• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

phase splitter issue

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Ideal cathode bias would be at 1/4 B+, for maximum headroom on both ends...
1/3 is often preferred, I'm not sure why... (same idle voltage drop across both
triodes of a twin?).

If input ever exceeds B+/2, the concertina becomes forward biased diodes to
both the plate and the grid. Both ends locked together travel in same phase...
Whatever's driving a concertina usually can't supply DC current, so both ends
will not travel together far beyond that collision...

Bias at B+/2 or higher would be a huge mistake. I still see it often drawn that way.
 
Last edited:
Has anyone ever built a concertina phase splitter and have the signal be perfectly split but they differed in amplitude?


The output impedance from cathode is much lower than from anode.
Then, if the load impedance is too low - but equal at both outputs - the output voltages can be different.

What is the load impedance of both outputs ?
And what are the value of anode/cathode resistors ?
 
No, the same current flows through both impedances so equal impedance means equal voltage however low that impedance is. The complication is that the impedance includes everything attached, so what looks to be equal may not be. For example, if the anode supply is decoupled and the decoupling cap goes open circuit then the anode load has increased so the voltage will too.
 
The output impedance from cathode is much lower than from anode.

Not so. Please see my article in Linear Audio which examines this theoretically and experimentally.

If signals are different at cathode and anode, the loads are different. As DF96 points out, the load isn't just the load resistors... One common error is changing the load from one side inadvertently through measurement. To maintain the equal loading, the measurement has to either not perturb the circuit whatsoever (i.e., the measurement device has infinite impedance) or two identical measurement devices be used simultaneously.
 
The circuit enclosed using an ECF80 performs the function brilliantly with near equal swing symmetry. Expect 25V+25V rms swing and no excuse for signal voltage discrepancies.

richy
 

Attachments

  • concertina example.jpg
    concertina example.jpg
    16.9 KB · Views: 2,149
Sorry... 33Kto 56K......I've got several thous of these tubes from differing manu's, Aero, Mullard, you name it all with varying current requirements... this supply resistor adjusted to get roughly 95-110V on the anode. The concertina should easily drive an o/p pair to obtain 10-15W class A. A Zobel snubber would be probably required across the 100K 1st stage anode res, je noch design..o/p tranny etc.

Good luck

richy
 
The circuit enclosed using an ECF80 performs the function brilliantly with near equal swing symmetry. Expect 25V+25V rms swing and no excuse for signal voltage discrepancies.

richy

Yes, Ive used this 'scheme' with different tubes, and it does work extremely well....

Particularly like the 'DC' coupling from the gain-stage, and getting rid of the auto-bias and the coupling-cap of the phase-splitter.

Ive adapted it further to replace the pentode by a strapped triode and have a MOSFET on top, its bias set to place the anode of the triode at 110V, and the phase-splitter grid DC coupled by 470K to the anode of the 'mos-Mu' stage

Absolutely great sound.:)
 
Well, that's the classic voltage amp- cathodyne used in millions of amps. It does work great. For stability, I'd be inclined totie the two B+ together, then adjust the plate load of the voltage amp to get the right DC conditions. Or use an adjustable CCS- the cathodyne has a very low input capacitance, so doesn't load down the voltage amp.
 
Looks like we cross-posted there, Sy....

Your suggestion of the CCS in the plate-load of the gain-stage is exactly how I ended up, trimming the bias to get the correct voltage for the plate to grid voltage....

The CCS/MOSFET in the gain-stage really does make it more flexible and the sound is great too....
--I use it as my 'generic' gain/phase-splitter now....
 
The output impedance from cathode is much lower than from anode.

Not so. Please see my article in Linear Audio which examines this theoretically and experimentally.

I have to correct my earlier statement where I said that the output levels of different outputs are unequal if the load resistance is too low.

One reason to this confusion is caused by the fact that the output impedance from anode is much higher than from the cathode.
(other maybe is that I have done all my consructions and analysis with LTPs)

The Concertina phase splitter seems to be that sensitive to unequal loads that even a level measurement done with my HP333A distortion/audio voltage meter gives noticeable difference from anode and cathode.

I made few tests to fully demonstrate the effect of unequal load impedance and it's effects.

I built a Concertina with 6N1P and with 68 kohms anode- and cathode resistors. Then I used a 680 kohms load resistor to load both outputs separately to see what happens.

The results are as follows:

1. Voltage measured from cathode and load connected to cathode: no level change

2. Voltage measured from cathode and load connected to anode: no level change

3. Voltage measured from anode and load connected to anode: level decreases some 9 %

4. Voltage measured from anode and load connected to cathode: level increases some 9 %

From these results it can be seen that the output impedance from the anode is practically equal to Ra (68 kohms) and from cathode the output impedance is much lower. (the actual value can not be determined with 680 kohms load)

I also tried to determine the optimum grid voltage i.e. the voltage that gives lowest distortion with high output level.

With Uo = 25 Vrms, Ub = 300 V, the optimum grid voltage was 85 V which is some 28 % of Ub.
 
If the outputs are loaded equally (and that is indeed the case with the cathodyne in situ), the source impedances at plate an cathode are identical and low (~1/gm).

Then, if the load impedance is too low - but equal at both outputs - the output voltages can be different.

This is entirely incorrect. Please see my article (with measurements) in Linear Audio Vol 0.
 
The Concertina phase splitter seems to be that sensitive to unequal loads that even a level measurement done with my HP333A distortion/audio voltage meter gives noticeable difference from anode and cathode.

What frequency are you measuring at ? There is a slight cäpacitive difference on the concertina, esp noticeable on square wave... the HP döes have a sizeable input capacitance. Old trick; Use a 4mm to BNC adaptor and use a X10 probe and scale accordingly. Somewhere I recollect there is a RCA 7199 data sheet around which describes the capacitive issue, if I can find it.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.