[Amps] Designing the Cleanest Linear with RF Negative Feedback

[Tom Cathey]

Hi Bob,

Thanks for the suggestion.

You make a good point of possibly eliminating the 6146 pre driver. I thought 
about that myself, but didn't have enough knowledge to know if I would have 
enough gain left to reduce the overall gain with  20 db of negative feedback 
afterwards.  Figured I'd just try it and see.

Well, I know that the 4CX-XXXX final [they're all similar] will need about 
300 volts of swing. And the OPA-695 will do 10db gain. That's 5mW to 50mW. I 
figured I'd put two OPA-695's  in cascade for 20db = 500mW.  Into 50 ohms, 
500mW =  5V.    So we have 5V to drive the 4CX-250 driver.  And, the 9:1 
transformer will make it X3 = 15V, right?  This is a 9:1 impedance ratio = 
3:1 voltage ratio...?

Anyway, with 20 db of negative feedback from the final to the 4CX-250 driver 
will mean a 20db reduction in gain.  I suppose a lot depends upon the gain 
of the 4CX-250 driver and the transformer.  This is where I am not sure how 
to calculate it and proceed. [20 Log db for voltage, etc.]

ie, Will an OPA-695 putting out 500mw into an X? step up ratio transformer 
that drives a 4CX-250, that drives 300v into a 4CX-XXX X final  with 20db of 
negative feedback have enough system gain?  And, will the 9:1 voltage step 
up transformer have absolutely no affect on the low distortion goal... even 
if a small space charge current is pulled? What if we need more than 9:1 
step up, can it be done with cascaded transformers without ill effects? 
BTW, I'm curious why you feel the transformer will have bandwidth 
limitations... are we talking of a broadband unun torroidal type?

Opinions?

73,
Tom, K1JJ



> Before you close the case Tom something else for you to consider. To drive 
> a GG Amp you place watts on the cathode. To drive a Grounded Cathode Amp 
> you place volts on the grid. This provides you with a cheap and easy way 
> to increase the gain of any stage. Simply insert a voltage transformer in 
> the input. With a 1 to 9 unit on the input of the 4CX250 stage you should 
> be able to chop out the 6146 and drive it directly with the OPA-695. The 
> draw back of this method is the available operating bandwidth. Seeing as 
> you intend to use this in a monoband application this should not pose a 
> problem. The other thing to consider is the distortion generated by 
> saturating the transformer core. Hardly a problem worth considering at 
> that power level.