diyAudio (
-   Tubes / Valves (
-   -   Transformer in the cathodes ?? (

Shoog 9th April 2009 08:41 AM

Transformer in the cathodes ??
Hi all,
I have been mulling over an idea for a simple little PP 6080 amp recently. My experience with this valve is that it has less than unity gain (despite the mu of 2). I was thinking how I might really drive the output impedance down without introducing gNFB. I originally thought of doing some cathode feedback (with the ratios involved it would be a simple matter to hand wind bifilar coils on the toroidal outputs).
However this got me musing about the possibility of putting the whole transformer into the cathodes and be done with it. This would perform very similarly to anode loading and also introduce the cathode feedback I wanted. The practical implementation would be to have CCS delivering 100mA in each of the PP pairs cathodes, and then tying the points above them together with the output transformer, which would be left floating between the cathodes. I would anticipate something like a 10ohm to 1000Kohm transformer.

Sounds like a neat idea - but what am I missing.


dsavitsk 9th April 2009 08:50 AM

The TCJ Push Pull calculator software has this basic design and this tube in it, so for $29 you can sim to your heart's desire.

Tweeker 9th April 2009 09:17 AM

It wont be simple once you ponder the driver stage. You need B+ + bias swing to drive it. It can be done, but it implies an additional psu for the driver or other complication.

Yvesm 9th April 2009 09:50 AM

Hi Shoog !

Wherever is the load -in plates or in cathodes- it must remain the same if you expect to obtain the same output power.

Your driver must deliver the usual "twice the bias" pk to pk level at each grid, plus the pk to pk level appearing between the cathodes, in short "very huge" but at almost no load, no Miller effect etc...

Consider using a bootstrap scheme ;)

My 2 pence.


Shoog 9th April 2009 09:55 AM

I know that the driver would be a bit problematic but there is a way of avoiding the need for an additional power transformer, which I have successfully used in the past.
Generate a negative rail from the main transformer and hang the driver stage off an input transformer. This kills two birds with one stone as it allows you to run a LTP with the whole gain of the triodes preserved, and secondly it gives you twice the voltage for the driver to play with.


Tarzan 9th April 2009 09:58 AM

A few days back I ran accross a schematic of an Electro Voice amp: Model A100.
This amp has the OPT in the cathode...
It can give you some ideas...

More here...

Worth a look.


Shoog 9th April 2009 10:04 AM


Your driver must deliver the usual "twice the bias" pk to pk level at each grid, plus the pk to pk level appearing between the cathodes, in short "very huge" but at almost no load, no Miller effect etc...
Since I will be running the 6080's at 100V and 100mA, which equates to about 30V bias, the actual driver voltage required is about 120V Peak to Peak, which should be doable with the anticipated 260V available from the two rails. This only leaves one further problem - how to get 120V from 2V input in a two stage design.


Shoog 9th April 2009 12:52 PM

I was just thinking about my proposal and came up with a slight problem. My original idea was to get the transformer to take over differential communication duties in the output stage. However even though the current seen at either end of the transformer would be the same and hence effectively zero, since the 6080's are likely to be mismatched the DC bias point for each side will be different so the DC voltage on either side will be sightly different. Since the resistance of the transformer is so low, even a slight voltage mis-match will produce a considerable current across the transformer - leading to saturation.

So I was thinking if I ground the center tap of the primaries and then place the CCS in the actual cathode leg of the 6080, between the triode and the transformer, this will ensure balanced current across the transformer, and the CCS will take up the voltage difference between the two legs. Unfortunately the tightly differential nature of the output stage is lost.

Having just thought about the last proposal, there is a chance that the CCS will spring up and down to the signal leaving their cathodes at a stable zero volts, ie no output. So a third proposal is in order. Split the primary into two halves, place a CCS between earth and one side of the primary, and the same for the other side. Current is forced to balance and the CCS will take up any voltage difference.


Tubelab_com 9th April 2009 05:12 PM

2 Attachment(s)
I have been experimenting with cathode follower based output stages for about two years now. They show considerable promise despite the drawbacks. The obvious drawback is the BIG drive voltage that is required. Assuming a "perfect" cathode follower (gain = 1) the drive voltage is equal to the voltage required across the output transformer. So the solution is to use a low impedance OPT and output tubes that can deliver big current. The 6AS7 is a start. I have switched to the 6336A or some big sweep tubes due to their high cathode current capability.

Cathode followers like all circuits are not perfect. The drive voltage requirements are higher than the output requirements. There is some distortion since the bias voltage on the tube varies with the cathode to plate voltage which varies with the signal, hence the drive voltage requirements are slightly non linear. What if we could keep the cathode to plate voltage constant across the tube?

Back in 1957 a brilliant engineer named MacDonald devised a circuit to solve both of these issues. I got so interested that I built a SE amplifier based on the augmented cathode follower circuit. It worked well. See this thread:

More cathode follower experiments are here:

These were all SE circuits but I have experimented with some push pull designs using some of the cheap sweep tubes that I got last summer on sale. Yes you can put a CCS in the cathode of each output tube, and tie the OPT between the two cathodes. I did this and it does work. You need to make the grid bias adjustable on one of the tubes to take out the offset. You can put an oil cap in series with the OPT if needed to solve the DC issue.

I have included a simulation of a push pull cathode follower. It uses a mosfet to keep the cathode to plate voltage constant. The simulator shows that 75 watts can be achieved in class A. The tube current is 300 mA which requires a 6336A or a big sweep tube. 600 V P-P of drive is required to reach 75 watts. I had a very similar circuit running on the bench last year. These experiments were all limited by my power supply and the available drive voltage. I now have a 1KW power supply and am working on a better driver board. More cathode follower experiments will follow.

Shoog 9th April 2009 06:15 PM

Thanks for that Tubelabs,
My needs are much more modest - a measly 7Watts will make me happy.
Looking at your circuit, how much voltage does the mosfet consume in doing its job.
I am currently running my preamp as a SLCF ala Allan Wrights FVP5. I am very happy with it and know the principles quite well. I have an improved version in the pipeline - but other things are drawing me away from finishing it.


All times are GMT. The time now is 12:28 PM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 18.75%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright ©1999-2017 diyAudio