Cathode follower with impedence balanced drive

[Tubetec]

There are plenty of online coupling cap calculators , some will even present you with the FR graph but they generally dont offer the component value on a slider with fast auto recalculate after you touch the control , instead your stuck typing numbers in boxes . Maybe theres a simple way to overlay the screen shots to build up a better picture of how a switched HPF would look .

The idea of using a cathode follower as the drive for an impedence balanced line came to me in my dreams the other morning , I half forgot then bang the light bulb flashed again .

Re-doubling my efforts to make the HPF fully balanced makes loads of sense , as it means I also have the option of running balanced sources to the output stage , the fact that every thing in the output stage is then symetrical , incuding the path to input grids should mean any PSU induced hum is very effectively canceled in the output transformer .

The other idea I had was the possibility using both grids to mix two unbalanced signals into a single output .
The output stage is based around the BBC standard ECC81 circuit used in their equipment of a certain vintage , notice the boot lacing on the grids /cathodes .
Id add a parallel fed output transformer to this set up so the theres two sets of outputs ,
The other interesting point about these pre amps is the PSU is very simple ,theres no local decoupling on the HT where it hits the OP TX's centre tap ,


Full details here ,
https://www.bbceng.info/ti/eqpt/C_9.pdf

 

[abbey road d enfer]

The output stage is based around the BBC standard ECC81 circuit used in their equipment of a certain vintage , notice the boot lacing on the grids /cathodes .

Weird arrangement IMO.
DC-wise there is positive feedback, which suggests problems with quiescent current stability.
AC-wise, cross-coupling is insignificant due to the grid capacitors.
The link says this circuit is described in the BBC journal in conjunction with LIM/5. I'd be curious...

I know, who am I to have the gall to criticize the BBC engineers?

 

[Tubetec]

The cross coupling arrangement helps balance out tube mismatch ,
heres the Lim/5
https://www.bbceng.info/ti/eqpt/LIM_5.pdf

 

[Tubetec]

Most preamps of this kind use a pot to adjust standing balance in the OP TX , this method dispences with the need for any adjustment .

 

[abbey road d enfer]

The cross coupling arrangement helps balance out tube mismatch ,
heres the Lim/5
https://www.bbceng.info/ti/eqpt/LIM_5.pdf

Thanks for the link. However, the claim that it "stabilizes both valves against individual anode-current changes" does not stand the test of scrutiny. On the contrary, any increse of current from one of the tubes is communicated to the other in a cumulative way, because of the DC positive feedback.
The reasoning is similar to the paradox of Zeno of Elea, which concludes that the arrow can never attain the turtle (or Achilles can never outrun the turtle).
The claim that it "ensures equality of anode currents "is correct, though.

 

[Tubetec]

The cap across the cathode is shown as non polar , yet a value of 2500uF, it has to be electrolytics back to back.

 

[Tubetec]

https://www.lundahltransformers.com/wp-content/uploads/datasheets/1930.pdf
This was what I had in mind as the parallel feed TX , for the sake of argument the primary windings directly connected to the anodes , a single cap to ground at the centre tap point , saves the need to match two components , transformer is capable of handling kV across the windings in any case, 300 or so volts wont bother it .

I see the LL1930 shows only single ended input in the wiring diagrams , windings in parallel ,
For push pull windings are in series , what ratios/impedences do I end up with when I series/parallel the secondary ?

 

[abbey road d enfer]

This was what I had in mind as the parallel feed TX , for the sake of argument the primary windings directly connected to the anodes , a single cap to ground at the centre tap point

I believe ther's a misunderstanding here; parafeed means the anodes are powered via choke(s), so tehre is no DC in the xfmr. I think what you describe is push-pull connection, where there is DC of equal value in each half primary.

, saves the need to match two components ,

Whar two components?

I see the LL1930 shows only single ended input in the wiring diagrams , windings in parallel ,
For push pull windings are in series , what ratios/impedences do I end up with when I series/parallel the secondary ?

 

[Tubetec]

LL1660 was what I was going to use as inductor , its a transformer with 6 windings , it has very high inductance and perhaps well balanced stray capacities across the windings , maybe its possible to wire a primary and two secondaries each side of a centre tap as an even larger inductance , alternatively tap a feed back or output signal with it configured as a either 2.25+2.25 : 2+2 for bridged load and 2.25+2.25 : 1 for lower loads .
www.lundahl.se/wp-content/uploads/datasheets/1660.pdf

 

[merlin]

Thanks for the link. However, the claim that it "stabilizes both valves against individual anode-current changes" does not stand the test of scrutiny. On the contrary, any increse of current from one of the tubes is communicated to the other in a cumulative way, because of the DC positive feedback.

The idea is that the strong tube forces the weaker tube to conduct more, but the weaker tube acts as deadweight on the stronger tube, so they kinda meet in the middle.
https://www.tubecad.com/2009/04/blog0163.htm(It has nothing to do with Blumlein btw, Broskie seems to have got his wires crossed there)

 

[abbey road d enfer]

maybe its possible to wire a primary and two secondaries each side of a centre tap as an even larger inductance

I don't see how it would be possible. The "center tap" would not be centered. That would be "possible" if the ratio was 2+2:1+1+1+1
I don't think you really need a higher inductance, particularly in view of the possible HF losses due to increased parasitic capacitance.

www.lundahl.se/wp-content/uploads/datasheets/1660.pdf

Some of these datasheets are a little cryptic.

 

[abbey road d enfer]

The idea is that the strong tube forces the weaker tube to conduct more, but the weaker tube acts as deadweight on the stronger tube, so they kinda meet in the middle.
https://www.tubecad.com/2009/04/blog0163.htm

This explanation doesn't work for me.
When tube A conducts more, it raises tube B's grid voltage, which in turn increases its cathode voltage, which results in increasing even more the grid voltage of tube A.
Since the DC gain is less than one, it does not result in catastrophic runaway, but still I can't admit it is more stable than basic cathode-bias.
I fully agree with the self-balancing effect, though.

 

[Tubetec]

More here ,
https://www.tubecad.com/2005/May/blog0046.htm

 

[abbey road d enfer]

It rightfully does not claim better quiescent stability, just better balance.

 

[Tubetec]

Then again the BBC will have had many thousands of these amps in service ,
instead of traditional cathode bias with a pot controling the balance that needs a service tech in a white coat to adjust every once in a while , the Blumlien circuit means good balance even as the tubes start to drift and age and no current metering circuit is needed .

 

[abbey road d enfer]

Then again the BBC will have had many thousands of these amps in service ,
instead of traditional cathode bias with a pot controling the balance that needs a service tech in a white coat to adjust every once in a while , the Blumlien circuit means good balance even as the tubes start to drift and age and no current metering circuit is needed .

I don't contest the rightfulness of the claims for better matching. I just think the assertion regarding supposed increased stability is an unsubtantiated and unnecessary claim.