Help on CCS PP differential driver stage for SE 845 amp

[swordsman]

Hi,

I'm trying to build a 2 stage amp using a twin triode tube via a 5K push-pull interstage transformer to drive a 845. I'm reusing a old chassis therefore the holes are cut for only one driver tube. The driver voltage swing needed is 400Vpp so as to provide some headroom into class A2. This input sensitivity of this amp should be able to be driven to max output power with a standard line in voltage eg, a CD player.

With such a high requirement on the driver, i'm thinking of a differential driver stage with a CCS on the tail. This will be similar to what allen wright did for his pp-1c el34 amplifier but not with the Cascade Triode stage inorder to get the lowest output impedance. The low impedance is necessary for the high capacitive 845 input.

Any genius care to share what tube can i use for this function and how i can implement one?

http://www.vacuumstate.com/schematics/pp-1c_s.gif

 

[SY]

Since the loading of an A2 (or AB2) amp is so nonlinear, I'd want a buffer between the diff amp and the outputs. Preparing to duck for cover, I will suggest that, lacking space for another tube, a couple of source followers would do the trick.

 

[Shoog]

May I suggest that in order to achieve the maximum gain from your LTP you need to drive both of its inputs (doubling your gain). A good way to achieve that would be to move your interstage transformer to the front end. I also think you will need to load both your LTP outputs down to make certain that balance is maintained with only one of the outputs facing a grid. I second SY suggestion about using a source follower. Infact if you do that then some simple grounded grid stage should achieve what you want without the need for the LTP.



Shoog

 

[swordsman]

details on driver requirement

I'm using fixed bias and will be running lower than usual voltage on the anode of the 845. For full power output, the peak voltage needed at the grid of the 845 is 115Vp, which is about 82Vrms. With a typical line input of 2Vrms, the gain needed is in the driver stage is 41. For a overhead of 3db in the driver stage, the needed gain will be approximately 60.

Using a differential pair, the gain is shared between the 2 triode section. The gain needed per section is therefore half which is 30. If you still follow me so far, each triode section will take a input of 2Vrms*1.414 = 2.83Vp and swing around 85Vp. This is not far from driver requirement in a 300B amp. There should be tubes available for this purposes.

One way to see my suggested circuit is to see the driver stage as a push pull amp with the interstage transformer as the output transformer.

 

[Shoog]

I see what you are planning, and it seems a good idea. However by only driving one input you achieve half of the gain of each of the triodes. In the case of say an ECC83 you have a mu of 100 so the max gain you can achieve is about 50. Drive both the inputs and that jumps to 100. I sure you know this - but half of the LTP gain is used to compare and amplify the ground reference.

Shoog