CCS current sink in the output of my 2 tube 6080 amp.

[Shoog]

Hi there all,
I know that this is an old topic but I want a bit of clarification before I make a final decision.
You may or may not have followed my thread about a 2 tube 6080 PP amp. I ended up with ECL82's driving a toroidal phase splitter and finally 6080's in a push pull configuration. I used seperate bypassed 330R resistor, with a parallel 2K in series with a 1K5 pot, to give me bias adjustment.That is each triode has it own bypassed cathode resistor. This works really well and sound good with excellent bass as is. I am sufficiently impressed that i think with a bit of tweaking it will sound much better than my set amp.

However I am concerned about bias drift over time, so I was thinking about various methods of automatic bias adjustment. I originally thought about using the "garter circuit" as mentioned in Tubecad, but I didn't have enough spare voltage and it required the insertion of another cap in the output grids.
After reading about a differential output stage in a "Vacuum State" power amp, I am now think about individual Constant Current sinks in each cathode. Again the inspiration has come from Tubecad. I was thinking of using simple bypassed LM317 CCS. Since the 6080 is biased at 30V the use of LM317's is a bit marginal so I was going to put a NPN transistor in series with it. The bias point of the transistor would be set with a resistor network at about 10V. This should let the transistor take the strain.

Is it a good idea to use seperate CCS's per triode or would I be better with just one per push pull pair ?

Are there any dangers to this approach?

Can I expect an improvement in detail and bass control ?

Thanks

Shoog

[alexmoose]

Hey, I am acutally struggling with the same problem right now, only I am worrying about the EL34, however you seem to know alot more about this than myself. the EL34 has a bias in upwards of 35V which is enough for an LM317HV, however I can't hanlde both EL34s on one LM317HV or the LM317.

My questions is some more basic than yours, but does anyone have any ideas as to how I can achive this? possibly with a stronger regulator?

I am so happy that you raised this question as it is directly related to my own, thank you!

[ErikdeBest]

Hi ALexmoose

As far as my common sense goes I would say that the LM317HV can handle two EL34's. With the cathodes tied together you will keep same bias voltage, just the current will double...and that is no problem for a heatsinked LM317!

And Shoog, I can't be of much help. I just know that Ian (gingertube) said he got the best bass from his baby huey by using 'bias blocks' for the EL84's. Also, DC balance in a toroid OPT seems to be a very important thing, so a way of providing constant bias would seem to be a good thing to do. Besides the traditional LM317 CCS I can only think on the 'bias blocks' used through Ian and the CCS boards from diyaudio.

erik

[Shoog]

Interesting article can be found at

http://www.glass-ware.com/tubecircui...o_Biasing.html

Seems that the raw LM317 can well handle the needed 30V of bias, so that makes the whole arrangement very simple. 4xLM317 and 4x 12R resistor, big heatsink (12 watts dissipation total).
The nice thing is that I built this amp in a very modular form. The bias arrangement is mounted on a seperate board and the cathode bypass caps are hanging of the valve bases. Therefore all I need to do is build a replacement board and switch them over. Couldn't be simpler.

Just need to get the LM317's now. I believe I have a pair stashed already.

Shoog

[ErikdeBest]

Hi Shoog

Are you still running the 6080's with 160VDC on the plate? I am just looking at Svetlana's datasheet for the 6AS7 and at 160V on the plate and a 100mA the bias would settle somewhere between -60V and -70V. What am I doing wrong in this observation?

Erik

[alexmoose]

In order to create enough headroom room would it help to put a Zener String between the LM317, and each cathode? I'm just guessing, Zeners kinda freak me out

[Shoog]

The actual situation is +B of 130V, and a cathode drop of 30V. This leaves 100V over the 6080. If you look at the datasheet I think you will find that with 100V over the tube and 100mA of plate current the bias will settle at about -30V.

Shoog

[ErikdeBest]

Hi Shoog

Thanks for the info! Now I understand the 30V bias - it's not 160V across the tube, rather 100V.

Erik

PS. Today I had my last exam. I still have to work some hours each week, but I really hope to put your FVP5 together - soon!

[alexmoose]

Originally posted by ErikdeBest
Hi ALexmoose

As far as my common sense goes I would say that the LM317HV can handle two EL34's. With the cathodes tied together you will keep same bias voltage, just the current will double...and that is no problem for a heatsinked LM317!

erik

okay, that re-assures me alot, but is there anything I can do to maybe get a little bit of headroom with the bias voltage?

[ErikdeBest]

Hi Alexmoose

I understand your point. You can try the tricks with zeners or a mosfet, cited by broskie http://www.glass-ware.com/tubecircui...Biasing_2.html

Or you build a bias block just like the one made by Gingertube for his baby huey. Just be sure to use transistors that handle the extra current - and heatsink them!

Erik