Idea: Low power subminiature tube amp

[Koreth]

So I've had this idea bouncing around in my head for a while. I have a nice 100W Boogie. Probably too much power for the show's I'm playing presently, and definitely a too much power for bedroom or living room practice. Originally I was just going to save up my pennies and buy a Fender Champ or something. But the idea of building my own seemed like fun, and then I found about about subminiature tubes, which seem almost like the perfect solution, save for the fact that they're a bit of uncharted territory as opposed to the more common, tried and true circuits. But, NOS subminis can be had for less than new production 12A*7s and EL84s or 6V6s, so there's that temptation too.

Looking at datasheets and playing with load lines, I figure I can get 5W from a push-pull pair of 5902 pentodes with around a 6kΩ load with the plates and screens at 150V. The Hammond 125 universal output tranny seems like a close enough match, with a 5.6kΩ or 5.8kΩ primary at 4Ω or 8Ω. Those load lines drop a bit below the knee at 150V screen volts, and output power looks like it would be closer to 6W, but how hard is it to drop screen voltage a little bit in the power supply so as to reign in output power and keep the amp from eating screen resistors for lunch?

For a preamp I was thinking of copying one of the classic Fender designs, like the 5E3 Deluxe as A) it's a fairly simple design that can be implemented with only a couple twin triode tubes, and B) the Deluxe I played recently in a local used gear store certainly sounded pretty. Really, since the Deluxe is push pull pair of 6V6s, I guess I'd be building a Deluxe, but with subminis for less power.

Sourcing a tranny for the power supply seems trickier. Looking on Hammond's website, it seems that all their power supply trannies for tube applications are rated for the higher voltages typically seen with the bigger, regular sized bottles. Unless my math is wrong, I don't think I'd need anything more than 115 or 120V on the PT secondary to get the 150-160 I'd need to power the plates and screens of the submini tubes, figuring there'd be a some small losses in the power supply itself. PTs with higher voltage and amp ratings seem more common. It seems like a senseless waste of money to buy an overly huge PT appropriate for a 25 or 50W amp to underutilize it in a 5W circuit. There's good design with plenty of headroom for reliability then there's wasteful overkill. But if I was to have to spend the money on getting some custom iron wound, I might as well abandon the submini idea and just go with a regular Deluxe or a Champ/Princeton clone. The latter would probably cost less money to build anyway. Subminiature tubes is probably biting off more than I can chew for a first build anyways.

Anyway, does anyone have any thoughts or advice regarding this idea?

[Tubebass]

Why not use a pair of 6.3 volt heater transformers back to back? The first one needs to be beefy enough to run the heaters with some left over to run the step-up to 120 volts for the B+. Say a 2A tranny for the heaters with a 1A tranny for the high voltage.

[defaced]

There's a guy named FrequencyCentral that posts on the stomp box forums that makes submini amps. He's posted schematics, his SMPS power supplies, the whole works. Check out freestomboxes.org, the build your own clone forum, and diystompboxes.com.

[Koreth]

I found some of Frequency Central's threads on diystompboxes.com. Very nifty. The switching power supply he's using looks like a fairly easy to build kit. I wonder though, for tonal reasons, do I really want to be using a switching power supply? Switching power supplies offer much better regulation than linear power supplies, don't they? Isn't the comparatively poor regulation of a linear supply and the resultant sag and compression part of the magic tube sound?

Taking a second closer look at Hammond's offerings, the 262F6 looks like it would be a good match. 6.3V @ 1.5A is perfect to supply the heaters of two 5902s and a two 6021/6111/6112s. 120V becomes almost 170V when through a bridge rectifier, which should leave room to regulate down to the target 150V under load, right? Or did I just commit a math error mixing up peak and RMS values there?

[Koreth]

Hmm, this might be tricker than I thought. Taking a closer look at the plate curves, various load lines and then the screen curves, even when the load line passes through knee of the 0V curve for the control grid, the screens start drawing enough current to exceed their dissipation rating. It seems like the screens of the 5902 get thirsty and quickly and would need huge screen stopper resistors, lower screen volts, or a lighter load. I might have to shoot for more or less power than the target 5W to keep the screens from melting.

Or should I just disregard all that since the best overdrive tones are to be had when the output tubes are about to melt and catch fire?

[defaced]

I wonder though, for tonal reasons, do I really want to be using a switching power supply? Switching power supplies offer much better regulation than linear power supplies, don't they? Isn't the comparatively poor regulation of a linear supply and the resultant sag and compression part of the magic tube sound?

I would be more concerned with hash in the audio band than sag. You are correct, SMPS do typically have better regulation, but I have read that his supply does sag from open circuit to fully loaded. If you've seen the nixie 555 design he's using, that's probably going to sag alot more than a supply based on a real SMPS, it will also probably have lower efficiency too. As for sag and compression, it all depends on what you're after. I like very stiff power supplies, others like droopy supplies, it's all a matter of taste.

[gwtekman]

Use a hammond 269AX for power transformer. It has already been done in a couple of circuits I have seen using 5902's

[Guitarist]

Perhaps consider one of these, but it may be too many Volts: 160V 50VA Toroidal Tube PreAmp Transformer Guitar Amp - eBay (item 370333426089 end time Jul-09-10 21:37:21 PDT)
The Hammond apears tried and true.
Those subm. tubes are cool, I imagine some from this forum will be talking about these in their hearing aids at somepoint

[drfunkphd]

Sub Mini Transformers

I am looking into a subminiature design myself and I have seen the Hammond 269 tranny used. My only concern is that the secondary voltage is 240-260V IIRC and the plates of most subminiature tubes have a max rating of 165V.

I want to run an amp like this from ~110V standard wall voltage, but would like to see a tranny with 140-160V coming off the secondary. How about running a tranny that has 110 and 220V primarys and a ~300Vsecondary, but hooking it up to the 220V winding? Would that work? I've never done it but in theory you're changing the in-out ratio so you should see half the secondary voltage at 110V.

I'm not as well versed with my transformers as other parts of amp circuits, so hopefully a seasoned amp guru can provide some wisdom and guidance...

[MWJB]

What kind of output W are you looking for? I'm not sure why you are looking for such low B+ voltages. There are plenty of 200-0-200VAC to 260-0-260VAC PTs about, some with 5V rectifier taps. If you're not bothered by a tube rectifier check out the PT at Amp Maker for guitar amp kits and parts

[d95err]

Switching power supplies offer much better regulation than linear power supplies, don't they? Isn't the comparatively poor regulation of a linear supply and the resultant sag and compression part of the magic tube sound?

You can get sag simply by inserting resistance between the power supply and the first filter cap. Gets you any level of sag you want even with a completely stiff powersupply.

[Koreth]

Wow, wasn't expecting this thread to come back to life until sometime next year when I'll actually have the money to put towards the project.

Yes, the 269AX does have a 250VAC secondary, but it is center-tapped. Unless I'm misunderstanding rectifier circuits, if a full-wave rectifier instead of a bridge rectifier is used, and it uses the center tap, we'd get only roughly half the secondary voltage, instead of almost all of it.

[drfunkphd]

The 269 AX is 125-0-125, so with a full wave rectifier you should get 125*1.41 = 176V with no load. The max spec for the plates on the 5902 is 165V, so maybe a bit of resistance to drop that voltage down a touch, but maybe not. I guess if you really want to do the math, you could find a way to do the output tubes in fixed bias as opposed to cathode bias and run the tubes a little cooler.

My only concern, mostly due to my design plans, is the 100 mA current rating. It should be fine for a pair of 5902's in PP and 2 (or hopefully 3) 6112's. The heater current looks like more than enough on the 6.3V tap.

For more straight forward plans, the 269AX looks like plenty of transformer to get the job done. If you had any ambitions of building a sub-mini SLO or Mesa, you might need to run separate trannies for the heaters, B+, etc. or get something custom wound...but that would just be silly.