Hello,
Basically: Do valve-based designs necessarily have to use large wound transformers to generate the voltages used there-in and also to match the output?
I ask because the use of such transformers has always put me off designing with valves - the transformers tend to large & heavy, hence expensive, difficult to design, and difficult to source.
So, are there any standard 'modern' approaches to valve design, i.e. using switched-mode techniques and the such like?
Cheers,
Andy
Basically: Do valve-based designs necessarily have to use large wound transformers to generate the voltages used there-in and also to match the output?
I ask because the use of such transformers has always put me off designing with valves - the transformers tend to large & heavy, hence expensive, difficult to design, and difficult to source.
So, are there any standard 'modern' approaches to valve design, i.e. using switched-mode techniques and the such like?
Cheers,
Andy
An OTL (output transformerless) Tube Amp with a SMPS wouldn't have any iron at all, so yes, it is possible. But that is kind of a challenge.
But the easiest (and autochthonous) way is to use conventional transformers. Big, heavy and maybe expensive, but simple passive components !
Each way to avoid transformers will lead to much more complicate active system.
The simple task of generating a HV supply usually needs just the Transformer, rectifiers and caps. A SMPS needs a controller, switches, active regulation, slope compensation, filtering etc...
The OPT is just an impedance matcher. If it's wound properly it can be regarded just by its ratio and you can ignore bandwith limitation and DC-R etc. Feedback is just an option in such a conventional amplifier.
Avoiding the OPT will make the whole amplifier much more complicated. Active global voltage feedback is a must to bring the output impedance down. But impedance isn't the whole story because the OPT also transforms current. But current isn't something which tubes are famous for. So you either need many of them in parallel or use some really big ones with all the associated drawbacks (more distortion (those tubes weren't designed for audio), gigantic heater current, etc.)
You get what I mean ?
'Modern approach to valve design' is kind of an oxymoron. Like Eco-friendly 400cci Hemi V8
'difficult to design' - No definitely not. Sheet of paper, pencil, curves - done. No feedback needed for basic operation, no four quadrant curves, no drive current (as long as you stay A1/AB1), no huge gate capacity etc...
'difficult to source' - Maybe some years ago. Thanks to the internet, you can get transformers from everywhere. Cheap ones from China or from your country like 1,2,3
Hope I cleared things up a bit
But the easiest (and autochthonous) way is to use conventional transformers. Big, heavy and maybe expensive, but simple passive components !
Each way to avoid transformers will lead to much more complicate active system.
The simple task of generating a HV supply usually needs just the Transformer, rectifiers and caps. A SMPS needs a controller, switches, active regulation, slope compensation, filtering etc...
The OPT is just an impedance matcher. If it's wound properly it can be regarded just by its ratio and you can ignore bandwith limitation and DC-R etc. Feedback is just an option in such a conventional amplifier.
Avoiding the OPT will make the whole amplifier much more complicated. Active global voltage feedback is a must to bring the output impedance down. But impedance isn't the whole story because the OPT also transforms current. But current isn't something which tubes are famous for. So you either need many of them in parallel or use some really big ones with all the associated drawbacks (more distortion (those tubes weren't designed for audio), gigantic heater current, etc.)
You get what I mean ?
'Modern approach to valve design' is kind of an oxymoron. Like Eco-friendly 400cci Hemi V8
'difficult to design' - No definitely not. Sheet of paper, pencil, curves - done. No feedback needed for basic operation, no four quadrant curves, no drive current (as long as you stay A1/AB1), no huge gate capacity etc...
'difficult to source' - Maybe some years ago. Thanks to the internet, you can get transformers from everywhere. Cheap ones from China or from your country like 1,2,3
Hope I cleared things up a bit
An OTL (output transformerless) Tube Amp with a SMPS wouldn't have any iron at all, so yes, it is possible. But that is kind of a challenge.
The simple task of generating a HV supply usually needs just the Transformer, rectifiers and caps. A SMPS needs a controller, switches, active regulation, slope compensation, filtering etc...
TRUE if you want a complex SMPS supply.
UNTRUE if you are happy with a simple flyback SMPS. Using the TL494N you only need a handful of components to achieve the HV.
Another way of achieving HV without a lot of iron is to produce a HV push-pull SMPS with post regulation. Not as efficient but easy to design and work with.
With modern speakers you will always need an OPT.
Someone may be able to come with an ES design that has the appropriately high input impedance needed or even a 4K LS.
Last edited:
My goal is to make a zero gain amplifier, with a valve at is heart such that the 'sound' of a valve can be added to the signal. (Long story
I wonder if such a thing already exists. Any ideas would be grrrrreeeat!
Consider building a tube based line stage (or buffer) and using a solid state power amp.
Do a search here for tube buffers.
Hollow state line stages/buffers typically require a power transformer but do not need output transformers. They can provide tube sound to the signal.
SY's heretical line stage comes to mind ( and there are many others)....
http://www.diyaudio.com/forums/tubes-valves/58757-heretical-unity-gain-line-stage.html
Last edited:
Yep, Chinese. One of those headphone amps which ppl seem to rave about it giving 'that' tube sound which is what I'm after really, plus the simple supplies.
Soooo, how to make it unity gain?
It won't be driving a load per se, no speakers or headphones. It will be placed in-line with other kit which will do driving / amping etc. Guess I wasn't too clear on that considering my original post.
I'm not averse to a little bit of gain btw.
I'll check those designs out now...
Thanks a-gain (pardon the pun, ho-ho),
Andy
Soooo, how to make it unity gain?
It won't be driving a load per se, no speakers or headphones. It will be placed in-line with other kit which will do driving / amping etc. Guess I wasn't too clear on that considering my original post.
I'm not averse to a little bit of gain btw.
I'll check those designs out now...
Thanks a-gain (pardon the pun, ho-ho),
Andy
Don't use that circuit.
If I were you, I'd grab a copy of Morgan Jones's "Valve Amplifiers" and study it in detail. Lots of ideas for unity gain circuits, the correct way to use CCS loading, and the fundamental principles of design that you ought to know. Now, if what you want is basically an effects box rather than a clean amplifier, you'd do better seeing how the guitar amp guys do it, and there, Merlin Blencowe's books are excellent.
OK, cool.
Hopefully there will be something as simple for me to try!
Valve design is obviously not my bag, baby. Too awkward to scratch up with all those high V's required etc. but a friend of mine has sent me a 'request'.
Thanks again for your help,
Andy.
Just for the record, & if you have a little time to explain, why "Don't use that circuit"?
Hopefully there will be something as simple for me to try!
Valve design is obviously not my bag, baby. Too awkward to scratch up with all those high V's required etc. but a friend of mine has sent me a 'request'.
Thanks again for your help,
Andy.
Just for the record, & if you have a little time to explain, why "Don't use that circuit"?
Do it as SY says or buy something like this: 6N3 (5670) Tube Buffer DIY Kit (Stereo) No Tube | eBay
Otherwise we will fill pages here and won't get very far.
Otherwise we will fill pages here and won't get very far.
It's a circuit designed to maximize gain from a given tube, the opposite of what you're looking for.
@SY - OK, gotcha! I know not my valves TBH.
@the_manta: That kit you've kindly pointed out to me above seems (basically, bar the JFETs for constant current) the same as the circuit I've reproduced, other than the fact it uses different valves, i.e. Chinese 6N3 (5670) instead of 12AU7's?
So I guess the circuit I had originally assumed as OK uses the 'wrong' valve? So I guess I should start by asking "which valve"?
Andy
@the_manta: That kit you've kindly pointed out to me above seems (basically, bar the JFETs for constant current) the same as the circuit I've reproduced, other than the fact it uses different valves, i.e. Chinese 6N3 (5670) instead of 12AU7's?
So I guess the circuit I had originally assumed as OK uses the 'wrong' valve? So I guess I should start by asking "which valve"?
Andy
No. Ask, "Which topology?" And you'll find the answers and (more importantly) understanding in the books I suggested.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.