Shown below, I need the output stage cathode to sit at half the supply. I could float a second supply on the centre of the driver supply to keep the reference point clean. I have a CT HV secondary in which I haven't necessarily decided to use the CT.
To save on space/transformers, and maybe chokes/diodes, can I somehow tap another rail from the secondary that is balanced on it, but floats upon the middle point?
Edit: Have considered a voltage doubler.
To save on space/transformers, and maybe chokes/diodes, can I somehow tap another rail from the secondary that is balanced on it, but floats upon the middle point?
Edit: Have considered a voltage doubler.
Attachments
Last edited:
Your 'output stage' won't have much gain with 1k of cathode degeneration.
However, to answer your question, yes you can make a centre-tapped supply from a centre-tapped secondary. The SS guys do this all the time. Just be careful to keep the charging pulses in their own loop, unless you want lots of buzz.
However, to answer your question, yes you can make a centre-tapped supply from a centre-tapped secondary. The SS guys do this all the time. Just be careful to keep the charging pulses in their own loop, unless you want lots of buzz.
There's not meant to be any cathode degeneration, but a direct connection to the balance point with low source impedance. The driver output sits at around 40V below this holding the output grid. The output B+ relative to its cathode is about 240V.
By the way the 14k resistor shown is just making DC balance... an incomplete fix and probably just confounding the schematic. My problem has been making a low enough Z tap to keep the stage decoupled from the supply, and it is further decoupled through a single smaller cap and series resistance.
The reason for picking the centre point for the output cathode is partly for the DC, and partly for the noise coupling conditions between driver and output, ie. ripple. But the problem is the stage modulating the supply.
I've considered tying the cathode to ground for AC while dropping 180V DC at the same time, (and simply reducing the ripple to the driver stage to deal with that part of the issue). Although putting a large electrolytic cap or other voltage source in series with the output doesn't really appeal.
Then there's shunt regulating across the stage to maintain a fixed voltage on the resistor below, which would drop the required lower half of the supply but would also collect distortions from imperfect regulation.
The driver supply is a balanced tapping of the previous supply stage (from which the output supply is derived) therefore if I keep the centre tap away from both these levels of the supply, the middle of the output and driver rails will coincide with each other. All I should need to do is keep the cathode in the middle of the rail which I'm not doing.
If I were to remove the lower 1k resistor I'd still more or less need to do the same thing and keep the cathode still?
If I were to remove the lower 1k resistor I'd still more or less need to do the same thing and keep the cathode still?
No, the output triode is fed from a hard centerpoint to +170V , to boot through an extra filtering 1k2 resistor, so it will have at most 140V available.
This looks like a very low power stage, say 1W RMS or less, a headphone amp?
And showing the input cap is, at least, misleading; you should use a connector symbol there.
With due respect, the design in general does not make much sense.
The somewhat obsessive multi inductor filtering does not certainly help.
Sorry.
Yes less than 1W, for use above 250Hz on efficient speakers.
Secondary is 385-0-385rms, sorry. This is just a screenshot from my simulator part way through working on this.
There is a mu-follower then common cathode driver which is direct coupled to the output stage. The mu stage uses a triode (lower) and a pentode. The intention is.. that capacitive coupling to the driver is a passive line level crossover for 250Hz. The driver is a 6V6 with cathode degeneration, 15V of bias and the plate sits at 220V wrt ground, holding the grid of the output whose cathode is 40V higher, and plate is another 230V away. It will also be a DHT.
This began as a balanced supply and as such, keeping noise zeroed at the centre point is the direction I'm considering with the direct coupling. This doesn't work for the input stage so ground has been reserved for the input reference. (A second benefit would be the relative lack of noise on the rail chosen to connect with the transformer core through ground compared to a conventional single-ended supply.)
Do you think this is not a good idea? How?
Do you have a suggestion for my first question?
That's not the schematic of anything, you are just cutting and pasting snippets at random and on first sight it may look like something functional ... but it is not.
By the way, you are grossly misstating voltages involved (now we know that 170V is actually 385V) , didn't even care to state the tubes involved, you are drawing a 6V6 as a triode, call a "mu follower" a triode with a constant current load, still show the input as grounded, so much nonsense that this is not even a circuit which can be discussed.
What could I correct?
Where to begin?
Where to end?
Most important: still don't know what you are trying to do .
Your question was:
Yes, I already answered that on post #9:
meaning: your cathode is already hard connected (no "floating" needed) to half supply; its plate is connected to full supply, or 170+170V .
That 170 must be read 385 does not change the issue, cathode always sits at half supply.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.