help designing 6L6G S.E. amplifier

[pilovis]

what about this?

 

[dirkwright]

what about this?

Sure! you need two tubes though. You can direct couple the first stage to the follower, and then use a CCD for the follower load. That way, you delete a coupling capacitor (usually a good thing).

 

[dirkwright]

Here's my take, but it needs a small B- supply....

 

[pilovis]

Here's my take, but it needs a small B- supply....

Nice!

 

[TheGimp]

Will that work? The current mirror will force equal current through both anodes. Therefore, any feedback to tube section B will be isolated from effecting tube section A.

I think you want independent current sources for each anode.

 

[dirkwright]

Will that work? The current mirror will force equal current through both anodes. Therefore, any feedback to tube section B will be isolated from effecting tube section A.

I think you want independent current sources for each anode.

It feeds back through the joined cathodes.

 

[dirkwright]

Nice!

You can make a simple -8 volt DC or so negative supply from the 6.3 volt AC heater supply. You'd need a bridge rectifier, a capacitor and just ground the + side of the rectifier. It's only about 2mA for this anyway. simple.

 

[dirkwright]

My idea for the input stage comes from something like the attached file. I may have made a mistake in what I shared but that's the basic idea. The document that it came from is here:
http://ocw.mit.edu/courses/electric...009/lecture-notes/MIT6_012F09_lec20_loads.pdf

I have not read it all yet.

Q3 and Q4 are the triode.

 

[dirkwright]

Yeah, maybe I have the connections a little bit off on the input stage. See here:
The SOHA II Headphone Amplifier

Then again, maybe not. The first stage is supposed to be inverting.

 

[Eli Duttman]

Some important points. The 12AX7 triode lacks the ability to drive g1 of triode wired 6L6s. CCS loading 'FQ7 sections will push the gain up to nearly the full μ of 20 and bypass caps. on the cathode resistors can be omitted. The 1045S previously mentioned is fine for the CCS here.

BTW, the place for a CCS in a differential gain block (LTP) is in the tail. Do not CCS load the anodes, especially if a tail CCS is used. Choke loading LTP anodes is fine.

 

[retrovox]

This data sheet may be helpful http://www.retrovox.com.au/STC807.pdf The 807 and 6L6 are from the same family of tubes, so the data will be valid.

I have uploaded STC 6L6G application report as well. 23 pages.

See http://retrovox.com.au/STC6L6G.pdf

David

 

[SemperFi]

Why drive a 6L6 into A2 when the plate load is 5kohms, and b+ this low?

 

[dirkwright]

BTW, the place for a CCS in a differential gain block (LTP) is in the tail. Do not CCS load the anodes, especially if a tail CCS is used. Choke loading LTP anodes is fine.

I'm curious, can you explain why CCS is a bad idea for the loads on the LTP? It's apparently done in the tube world and it's common in the solid state one.

I agree the best place for the CCS is in the tail of the LTP.

 

[SemperFi]

I can see potential problems if the CCS drift and you get serious shift in your quiescent voltages. The CCS in the tail can fight the CCSs in the plates in a way. But if you shunt the CCS with a high value resistor this shouldn't be a problem, or make sure eventual drift is common and tracks.

 

[SemperFi]

Btw Maurizio, you are aware using b+ of 240volts will produce a very low output power? Usually these tubes are used with > 400V. It will work, but you'll have loudness troubles.

 

[dirkwright]

I can see potential problems if the CCS drift and you get serious shift in your quiescent voltages. The CCS in the tail can fight the CCSs in the plates in a way. But if you shunt the CCS with a high value resistor this shouldn't be a problem, or make sure eventual drift is common and tracks.

I don't know what others have done about this possibility, or even if it's a serious problem. I wish I knew more about circuit design.