Tubelab SE bias current keep on rising

[drazmanithnin]

I've completed my 300b Tubelab SE. My B+ = 445.Bias voltage for 5842 at 175v and negative bias voltage at -150v per 300b. I used 36k 3w resistors for R14 and R25. Initially the bias current was set at 65ma for the 300bs but then the current slowly keep on rising exceeding >100ma. I keep on tweaking the R12 R23 pots until I reached the pots end.
Help please!

[rknize]

Wow...you are running a lot of B+. I wonder if you are getting runaway.

[drazmanithnin]

I've changed the power transformer. The B+ is now 350V. I've done all the modifications as suggested from Tubelab webpages. The problem still persist.

[rknize]

Can you put a meter on the source of one of the FETs (the leg of R14 or R34 that goes to Q1) to see if the output tube bias drifts? I wonder if the FETs are overheating. If the grid bias is steady but the plate current keeps drifting then it sounds as if something is wrong with the tubes? They should be happy at 350V.

[SGregory]

What does you B+ do when the current runs up? At that kind of bias draw you should see a significant reduction in B+. If your grid bias is really at -150V and your current runs away then the tubes are way out of spec. Hard to imagine both tubes failing simutaneously the same way.

rknize is correct. You should look toward the fets. If you have enough meters and clips set up watching B+, grid voltage, and bias current and report back.

I had a similar problem as a result of my own error. I was having a hard time controlling bias, plus it was really distorted. It turns out that I was reading my meter wrong for bias current. I was assuming that 0.065V was 65mA of current. Worked all day and into the night on the problem. Woke up the next morning and had an "AH HA" moment that the resistor was 10R and therefore 0.065V is in reality only 6.5mA. After that I felt stupid. But I soon got over my stupidity when I set the bias to .65V (65mA) and enjoyed the music.

[rknize]

I've been tricked by my ultra-cheapo digital meters the same way. The display is so bare-bones (no decimal point) that you have to remember what the multiplier of both the meter and the resistor is before interpreting it.

[adwsail]

I'm having the same problem with my SE. I got a set of Sophia Mesh Plate 300B and cannot get the bias to stabilize ON ONE TUBE. It makes no difference which socket I put it in the one tube will NOT stabilize. The other tube will stay pretty constant, + or minus 2 ma, but the other tube has gone over 75ma on several occasions and over 100 onn one occasion. I just dont trust the amp at this point so I'm checking the bias every 15 minutes. Not a good way to enjoy an otherwise stellar amp. I'm about ready to contact the good folks at Sophia Electric and see about a new tube.

[rknize]

I would do that. They are made in China, IIRC, just like the 300BS Shoogoo mesh tubes I tried. They were stable, but the construction was not very consistent. I would expect that at Sophia prices they would replace the bum tube.

One more piece of data you can get is to put a meter on the tube grid. If the grid voltage stays steady when the bias current drifts, then you have more ammo that the tube is questionable.

[adwsail]

Just contacted Sophia and they want BOTH tubes back for replacement. Can't ask for much more than that.

[rknize]

Awesome. I'm sure they are quite used to it given their source.