybe? Try tightening the grip on the pin if possible and see if the problem remains.

20

I've had this happen. It's occurred only once on my ST70 with a well used quad of XF2 Mullards. The tubes tested great for emission, transconductance and plate current matching and biased up in the amp. After playing for half an hour I returned to the room to check bias. Zero reading on one tube in rear socket...The room was brightly lit with midday sun so I didn't notice that the rear tube on that channel had no filament glow. I tapped the tube in socket and it fired right up. The pins on these were thinner than the Russian tubes I'd been using and perhaps worn from being in and out of sockets for 50 years...so it wasn't making contact. Sound quality hadn't been affected but volume wasn't high either. That channel had been limping along on one tube for probably 20 minutes. I tightened the socket and all's well. Those tubes still work well and all of this is a testament to the quality of the old tubes.. I just don't see a modern JJ living through 50 years of intermittent service.

If the cathode resistor opened, the meter would actually indicate much too high of a "current flow". Actually of course, there would be no current flow at all, because the voltage is not being developed across any resistance in this case. In reality, the meter would actually be indicating the voltage that the tube would basically bias itself to, to achieve a cutoff condition if the cathode resistor were open.

If the the problem is in fact the tube, then either an open screen or open cathode connection internally within the tube would cause this scenario. Any other open connection would either cause the heater not to light, the screen grid to become a light bulb, or the plate to become a red light!

Dave

Thanks Dave, for clarifying the possibilities. In summary, is this what you are suggesting?

(1) Control grid open or bad connection- Anode overheats
In this case, what voltage does the control grid float to?

(2) Anode open or bad connection, screen has HV- screen overheats

(3) Cathode or screen open or bad connection - no bias reading, tube off

(4) Cathode resistor open - strange bias reading, tube off

HI Dan -- you've pretty well got it down. If the control grid became free -- it would rise to 0 volts from a negative value if traditional fixed bias is used, and to the value of the cathode voltage if cathode bias is used. Either way, the tube is clearly in distress at that point, and won't last long. Both the plate and screen will have dissipation levels well in excess of their ratings.

Dave