Wow. What is that thing 
What an amazing beast.
There is always the compactron, I guess. Three valves in one. For the moment I think I have to admit you need two valves for a guitar amp.
Hmm. Two ECL82s. You could have two triodes as preamp and the pentodes in parallel for class A output. Or even use one of the pentodes used in the preamp for loads of gain.
Pete
What an amazing beast.There is always the compactron, I guess. Three valves in one. For the moment I think I have to admit you need two valves for a guitar amp.
Hmm. Two ECL82s. You could have two triodes as preamp and the pentodes in parallel for class A output. Or even use one of the pentodes used in the preamp for loads of gain.
Pete
I think Compactrons were only made and sold in the USA, so they may be a little harder for you to source in the UK.
You also need a weird 12-pin Compactron valve socket (there were new ones available when I bought them around 2010, but I have no idea of current availability.) And then there is the trouble with unwanted feedback between the closely spaced elements inside the single glass envelope.
Someone on this thread pointed out that there are some valves that contain two pentodes in one bottle. (Maybe they are compactrons, since it would seem you need at least 10 pins for two pentodes plus heater?).
You can certainly get a lot more voltage gain out of a pentode than a triode, but will it be enough? And will unwanted electrostatic coupling and instability rear their heads? I don't know.
I've read that you get better-sounding valve distortion by slightly distorting several successive valve stages, rather than by massively overdriving one single stage. My own limited experience tends to support this view. So there may be some good reasons to use multiple valves in a guitar amp, at least if any kind of overdrive is intended.
So my own interests drifted away from using a minimal number of valves, and towards using oddball dirt-cheap valves instead. I think it's fun to make a nice-sounding amp with a handful of valves that only cost you a buck or two each.
Lots of fun possibilities there!
There is also the possibility of using the two triodes to make a long-tailed-pair (or differential amplifier, as we solid-state types generally call them) phase inverter, and then using the two pentodes as push-pull outputs.
This will be a complete push-pull power amp with only two valves. You would still need a preamp, but could get away with only one 12AX7 there, like those "18 watt" Marshalls everyone got so worked up about a decade ago.
One nice thing about this arrangement is that one could cross-wire the triodes and pentodes from the two bottles, which would make the output from each bottle out of phase with its input. Perhaps this will help stability.
For a while now, I've had an amp like this on my mental list of "Projects I'll build some day." Only I usually think about the 6LY8 and 6HZ8 valves, both of which I already have in the junkbox, err, treasure chest, and not the ECL82.
-Gnobuddy
I know there is a reason, but in the spirit of asking the impossible:
Given the topology of the long tailed pair.. why not use a pentode LTP driving a centre tap transformer directly? The output pentodes become their own phase splitter. Just wondering.
For me, two ecl82 or 86. The triodes are the two stage preamp with time stack. The pentodes run in parallel single ended. If you want it you can get twice the power. All from two valves.
Of course push pull pentodes would give you four times the power, but you use at least one triode up as a phase splitter. Whatever you do that way you end up with three glass bottles..
Given the topology of the long tailed pair.. why not use a pentode LTP driving a centre tap transformer directly? The output pentodes become their own phase splitter. Just wondering.
For me, two ecl82 or 86. The triodes are the two stage preamp with time stack. The pentodes run in parallel single ended. If you want it you can get twice the power. All from two valves.
Of course push pull pentodes would give you four times the power, but you use at least one triode up as a phase splitter. Whatever you do that way you end up with three glass bottles..
I'm sorry to say, I don't think this will work out - this layout will have the same (inadequate) voltage gain as the amp Snaarman has just built. The cathodyne has 0 dB voltage gain (unity), the output stage will typically lose a few decibles (less than unity gain), and there will be only the gain from the first triode. Which isn't enough.
What you described is a typical topology for just the power amplifier section of many classic guitar amps - a Fender Princeton Reverb, for instance. But you need a preamp with two more triode gain stages in front of this if you plan to also use a lossy tone control ("stack"), and that is still only enough gain for clean tones or very slight overdrive.
-Gnobuddy
You can use a centre-tapped transformer as a phase splitter, certainly. If you do, you do not need a long tailed pair, which has only half the voltage gain of a single triode!
The catch is finding a suitable transformer, and driver triodes beefy enough to handle the much heavier load of the coupling transformer primary.
That heavier load in turn is dictated by the fact that you can only wind so many turns of copper wire on a reasonable-sized transformer. The reactance of the primary coil has to exceed the intended input impedance at the lowest frequency of usage. If, say, you wanted 100k input impedance at 83 Hz (the lowest guitar note in standard tuning), you would need a primary inductance of 192 Henries, which is far too big to be practical.
So in practice you will be stuck with, at best, something like a 10k or 20k load for the driver triode. This immediately rules out high anode resistance (ra) triodes like the ones in a 12AX7. I don't think the triodes in an ECL82 will cope, either - datasheet mu of 70 and gm of 2.2 mA/V combine to suggest an anode resistance of about 32 kilo ohms, too high to drive a 10k interstage transformer.
There is a twisted evil thing called "self-split output stage" found only in tiny DIY amps. It is a terrible idea, and best avoided. I don't think this is what you meant, but someone might bring it up. If they do, just say "No!"
Wouldn't that put the input stage (triode) in the same bottle as the output stage (one of the two pentodes)? The close proximity might cause the same instability issues the Compactron amps suffer from.
-Gnobuddy
Got a link to share? My Googling comes up useless - no schematics / circuit diagrams, just legalese.
Google is the software version of Uriah Heep - a narcissistic sycophant who tells you what you want to hear, but only in a way that's always self-serving.Google software is constantly trying to figure out what your interests are based on your previous search and browsing history, and displays results accordingly.
This means that when I Google for "singled ended push-pull", I almost certainly see something completely different from what you see when you Google the same phrase.
-Gnobuddy
Switchable Single-Ended, Push-Pull Guitar Amp
and IEEE reference some se/pp stuff, but I was thinking of the self split topology also - not sure why that has been dismissed ( it is much easier for me to find an adequate pp OPT than an se OPT - I need to look for workarounds like self-split)? Snaarman was close, seeing the similarities with a LTP/diff amp.
and IEEE reference some se/pp stuff, but I was thinking of the self split topology also - not sure why that has been dismissed ( it is much easier for me to find an adequate pp OPT than an se OPT - I need to look for workarounds like self-split)? Snaarman was close, seeing the similarities with a LTP/diff amp.
Last edited:
Thanks for the link!
The word "switchable" makes all the difference. This is a perfectly ordinary push-pull amp, but you can choose to switch off one output valve, and let it idle.
From what I can see, the concept works perfectly to cancel DC if both valves are biased to operate in pure class A. If in class AB, there will be some net DC current through the output transformer in single-ended operation, but considerably less than a single valve in single-ended operation.
Here's why:
1) For well balanced phase splitting, you need the largest possible tail resistor, ideally, a current source.
2) To deliver output power, you need the smallest possible tail resistor, ideally zero.
So you have a fundamental engineering conflict. Either you get crappy phase balance and heavy distortion, or lousy efficiency and limited power output.
It gets worse:
3) The second valve is driven by the cathode of the first. This means there is no drive to the second valve if the first valve cuts off. This means the first valve has to operate during the entire audio cycle. This means even crappy poorly balanced "self split" only works if the output valves are biased to pure class A - which, once again, limits power to the speaker, and is very inefficient.
There's a comedy TV show in the USA known as "Saturday Night Live". Many years ago, they did a skit about a new (fictitious) consumer product, Shimmer:
Audio: YouTube
Description: SNL Flashback: 'New Shimmer', the floor wax that's a dessert topping ~ Holdout Sports
Shimmer, as it turns out if you listen to the clip, is both a floor wax and a dessert topping. Cue laughter...
That's the trouble with a self-split output stage. It's trying to be both a floor wax and a dessert topping (both a phase splitter and an output power stage). It ends up being awful at both tasks.
Understood. You could use a second valve or a MOSFET current source to draw just DC current through the second half-primary of the OT, like the "switchable SE/PP amp" you've been discussing. That would let you use the PP OT in a single-ended amp.
The "self-split" circuit, on the other hand, is a can of worms, requiring mutually conflicting engineering requirements that cannot be met. One of these can spit out a few tens of milliwatts of heavily distorted audio at very poor electrical efficiency, and there may be a few people who are looking for just that in a buzzy flea-powered guitar amp. For everyone else, just say "No!" to self-split!
The trouble is, the self-split output stage is a "STP", a short-tailed pair, not a long-tailed pair. The "long" in LTP refers to the large resistance value of the "tail" resistor. But a self-split circuit cannot use a long tail, as a large resistor here will not allow enough current to flow to deliver substantial power to a loudspeaker.
Unfortunately, the short tail (small cathode resistor) ruins all the good characteristics of an LTP phase splitter. It simultaneously ruins all the good characteristics of a proper push-pull output stage.
The self-split circuit is a cake that is not only bitter, it also has a dead rat on top of it.
-Gnobuddy
Last edited:
The attached seems to be a HIGH gain single-bottle amplifier with some pretense of power. By thumb-count the (bypassed) gain is 5mV input for full half-Watt output, which is generally more gain than you can stand. Hence I drew it unbypassed and expect 20mV sensitivity, which aint bad. However this is a HEAP of gain from first grid to power plate, and maybe can't be stable.
http://www.mif.pg.gda.pl/homepages/frank/sheets/123/6/6AR11.pdf
The 'power' stage runs essentially at book values. (B+ is 150V) The 1st stage runs much lower current for more voltage gain. Yes, Vg2 is very low. (G3s strap to cathodes.)
http://www.mif.pg.gda.pl/homepages/frank/sheets/123/6/6AR11.pdf
The 'power' stage runs essentially at book values. (B+ is 150V) The 1st stage runs much lower current for more voltage gain. Yes, Vg2 is very low. (G3s strap to cathodes.)
Attachments
The 6AR11 twin pentode looks fun, and they are not too expensive. But they seem to be signal devices not power devices. Nice idea though.
I think the single vale guitar amp requires you to cheat a bit. The FET input stage and ECL82 shows promise. I can run the FET stage off the pentode cathode voltage providing I don't take too much current and decouple properly.
What happens next (style of tone stack) depends on what sort of second hand cabinet I can win on eBay. That dictates how many pot holes the metal has and how many controls the preamp offers. That affects the FET stage somewhat.
See. This is how I design gear. Backwards
I think the single vale guitar amp requires you to cheat a bit. The FET input stage and ECL82 shows promise. I can run the FET stage off the pentode cathode voltage providing I don't take too much current and decouple properly.
What happens next (style of tone stack) depends on what sort of second hand cabinet I can win on eBay. That dictates how many pot holes the metal has and how many controls the preamp offers. That affects the FET stage somewhat.
See. This is how I design gear. Backwards
About the single ended push pull thing I found this:
A Single-Ended Push-Pull Audio Amplifier - IEEE Journals & Magazine
It is kind of weird because it requires a B- and there is no center tap, but still very interesting.
[/IMG]
A Single-Ended Push-Pull Audio Amplifier - IEEE Journals & Magazine
It is kind of weird because it requires a B- and there is no center tap, but still very interesting.
Last edited:
One of my more depressing realizations is the thought that I have contributed virtually nothing significant and original to the human race. I haven't even so much as invented a new type of knot - if my mommy hadn't shown me how to tie my shoelaces when I was a few years old, and I had grown up alone on an island, I would probably still be tripping over my untied shoelaces and falling out of my shoes. (Of course I don't have the brains to have invented shoes, either!)
We ordinary humans get by because of those one-in-a-million geniuses who come up with all the good ideas for our entire species. It took an extraordinary genius to invent the wheel, but once that was done, the rest of us can just about manage to figure out which side of a wheel-barrow is up.
A related thought: with 7.7 billion humans currently on the planet ( World Population Clock: 7.7 Billion People (2019) - Worldometers ), I am only the tenth decimal place in this vast sea of humanity. So insignificantly small that it's hard to fathom. Perhaps it's no wonder I haven't set the planet on its ear with my amazing contributions.
-Gnobuddy
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.