6V6 Push Pull

[J_Hart_14]

So, this is my second time to post something. I have built this small 6V6 push pull amplifier out of some spare parts. I have included my schematic of the amplifier section. This includes some numbers and values. Just fyi I dont know much about the output transformer, I stole it off a Bogen Amplifier that had a pentode power tube and triode in the same glass tube El86 or something? At any rate I beleive it to be rated from 6-10K Ohms on the primary, it also has a 70volt line isolated secondary tap. That aside;

First: I would like to add a feedback loop which I have not done yet; i would like suggestions.

Second: Also I would like to see some math for the feedback if someone could do so?

Thirdly: I have indicated in red a grid stopper resistor, is there a reasonable value for what I have drawn? Also being as there should theoreticaly be no input current I coud pick this to be 100K as well?

Lastly: Any suggestions on the circuit would be much appriciated.

Any information or suggestions would be much appriciated. Thanks again,
Jonathan

[J_Hart_14]

Here is the schematic

[chrish]

Sorry I am not qualified to comment on the circuit, but I am interested to know what program you used to draw the schematic.

Thanks,

Chris

[J_Hart_14]

I just used Power Point, somewhat painstakingly but it worked. I do not have a good schematic program so I just found a way around it. Hope it helps.
Jonathan

[mach1]

Unecessarily complex. You can drive a 6V6pp with a single 12AT7 long tailed pair. Type in 'el cheapo' and do a search of the forum. Also do a search under 'musical machine' for a slightly different approach.

[andyjevans]

As above. Except I'd use the ECC40 which is a terrific sounding tube, very popular in European equipment. Three stages of 12AU7 is out to lunch and then some.

Actually I'd use a 1J6G as driver. Just discovered this little jewel of a double DHT and can't get enough of it - have it in my linestage. You'd need a preamp since mu is about 15. And a DC filament supply since it's directly heated, but that's simple enough and the tube is cheap.

andy

[jerluwoo]

Your circuit will not work. Here is a scheme to give you a starting point that should work correctly.

[J_Hart_14]

Originally posted by jerluwoo
Your circuit will not work.

I dissagree, I have it built currently and it operates. It could be debated how well it works but that aside it works.

Thanks for the input that it is overly complicated, now that I step back it does look a little excessive all for just a little output. A simpler solution could be had. I will take a look at a simpler circuit.

Again to repeat however, if I wanted to put a global feedback network is there some mathmatical solution to go about this? Local feedback I somewhat understand; however, once I need to take the tansformer into account I get a bit lost. Do I just make a voltage divider based on what I expect at one of the secondary taps? Then just assume that voltage divider to feed back an overall % compared to the input?

[Poindexter]

I mainly agree with those above; the main problem with your circuit is that it has way too much front end gain, so the noise level will be a little higher than optimum. You could remove that third front end stage and drive directly off the split load inverter, it would improve the sound, I bet.

If your outputs are operating in triode, as shown, I think it would be a mistake to apply global feedback around the circuit.  You would probably be disappointed in the reduction of space, image, and inner resolution.  Feedback was used in the pentode amps of the Bogen period because the output impedance of the amp would be unacceptably high without it, thus the inclusion of extra gain stages to supply Av that would be subsequently lost in the NFB.

If you wish to try NFB, take a wire from the top tap of the output transformer secondary, through a resistor of about 15K - 18K (try different values experimentally), but not to the top of that 10K cathode resistor; you gonna put that voltage on your output secondary winding.  Split that resistor into 8.2K on top and 1.8K on the bottom, and attach the NFB loop to the junction.

This will give you 6-8dB of NFB to listen to, and not send the amp into oscillation.

Spearmint on!

Poinz

[jerluwoo]

I say that it doesnt work because of your phase splitter. As it is the tube is biased well into cutoff. Since you have the grid referenced to ground then you have that 56k resistor from cathode to ground biasing the tube. It would work like that if you were using a negative voltage supply, but not as is. Also the input tube is biased very close to cuttoff as well. It should only need a cathode resistor of 560 to 620 at most with the 100k plate resistor. Take some voltage measurements at the plate of the input tube and the cathode of the phase splitter tube and you'll see what i mean.