Help needed with white cathode follower input Z / SRPP driven

[KrIVIUM2323]

Hi,
could someone explain me how to determine/calculate the input Z of the circuit in the following datasheet please:

http://www.jensen-transformers.com/datashts/10k611m.pdf

In next schematic it is said the 12au7 SRPP could be direct coupled to the white cathode follower of 10k611m datasheet but i don't understand choosen RK values of 1,2k in srpp.

http://www.jensen-transformers.com/as/as054.pdf

I tried to work backward using formulae given by M.Blencowe,

Rload= Rk(µ-1,5)-Ra/2

but it doesn't compute.
I'm surely missing something but don't see what.

Thanks.

 

[ruffrecords]

Hi,
could someone explain me how to determine/calculate the input Z of the circuit in the following datasheet please:

http://www.jensen-transformers.com/datashts/10k611m.pdf

It is very very high as it is just the grid of the tube. Essentially the input impedance looks like a capacitance of a few picofarads.

In next schematic it is said the 12au7 SRPP could be direct coupled to the white cathode follower of 10k611m datasheet but i don't understand choosen RK values of 1,2k in srpp.

http://www.jensen-transformers.com/as/as054.pdf

I tried to work backward using formulae given by M.Blencowe,

Rload= Rk(µ-1,5)-Ra/2

but it doesn't compute.
I'm surely missing something but don't see what.

Thanks.

It is most likely chosen to set up a suitable quiescent current. In this application there is not a fixed load so optimisation formulae like Blencowe's do not apply. At dc the tube looks like a resistor roughly equal to mu.Rk plus ra which I reckon to be about 30K. So the quiescent current is about 250/60 mA or about 4mA.

Cheers

Ian

 

[KrIVIUM2323]

Hi Ian,
thank you for your answer!

It is very very high as it is just the grid of the tube

This is exactly what it seemed to me.

In this application there is not a fixed load so optimisation formulae like Blencowe's do not apply.

When you say application you're talking about the circuit presented in as054 by Jensen not the combination of the two circuits, right?

This is where i'm lost and don't understand choice of SRPP for this last amplifier.  Mr Jensen know about his stuff so there is surely a reason for this choice.

Ok, let's say that the two applications presented here are a whole circuit and i would like to 'optimise' the rk in SRPP to drive the WCF, -assuming it is interesting for operation, which could be pointless as load is very high which is better that the other way around with SRPP- which value of Rload to choice?
470K, 1m, more? Pointless and better stay with same calcution you gave to determine convenient quiescent current?

At dc the tube looks like a resistor roughly equal to mu.Rk plus ra which I reckon to be about 30K. So the quiescent current is about 250/60 mA or about 4mA.

I suppose you mean 250/ (30K+30K)= 0,004ma approx.

 

[ruffrecords]

When you say application you're talking about the circuit presented in as054 by Jensen not the combination of the two circuits, right?

Right, there is no mention of what the load might be in the mic stage schematic.

This is where i'm lost and don't understand choice of SRPP for this last amplifier.  Mr Jensen know about his stuff so there is surely a reason for this choice.

I agree and neither do I understand the choice of cascode for the first stage. Yes, it gives you lots of gain but it is well known for its poor distortion. Equally, the SRPP stage produces more distortion than other topologies (like the mu follower for example) and the 12AU7 is well known for its poor distortion so overall I would say Mr. Jensen was probably aiming for a colourful tube design.

Ok, let's say that the two applications presented here are a whole circuit and i would like to 'optimise' the rk in SRPP to drive the WCF, -assuming it is interesting for operation, which could be pointless as load is very high which is better that the other way around with SRPP- which value of Rload to choice?
470K, 1m, more? Pointless and better stay with same calcution you gave to determine convenient quiescent current?

4mA is a reasonable current at which to run a 12AU7. I expect it will drive a 10K load with no problem so if you were using an unbalanced system you could take the output direct from the SRPP. However, the White follower has unity gain and the output transformer loses 12dB or 18dB so this design needs 12dB/18dB of extra gain to make up for this loss and this is pretty much what the SRPP stage provides. The overall gain from the mic is therefore mainly due to the input transformer (20dB) and the cascode circuit (40dB?). There are others ways to skin this particular cat but all the stages add up to an OK if coloured overall design.

At dc the tube looks like a resistor roughly equal to mu.Rk plus ra which I reckon to be about 30K. So the quiescent current is about 250/60 mA or about 4mA.

p

I suppose you mean 250/ (30K+30K)= 0,004ma approx.

Yes.

Cheers

Ian

 

[KrIVIUM2323]

Ian thank you very very much!

Ok i did some calculation using different Rload and it now make much more sense. 

4mA is a reasonable current at which to run a 12AU7. I expect it will drive a 10K load with no problem so if you were using an unbalanced system you could take the output direct from the SRPP.

This is what i end up with, rk choosen to give quiescent current with correct margin for tube behavior.

This raise a new question for me, 12au7 is near equivalent in noval of 6sn7. For common ground cathode circuit i've read and see often quiescent point of approximately 8ma as it's supposed to be more linear for 6sn7. Do you think approaching this quiescent current will be of benefit with srpp?

There are others ways to skin this particular cat but all the stages add up to an OK if coloured overall design.

I agree.

One more time THANK YOU Ian!
Cheers.

 

[ruffrecords]

This raise a new question for me, 12au7 is near equivalent in noval of 6sn7. For common ground cathode circuit i've read and see often quiescent point of approximately 8ma as it's supposed to be more linear for 6sn7. Do you think approaching this quiescent current will be of benefit with srpp?

In my experience the 12AU7 is nothing like a noval version of the 6SN7 but the 6CG7/6FQ7 is an exact noval equivalent of the 6SN7. However, that is beside the point. It is generally true that triodes are more linear at higher currents so a 6SN7 is likely to be more linear at 8mA than at 4mA and the same is true for the 12AU7. This does double the plate dissipation from about 500mW per plate to 1Watt per plate which is within spec so it should be safe. However, no matter where you set the operating point, the fact remains that the 12AU7 has three times the intrinsic distortion of a 6SN7 so whether this is a worthwhile improvement is anyone's guess.

Cheers

Ian

 

[KrIVIUM2323]

6CG7/6FQ7 is an exact noval equivalent of the 6SN7

Thanks, it enlights some choice of tube in some famous vari mu circuits final gain stage, i was wondering about too.

12AU7 has three times the intrinsic distortion of a 6SN7 so whether this is a worthwhile improvement is anyone's guess.

 

I've seen that before in Moragn Jones's 'tube amplifier' i think. When i said equivalent i was thinking about values and parameters. I definitly prefers octals to novals in trials i've made.

Thanks fro the input.
Best regards.