I could use this for another thing...more exciting for me!
What are you using as primary load in the OPT? 5K?
Yes, I assumed 12:1 for the transformer turns-ratio, so about 5400 ohms.
The shunt cascode circuit is well-proven - I use 3 stages of this circuit to get from MC (Denon DL103) to 300B grid!
I assume, you wanted 249.4V Vcascode, instead of 250.6V. And even if you set it precise, and well stabilized, output tube will see very random and dependent on weather in Zimbabwe bias voltage. And if your servo in cathode of an output stage will somehow compensate it, anode current of the first stage tube will depend on Amazon rainforests.
Last edited:
hi Anatoliy,
That's OK. The dc level of the driver output, nor the supply voltage of the driver (nor the output stage) are not precisely specified.
The servo sets the output stage current, regardless of any of the voltages - within reason.
The cascode voltage should be regulated for best performance - not for dc reasons, but because we prefer low impedance drive of the base voltage.
It's OK - we are not using the BJT for amplification.
Its job is simply to fix the anode voltage of the triode, and steer the triode's current variations into the load resistor.
The sound of it surpasses the usual triode stages easily, and often you can replace 2 stages of triodes with one shunt cascode.
Comparing the sound of shunt cascode (eg EF80 triode) driving 300B to the often-seen 2-stage 6SN7 is very revealing.
I would like to try with the 45 PP to get 17W. The bias of the output stage is the same for the 5.5W class A amplifier and the 17W class AB2 amplifier (i.e. 275V/36 mA and 3.9K). Only difference is the drive. The OPT in your design is replaced by a SE-to-PP interstage like the Lundahl LL1660 or LL1692A.
Last edited:
Despite of CCS anode current supply, load line of the triode is almost vertical. And "steering" is not linear, since collector current is the sum of emitter and base currents. Base current is not linear.
Yes, this could be interesting. Maybe worth a new thread?
For the interstage maybe Sowter 8423 also worth a look -
INTERSTAGE TRANSFORMERS FOR VALVE TUBE AMPLIFIERS
That's correct. This is intentional! Many triodes, and triode-connected pentodes work excellently this way.
Not quite correct. Collector current is the **difference** of the emitter and base currents. And if Darlington connexion is adopted, almost all base current can be returned to the output - making the difference extremely small.
Or, we can use PMOS device to eliminate it.
When you consider -
- the high gain (frequently good enough to replace 2 triode stages)
- no signal current in power supply
- input and output both referred to ground
- very high power supply rejection
- easily adapted to dc coupling
it becomes clear how useful this architecture is.
Yes the Sowter is good as well.
I will send you a pm, sooner or later, if you don't mind.
End of OT.
Last edited:
Pardon, mistyping. Of course collector current is the difference, but base current depends on collector current non-linearly. Look at curves of dependence of beta on collector current.
Rod, I would never use such "steering" without negative feedback, and I would suggest you to close servo loop through the 1'st anode current source. Such a way you will keep all together under control.
Try it in non-virtual reality and see what I mean.
Rod, I would never use such "steering" without negative feedback, and I would suggest you to close servo loop through the 1'st anode current source. Such a way you will keep all together under control.
Try it in non-virtual reality and see what I mean.
Please see the Hfe characteristic of the 500V PBHV9040Z. This excellent new NXP transistor has effectively constant Hfe across the full range of current required in our driver applications.
http://www.nxp.com/documents/data_sheet/PBHV9040Z.pdf
And with darlington connexion, the base current error is almost totally removed.
It's OK - I have tried it for the last 8 years in the driver for my 300B, completely open-loop. Very real reality!
Still, I agree that dc-stabilisation of the current source (or the triode bias) might be useful, and have been considering circuits for this.
Attachments
About the sound I cannot say anything. But I cannot agree that the 2N5401 in the circuit you posted is not used in amplification. From a 4P1L you can get about 20dB gain, max. The gain at the collector of the bjt is about 30dB. If the bjt doesn't amplify you have to ask yourself where the extra 10dB is coming from? The output stage can be safely left out.
4P1L in his device has current output driving emitter of the "common base" BJT stage that is loaded on resistor. Both devices amplify the signal. However, if Rod likes the sound of his setup, and is satisfied by it's stability of parameters, it is his choice. But as I said, I would stabilize DC parameters of all directly - coupled thingy together, and definitely would add negative feedback around it. Religious people then would have a chance to compare and decide, which one is better according to their religion, by sound only, without knowing what ius inside.
Seems to me it's a classic transimpedance amplifier.
I completely agree. I was just pointing out that it's not a two stage amplifier, rather it has three stages, the first two being formed by two DC coupled active devices: a tube and a bjt.