In over my head with tubes!!

[jaysee]

I own a Randall rg50tc. Basically its one of Randall's few tube amps, it's a 50 watt combo. Stock it has 2 el34's. The amp is well suited for high gain styles. Even the clean channel breaks up very readily, and with clean boost takes on even more distortion. In the back of the amp are test points for checking bias, there are two trim pot's, one for each tube

As my music preferences have come to include lighter rock rather than only metal music I've been somewhat limited by the rg's excess gain. I found this post about a mod for the rg to clean it up a bit:

http://acapella.harmony-central.com/sho ... G50TC-mods

As the poster stressed how simple the mod was I figured I would try it. I tried running a 12at7 in the other preamp positions and didn't like the harsher high frequency breakup so I left the preamp alone. I decided to try the power amp modification mentioned in the post. Just to try it out I borrowed 2 Tung-sol 6l6gc-str tubes from my peavey ultra plus and jumped the 10k resistor at R147 (circled in the schematic attached). I liked the tone and was able to bias the tubes at 35mv as suggested.

After several weeks I wanted to get the peavey up and running so I bought a matched set of svetlana 6l6gc tubes. I got a hard/blue set from musiciansfriend, hoping to further help clean up the amp's tone. I dropped the tubes in and found the lowest bias I could achieve was around 43mv with the trim pot cranked all the way. (this is just under 70% bias at idle) I continued to play as I thought the bias might drift down as the tubes broke in. After playing more I noticed with the lights off I could see a faint red plate on both tubes.

I suspect the screen could be red plating, as I read somewhere that the screen voltage is measured in the tube bias. I'm new to tubes, how would I measure the screen voltage or bias or.. And is this assumption justified? Thanks! I really appreciate any help

[ampgeek]

Welcome Jaysee!

Yes. The current measured at the cathode includes the contribution by the screen. Typically ~10% of the total.

You can determine that by measuring the voltage drop across the screen resistors (R113 and R106) and using ohms law. Current (in amps) will be the voltage drop divided by the resistance. 1000 ohms in this case.

Screen power dissipation will be the voltage drop that you measure times the current that you calculate.

The tube data sheet gives the max dissipation for the screen grids.

Good luck!
Dave O.

[xtian]

The photo you posted shows tubes operating normally, not red plating.

[jaysee]

The photo you posted shows tubes operating normally, not red plating.

The camera on my phone didn't really do it justice, but in the middle of the plate it glows red after running for several minutes.


I also didn't mention earlier that after first testing the svetlana 6l6gc's and finding the bias too high I put the tungsol str's back it and they biased correctly. Is there some fundamental difference between the two tubes? This finding leads me to assume I didn't blow a screen resistor.

[jaysee]

You can determine that by measuring the voltage drop across the screen resistors (R113 and R106) and using ohms law. Current (in amps) will be the voltage drop divided by the resistance. 1000 ohms in this case.

Screen power dissipation will be the voltage drop that you measure times the current that you calculate.

The tube data sheet gives the max dissipation for the screen grids.

I measured the voltage across R106 (R113 was harder to get to) and found 1.96v drop across the resistor. I'm getting a little hung up on the math though, it is simple but I think I might be reading your post incorrectly

V=I*R but I have voltage, and want current so I=V/R
V=1.96 and R=1,000 so I=0.00196 amps

then you say W=V*I
V=1.96 and I=0.00196 so W=0.0038 watts

The max screen dissipation for the EL34eh is 7.8 and for sv6l6GC 5watts so I was expecting to find a value between these, but my found value is excessively low. Is this typical/expected for an amp at idle? Is there error in my reading/calculation or does this mean my screen dissipation is in check and something else is causing my bias issue?

Thanks for the posts so far

http://www.dougstubes.com/el34eh.pdf

http://www.ne.jp/asahi/evo/amp/device/6l6gc.pdf

[martin manning]

You've got the screen current right, but screen dissipation is the product of that current and the voltage from cathode to screen.

If you measure cathode current by measuring the voltage across the 1 ohm resistors at the cathodes, subtract the calculated screen current, and multiply that result by plate-cathode voltage (just plate voltage in your case), you are left with just the plate dissipation. Since screen current is about 2mA, and your cathode current is 43mA, if you calculate plate dissipation without subtracting screen current you get an answer that is almost 5% high- a conservative result.

Different examples of the same tube type will require different bias voltages to achieve the same idle current. It is a little odd that you are seeing red-plating at that relatively low current, which appears to be well below the maximum. I think that the bias supply in that amp is borderline for biasing 6L6's, though.

[ampgeek]

Yep...my bad!

I knew that but my fingers were firing faster than my neurons and I didn't think it through.

Please forgive me for misleading you.

Cheers,
Dave O.

[jaysee]

Don't sweat it Ampgeek!

Last night I unjumped R147 and put the original El34's in to undo the whole modification.. I found that one tube has what looks like cathode oxidation rattling around within the tube and the other one has become microphonic!

So because I'd rather not but another new set of tubes I'd like to get the 6l6's to work.

This is from the original post of this mod:
"Normally it sweeps from -40V max on up/less negative. I removed R147 10K resistor and replaced it with a jumper. so the bias starts at -55V. A good starting point. I put in the 6l6GC's and biased them at -50V with a corresponding plate current of 30mV which gives you a 13-14 watt dissipation at idle, a comfortable level for both the tubes and the output transformer."
http://acapella.harmony-central.com/sho ... G50TC-mods

Looking at the schematic, jumping R147 is not giving me enough negative sweep. Would it be feasible to jump (or run another resistor in parallel to) R111 and R114 to achieve a sweep further negative?
My issue is not knowing where to start in order to calculate what resistance would put me in a safe spot. Right now my bias starts at around 43mA, I would like it to start around 30mA.

Once again, I appreciate the help

[jaysee]

Also, I want to be sure what I'm seeing is red plating. With a normal functioning tube is there any red coloration of the plate? What I see is around the creases in the plate closest to the center of the tube, a faint red/orange glow. It is more easily seen with the lights off. It is not coming from the filament, I know that much.

Would a tube run at or slightly about 70% dissipation show any heating of the plate?

Thanks for bearing with me

[Structo]

Your screens could be glowing.

[Tubetronix]

Looking at the schematic, jumping R147 is not giving me enough negative sweep. Would it be feasible to jump (or run another resistor in parallel to) R111 and R114 to achieve a sweep further negative?
My issue is not knowing where to start in order to calculate what resistance would put me in a safe spot. Right now my bias starts at around 43mA, I would like it to start at 30mA.

R111 and R114 carry essentially no current, and are necessary to create the load for the phase inverter output, so no for two reasons. The problem is the transformer winding voltage is too low, or said another way there just isn't any more voltage available.

[Reeltarded]

I am a bit surprised that this amp can't easily bias in such a narrow range as to handle 34s and 6l6.

[martin manning]

They are being clever using the same winding for the bias and the bipolar power supplies. It works for EL34's but is a bit lacking for 6L6's.