True KP, but with less gain from the stage with NFB applied, the common cathode gain stage say, then stages after this need more gain no?
I only drew the OP stage, you use droppers for each stage or whatever, see VW - The Valve Wizard - last page bottom diagram.
Just bung a 470 ohm or 1k on it, suck it and see. You find everything becomes much clearer once you apply what you've read on the bench. You can read too much till your head feels detached from your body : )
Andy.
I only drew the OP stage, you use droppers for each stage or whatever, see VW - The Valve Wizard - last page bottom diagram.
Just bung a 470 ohm or 1k on it, suck it and see. You find everything becomes much clearer once you apply what you've read on the bench. You can read too much till your head feels detached from your body : )
Andy.
Yes, that's right. If you want the same gain as before when applying NFB (i.e. closed loop gain), you need to increase open loop gain. But this doesn't affect any signal level within the loop. This is what we've been talking about.
Best regards!
I would cross one Ocean at a time, either the Atlan tic or the Pacific but not both non stop, one after the another.
Or solve *one* variable first, not two at the same time.
Meaning either build a new circuit or a new layout; doing both at the same time introduces many variables and when solving problems which will certainly rise their ugly heads will be more complicated.
That said, the layout you show in pictures looks very sensible, why would you want to mess with it?
Or solve *one* variable first, not two at the same time.
Meaning either build a new circuit or a new layout; doing both at the same time introduces many variables and when solving problems which will certainly rise their ugly heads will be more complicated.
That said, the layout you show in pictures looks very sensible, why would you want to mess with it?
Resistors in a power tube G2-circuit should not be used to limit the voltage/dissipation there, only what is sufficient to quell oscillation. The one exception seems the EL34, where a relatively high 1K is used; this drops the maximum output somewhat, but shows better distortion behaviour.
This is because initially one sets out to keep Vg2 constant (variations affect maximum output power ability). Then one goes and make it 'unconstant' again by adding relatively high series resistors .....
El Sid, go look at the complete Mullard Valve data for EL84 (some 12+ pages). There you will find many characteristic graphs for a great many applications. Do note how the distortion varies.
One further effect (graphs for this allso given in complete EL84 data sheets): This is the so-called 'musical output' conditions (Warning: NOT to be confused with the PMPO nonsense!!) In these comparatively high capacitor values are used. The logic is that under typical musical conditions (low mark-space ratio, i.e. comparatively long low-signal periods between peak demands), large enough filter caps give a close semblance to fixed bias conditions most of the time.
Again going to relevant data sheets, it is shown that some 20% higher maximum output is available and at reduced distortion. (Apology for not showing dada sheets here; my copies are worn and creased from decades of use. Not from any periodical; my workplace subscribed to separate data sheet folders from manufacturers.)
I have often done this. Noticable difference very small, still. {By high values I mean power dc. caps of some 470µF - 2200 µF. I also regularly used cathode bypass caps of several thousand µF.
This is off the classic path but worth the effort in most cases.
This is because initially one sets out to keep Vg2 constant (variations affect maximum output power ability). Then one goes and make it 'unconstant' again by adding relatively high series resistors .....
El Sid, go look at the complete Mullard Valve data for EL84 (some 12+ pages). There you will find many characteristic graphs for a great many applications. Do note how the distortion varies.
One further effect (graphs for this allso given in complete EL84 data sheets): This is the so-called 'musical output' conditions (Warning: NOT to be confused with the PMPO nonsense!!) In these comparatively high capacitor values are used. The logic is that under typical musical conditions (low mark-space ratio, i.e. comparatively long low-signal periods between peak demands), large enough filter caps give a close semblance to fixed bias conditions most of the time.
Again going to relevant data sheets, it is shown that some 20% higher maximum output is available and at reduced distortion. (Apology for not showing dada sheets here; my copies are worn and creased from decades of use. Not from any periodical; my workplace subscribed to separate data sheet folders from manufacturers.)
I have often done this. Noticable difference very small, still. {By high values I mean power dc. caps of some 470µF - 2200 µF. I also regularly used cathode bypass caps of several thousand µF.
This is off the classic path but worth the effort in most cases.
JM you are indeed right, so what I am trying to do first is decide a) where to supply the g2 from, b) what the supply circuit will look like R, RC, RCR, etc, and finally the values.
Dankie Oom. Unfortunately I don't have the very detailed data sheets you do. I have the below graph for distortion in a PP application with Va and Vg2 around the voltages I have in mind. One question I have about high cap values, because I have thought about it, is when you are powering up and have high caps all over the place, may you not potentially have a very high current on startup as all these caps charge?
Last edited:
This project stalled a bit while i've been busy with the electrical design and work, life etc. I needed a bit of a break from thinking, so decided to make the prototype enclosure using bits and pieces of the exsting amp.
I dismantled this....
....and after much measuring, calculating, sawing, drilling, screwing and cursing, ended up with this:
Even as a rough prototype, I am pleased with how it looks. Already I have ideas for some improvements. As in the concept drawing, the base will be the same width as the top.
The power transformer will be inverted so the wires aren't exposed at the top and go through the base directly. As it stands the core is aligned with the OPTs so will rotate it. I originally wanted to show the maker's name but that's not sensible.
Unfortunately the insides are not yet populated
I dismantled this....
....and after much measuring, calculating, sawing, drilling, screwing and cursing, ended up with this:
Even as a rough prototype, I am pleased with how it looks. Already I have ideas for some improvements. As in the concept drawing, the base will be the same width as the top.
The power transformer will be inverted so the wires aren't exposed at the top and go through the base directly. As it stands the core is aligned with the OPTs so will rotate it. I originally wanted to show the maker's name but that's not sensible.
Unfortunately the insides are not yet populated
That actually looks pretty good. It's reminds me of 1930's power station racks or similar. I find it's therapeutic to do some metal bashing when all the theory clogs up the old noggin. It helps resolve the circuit, gives you something to pin your ideas on.
Well done, Andy.
Well done, Andy.
Well it's a year later 😡, and I've finally got some time to work on this again. All the external components have now been wired into the interior. I used some spare corrugated cable sleeve I had lying around (left over from installing a reverse camera on my car) to cover the wiring and look "industrial", as well as keep poking fingers away from the transformer terminals.
Next up is putting the power supply caps and resistors onto a piece of veroboard I'll attach to the screws holding the choke and wire that up.
I did make a cage for it it a few months back from some old wire mesh I had lying around outside:
Next up is putting the power supply caps and resistors onto a piece of veroboard I'll attach to the screws holding the choke and wire that up.
I did make a cage for it it a few months back from some old wire mesh I had lying around outside:
Last edited:
- Status
- Not open for further replies.