I have 8 6AR6. These are Tung Sol tubes similar to 5881. I decided to modify my ST70 to use them. I rewired the tube sockets and triode strapped them with 100 ohm resistors between anode and G2. I added a 350 ohm resistor at the bottom of the cathodes to form a combined cathode and fixed bias. I then adjusted the bias with 400 volt across the tubes at 45 ma idling current for each tube.
6 of the 8 tubes worked fine. However, two of them just cannot be adjusted. There was no current before the heater warmed up. The moment the heater warms up, bias current increased immediately. The 350 ohm wire wound resistor started to smoke if I did not shut the amp down right away.
Are those two tubes shorted internally?
Could it be possible that there were some oscillations that caused those tubes to run away? Can I increase the triode strap resistor to 1 K ohm and solve this problem? I thought 6AR6 are rugged tubes. I have never seen this with any other tubes I have. The tubes are used but with no sign of age (no smoke on the top of the tube).
6 of the 8 tubes worked fine. However, two of them just cannot be adjusted. There was no current before the heater warmed up. The moment the heater warms up, bias current increased immediately. The 350 ohm wire wound resistor started to smoke if I did not shut the amp down right away.
Are those two tubes shorted internally?
Could it be possible that there were some oscillations that caused those tubes to run away? Can I increase the triode strap resistor to 1 K ohm and solve this problem? I thought 6AR6 are rugged tubes. I have never seen this with any other tubes I have. The tubes are used but with no sign of age (no smoke on the top of the tube).
I've never known a tube to short. Tubes usually don't conduct on initial start up, but the more it heats the more it conducts till it reaches operating temp. They usually Open due to a burnt out filament, loss of vacuum or overheated.
You didn't state what value are you using for a Plate/Anode resistor?
Or are you implying your running it as cathode followers with a 350 ohm cathode resistor and 400 volts at the plate
You didn't state what value are you using for a Plate/Anode resistor?
Or are you implying your running it as cathode followers with a 350 ohm cathode resistor and 400 volts at the plate
There are several possibilities. The tubes could be gassy. They could have a grid-to-cathode short. They could be oscillating. And (I've actually seen this) the internal grid connection may have come loose.
Does the problem follow the tube, i.e., do you have the problem with the same tubes if you move them to different sockets? Are you using grid-stopper resistors?
An OT question: are you the same Bing Yang who does TPEs? If so, small world!
Does the problem follow the tube, i.e., do you have the problem with the same tubes if you move them to different sockets? Are you using grid-stopper resistors?
An OT question: are you the same Bing Yang who does TPEs? If so, small world!
Jeb-D.,
These are power tubes. The plates were connected directly to OPT. I have hooked up a dummy 8 ohm resistor as output load.
SY,
Yes, the problem is tube dependent. I had grid-stopper resistors at 1K ohm as used in the original Dynaco circuit. Should I change the 100 ohm triode strapping resistor to 1K ohm? Will it help? All resistors are metal film. Could carbon composite help?
These are power tubes. The plates were connected directly to OPT. I have hooked up a dummy 8 ohm resistor as output load.
SY,
Yes, the problem is tube dependent. I had grid-stopper resistors at 1K ohm as used in the original Dynaco circuit. Should I change the 100 ohm triode strapping resistor to 1K ohm? Will it help? All resistors are metal film. Could carbon composite help?
Are the grid stoppers soldered in as close to the tube socket pin as possible? 1k should be plenty, but the position is critical. Screen stoppers are usually not that critical for triode-strapped pentodes or beam power, so it's more likely that you've got a few defective tubes. I've even seen tubes with dead shorts between grid and cathode, so anything's possible, but gas strikes me as the most likely bad actor.
That can be checked with a little work setting up a test jig. What sort of test equipment do you have? Scope? Bench power supply?
That can be checked with a little work setting up a test jig. What sort of test equipment do you have? Scope? Bench power supply?
I moved all resistors closer to the pins. I also replaced the grid stopper to 1K carbon composite. One of the bad tube now works. Just as I was happy at the results, one of the 100 ohm triode strapping resistor flashed and burned.
This is rather odd as the resistors are 2 watt metal film ones. Both the tube and the resistor were working for at leat a few hours before. All I did is to resolder and move the resistor closer to the pin. There had to be quite an amount of current going throug the resistor to burn it up! Should I change the triode strapping resistors to 1K to reduce current flowing through it?
This is rather odd as the resistors are 2 watt metal film ones. Both the tube and the resistor were working for at leat a few hours before. All I did is to resolder and move the resistor closer to the pin. There had to be quite an amount of current going throug the resistor to burn it up! Should I change the triode strapping resistors to 1K to reduce current flowing through it?
That's a common symptom of oscillation. If this tube is like a 6L6/5881/7027, the screen current will run from 5-25mA per tube. That's sorta high for 1k, but it's worth a try to see if it settles things down (current limiting effect is minimal- it will still pass nearly half an amp before the resistor vaporizes).
I suspect that the compensation really needs to be readjusted because of the lowered rolloff from the output stage Miller effect, but you'll need a scope and square wave source for that.
Note that the plate ratings of the 6AR6 are lower than the original EL34 (19 vs 25W), so be sure you're not simply just running them over the edge and you've only got a few tubes complaining.
I suspect that the compensation really needs to be readjusted because of the lowered rolloff from the output stage Miller effect, but you'll need a scope and square wave source for that.
Note that the plate ratings of the 6AR6 are lower than the original EL34 (19 vs 25W), so be sure you're not simply just running them over the edge and you've only got a few tubes complaining.
I replaced the grid stopper resistors with 3.3K ohm, kept the triode strapping resistor at 100 ohm for one channel and 1K ohm for the other one, moved the 10 ohm current sending resistor as close to the pin as possible. When triode strapped, the average current going through the resistor is about 3.3 ma. So the screen voltage for the 1K ohm side is about 3 volt lower than that at the plate.
One of the tube still does not like the set up. It now slowly conducts more current instead goes up in a second. I decided not to bother with it. I managed to get two pairs matched within 1 ma. The cathode bias is at 32 volt (along with fixed bias at -26 volt). The dissipation is right at 18 watts (400V x 45 ma).
The gain of 6AR6 is lower than EL34. One channel started to hum after the change. Thurned out that one of the 7199 is weak. I changed it with a new one. So far so good. I am not sure the sound will be better than my existing Sovtek EL34G. My note indicated that the Chinese 6L6G actually sounds a little bit better than EL34G. Hopefully 6AR6 will be better than my HK 6V6 PP and a DIY 2A3 SET.
Thanks SY for the help.
One of the tube still does not like the set up. It now slowly conducts more current instead goes up in a second. I decided not to bother with it. I managed to get two pairs matched within 1 ma. The cathode bias is at 32 volt (along with fixed bias at -26 volt). The dissipation is right at 18 watts (400V x 45 ma).
The gain of 6AR6 is lower than EL34. One channel started to hum after the change. Thurned out that one of the 7199 is weak. I changed it with a new one. So far so good. I am not sure the sound will be better than my existing Sovtek EL34G. My note indicated that the Chinese 6L6G actually sounds a little bit better than EL34G. Hopefully 6AR6 will be better than my HK 6V6 PP and a DIY 2A3 SET.
Thanks SY for the help.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.