Ok, firstly you have to bear in mind I'm in the UK, there is no Edcore here to come to the rescue, I have no money for custom coils either, so this project idea is using cheap rubbish that is to hand! So the initial conditions are:
Small, simple, cheap and bombproof:
1) No silicon except perhaps the rectifier
2) SRPP (I like the sound, it's simple, output impedance is fixed(ish))
3) Double ended: No output capacitor needed
4) Uses toroid mains transformers as OPTs (size?)
5) Compact and cheap 6n6p double triode construction
6) No chokes
7) 6n6p as I'm sure it can do more than small signal
8) Ideally over 4W output, gain as per integrated amp.
So the idea comes from the use of SRPP 6n6p tubes for OTL headphone amps, and noticing that I don't often need a huge amount of power for speakers. Also a little inspiration comes from the oddwatt schematic, and from tubelab getting huge power from small cheap tubes.
So the idea is this:
To use one double triode tube (6n6p) on the bottom and one on the top, so filament voltages can be separate(lifted) for top and bottom. So this is a 2 6n6p tubes per channel - 4 tubes for stereo. Then use one triode in each tube to construct 2 SRPP stages, driven from a LTP phase splitter to bridge the output. I.e. this is a dual ended push-pull. A trimmer may be needed to null the DC output.
The only figures I have estimated is about 18mA per 6n6p, for the UK it's fairly simple with back-to-back transformers to get 240V RMS = 340V line, PSRR should be good for the output stages.
If the LPT phase splitter does not create enough voltage for full power another stage would be needed - I have not yet worked this out yet.
Also I'd want to match the SRPP impedance (x 2 as it is bridged) via the OPT to about 8 and 4 ohms if the OPT has a tap (most toroids have a tap - 12-0-12V etc) - which I guess would effectively be the OPT ratio needed.
So is this idea stupid, crazy or interesting?
Has anyone made anything like this and was it any good?
What sort of power would you expect from what line voltage?
Should it need 2 or 3 stages in it?
What do you think the (ideal) output impedance from the SRPPs to be (Pointers to ways for me to calculate this would be welcome!).
Thanks in advance for any suggestions - I haven't bought anything yet except I have a couple of spare 6n6p's to experiment with..
Small, simple, cheap and bombproof:
1) No silicon except perhaps the rectifier
2) SRPP (I like the sound, it's simple, output impedance is fixed(ish))
3) Double ended: No output capacitor needed
4) Uses toroid mains transformers as OPTs (size?)
5) Compact and cheap 6n6p double triode construction
6) No chokes
7) 6n6p as I'm sure it can do more than small signal
8) Ideally over 4W output, gain as per integrated amp.
So the idea comes from the use of SRPP 6n6p tubes for OTL headphone amps, and noticing that I don't often need a huge amount of power for speakers. Also a little inspiration comes from the oddwatt schematic, and from tubelab getting huge power from small cheap tubes.
So the idea is this:
To use one double triode tube (6n6p) on the bottom and one on the top, so filament voltages can be separate(lifted) for top and bottom. So this is a 2 6n6p tubes per channel - 4 tubes for stereo. Then use one triode in each tube to construct 2 SRPP stages, driven from a LTP phase splitter to bridge the output. I.e. this is a dual ended push-pull. A trimmer may be needed to null the DC output.
The only figures I have estimated is about 18mA per 6n6p, for the UK it's fairly simple with back-to-back transformers to get 240V RMS = 340V line, PSRR should be good for the output stages.
If the LPT phase splitter does not create enough voltage for full power another stage would be needed - I have not yet worked this out yet.
Also I'd want to match the SRPP impedance (x 2 as it is bridged) via the OPT to about 8 and 4 ohms if the OPT has a tap (most toroids have a tap - 12-0-12V etc) - which I guess would effectively be the OPT ratio needed.
So is this idea stupid, crazy or interesting?
Has anyone made anything like this and was it any good?
What sort of power would you expect from what line voltage?
Should it need 2 or 3 stages in it?
What do you think the (ideal) output impedance from the SRPPs to be (Pointers to ways for me to calculate this would be welcome!).
Thanks in advance for any suggestions - I haven't bought anything yet except I have a couple of spare 6n6p's to experiment with..
Yes, that!
Glad you endorse the 'workingness' of it, I could not find an example of it via google so I wondered what I was missing 🙂
Of course since I find this: http://www.diyaudio.com/forums/tubes-valves/142236-6n6p-transformer-output-impedance.html
Which gives a 3.5W output from a pushpull 6n6p, from that thread it uses the excellent 6n2p (that tube is sounding massively good on my GU50 SE) - and I see the aim is for an output load of about 15k - derived from Ri 6N6P = apr. 1.8 kohm. As it's bridged I suspect a 30k load may suit.
So driving an 8R speaker with full power transfer gives (I'm a bit new at this bit so may be completely wrong):
W = V1.V1/30k = V2.V2/8ohm,
so V1.V1/V2.V2 = 30000/8 = 3750
so the ratio = sqrt(3750) = 61:1
Which for a 240V toroid is a 240/61 = 4V secondary. Probably easier to buy a 6V one, but it seems in the realms of reality.
Perhaps a touch of feedback from the cathode of the top SRPP tube to the cathode of the driver tube would reduce the impedance to the transformer so it would have a bit more grip.
If I can swing 240V (which appears easy in bridged as it's 120V per side) I'd then get 6V out, which into 8ohm = 36/8 = 4.5W. If I swung the full 340V voltage the theoretical limit = 8.5W, so in the real world I may expect 6W out of this then - does that sound right?
So to get that swing from 500mV I'd need a total 'side' gain of (340/2)/0.5 = 340, or a gain of 18.5 on each stage of a two stager and 7 of a three stager. So perhaps if I used a 6n2p SRPP as a driver this may deliver a two stage amp with sufficient gain, using 3 tube envelopes per channel and very little wiring.
Sounding good so far if any of my assumptions are correct!
Not trying to rain on your parade, but using a double H (half) bridge output stage really does not make sense for tubes. Notice that the center of the output winding is at virtual ground. So the primary is having to handle two X the AC primary voltage (or 4X the impedance). Then the SRPP stages are requiring twice the B+ voltage of a normal P-P stage. Not to mention the need for floating filament supplies and twice as many driver stages. Trying to use a power xfmr for the output xfmr at such high Z will be near disastrous.
If you want to use a power xfmr for the OT, then the only hope is a Circlotron output stage, which uses 1/2 the usual primary turns.
If you want to use a power xfmr for the OT, then the only hope is a Circlotron output stage, which uses 1/2 the usual primary turns.
If you want to make an inexpensive amp with 6N6P,what about the inverted triode amp on Steve Bench's site substituting 6N6P for,as far as I remember,5687's. It looks like an absolutely fascinating project!
SA: All noted, effectively I'm adding a (double) tube (doubling the output impedance and swing) and a phase splitter to eliminate a capacitor. Perhaps that's a bit nuts...
Steve Bench has some interesting stuff on his site doesn't he?!
Maybe I need to learn how to make transformers...
Steve Bench has some interesting stuff on his site doesn't he?!
Maybe I need to learn how to make transformers...
- Status
- Not open for further replies.