"Superfly" - submini tube version of Doug H's Firefly

[Brymus]

So tell us more about the DM160, are you suggesting using it as a visual thing (very cool - how would one incorporate it?) or do you mean that they could be used as an amp?

Well I was suggesting a Superfly variant using the DM160. The fact that they also glow is just a bonus!
They are directly heated, but that shouldn't make things very difficult. PP should be possible with a little lateral thinking. A paraphase arrangement would be possible, if nothing else. i.e., Hang a cap+potential divider off the anode of the output valve that is driven, and use the resulting signal to drive the second valve's grid.
Here's a very crude video of a DM160 frequency indicator I built last week for a forthcoming competition
http://s81.photobucket.com/albums/j207/merlinblencowe/?action=view&current=CIMG4628.flv&newest=1
After this I will probably dismantle it and try something Superfly-esque with the tubes.

OMG that is so freakin cool !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
I want an EQ made with those...uhm RICK ?

[merlinb]

I wanna build one too! I'm ordering a bunch of DM160 right now. Will you share your schematic? 

Yeah, I'll work something up and post it here. Might give you some ideas regarding biasing the little directly-heated buggers too.!

[merlinb]

Ok, here's the gist of my frequency indicator. I'm not saying an audio amp should necessarily use the same approach, but it could work.
I ended up using only 4 inverter stages for the PSU, which gave me about 50V from 12V input, but for audio I guess all six would be preferable (it's a hex inverter chip) especially if its to run off 9V, or use Rick's SMPS of course. This circuit runs at about 300kHz, so I guess it won't cause audible problems.


This shows how all the tubes are biased to about -2V. For audio, to prevent feedback between stages via the heaters, I think you would want to split each grid leak into 2 parts and decouple the junction of the two resistors.


In the frequency indicator I just had an opamp input stage which feeds eight low Q bandpass filters http://sound.westhost.com/project63.htm each tuned to frequencies from 50Hz to 10kHz. Each filter then drives a tube.

[merlinb]

And here's a doodle of what I was thinking a DM160 Superfly might look like. I've added variable bias for prototyping- I wouldnt expect to use it in the final version.

[frequencycentral]

The DM160 datasheet says each heater requires 1v / 30ma, so if you ran them parallel that'd be 1v / 120ma. I'm calculating a 91 ohm / 1.32 watt resistor (doesn't exist) would drop 12v to 1v. Therefore an 82 ohm 2 watt resistor in series with two 1n4001 (each drops 0.6v) should get very close.

[merlinb]

says each heater requires 1v / 30ma, so if you ran them parallel that'd be 1v / 120ma.

120mA, you're crazy man! All that wasted power!
I'm gonna figure out a way to keep the current low, so a 9V battery could be used (and last more than 5 minutes). The output tubes can at least be in series if self-split is used.

[frequencycentral]

says each heater requires 1v / 30ma, so if you ran them parallel that'd be 1v / 120ma.

120mA, you're crazy man! All that wasted power!
I'm gonna figure out a way to keep the current low, so a 9V battery could be used (and last more than 5 minutes). The output tubes can at least be in series if self-split is used.

Hahaha! 120ma is 30 less than a single 12AU7 @ 12v. 

[merlinb]

Hahaha! 120ma is 30 less than a single 12AU7 @ 12v. 

Pah, we're working on the nano scale now- can't afford that kind of frivolity! Anyway, how much cooler would it be to be battery operated?!

Ok, it can be done with two 1000uF caps. The smallest I've found on Rapid are 12 x 10mm, which I think is actually pretty good.
http://www.rapidonline.com/Electronic-Components/Capacitors/Radial-Electrolytic/85-deg-C-Radial-electrolytic-capacitors/62556

[frequencycentral]

OK, I'm seduced by battery powered pico tube amp. Am I right re the maximum voltage is 100v? If so a SMPS would be overkill as they require a fair amount of milliampres themselves, and would need taming to get down to 100v. So a charge pump would be a better option. I wonder what consumes less milliampres (and takes up the least space), your 40106 pump, a MAX pump or a 555 pump ?

[merlinb]

So a charge pump would be a better option. I wonder what consumes less milliampres (and takes up the least space),

Good question. I know mine is at least 70% efficient. It needs fewer diodes and caps, but the chip is bigger, so I don't know which is smaller.

Although, even operating every valve at 100V 400uA is only 160mW. Even with 50% efficiency that's only 36mA from the battery!
At 50V 200uA for every valve the battery current is less than 10mA worst case.
(Plus filament current of course)

I think we're looking at about 1mW audio output power- enough to hear by, and more than enough to drive headphones. How much power do you get from a superfly?
I might try a hybrid at some point and scrap the output transformer.

[frequencycentral]

Well I just ordered some DM160, so I'm ready to go soon. I've got a couple of amps to build first, I just got a commission for a Superfly with master volume, Vox tonestack and a tremolo section. And then my Obsession, when the tube cages finally arrive from HK.

I'm amused that the battery to run the pico amp will be the single biggest component. That and the OT. I'd imagine it's going to end up being the same size as a Superfly, just less room inside. And with the cool blue glow. 

[slacker]

That 40106 charge pump looks really interesting, never seen one like that before. Might have to give it a try

[merlinb]

I'm amused that the battery to run the pico amp will be the single biggest component. That and the OT. I'd imagine it's going to end up being the same size as a Superfly, just less room inside. And with the cool blue glow. 

Yes, in fact the battery is so big I'm wondering if it won't need to go in a bigger enclosure than the Superfly?
Maybe we can find a smaller OT, like a PCB audio matching transformer, and bend it to our needs.

[frequencycentral]

I'm convinced it will fit in the same size enclosure. Take a look at the Superfly gutshots. Have everything in the same place, with the four DM160 in a row sticking out of holes in the top of the box, the battery would be where the knobs are. The (9mm) volume knob on the top of the enclosure at the opposite end to the OT. I'd still use a 125A, as the tubes really need to poke their little heads out to get the best visual effect, so I'm thinking the same roll bars too. It just needs a very compact board layout, my speciality....

[merlinb]

Hmm, maybe your enclosures are deeper than mine (25mm). I can't get a PP3 to go in unless it's lying down on its back, rather than on its side, so it takes up a lot of room. Ill be impressed if you manage it!

EDIT: I'm an idiot- batteries do fit edgeways into my enclosures. What a dingbat.

[frequencycentral]

If you kind of squint at this one you'll get the idea: http://i210.photobucket.com/albums/bb292/frequencycentral/000_0009-13.jpg The enclosures I use are Eddystone 111 X 60 X 27mm, batteries do fit in them as shown. Anyway, if you put in the legwork on the circuit design I'll do the layout and drill you an enclosure.  

[merlinb]

If you kind of squint at this one you'll get the idea: http://i210.photobucket.com/albums/bb292/frequencycentral/000_0009-13.jpg The enclosures I use are Eddystone 111 X 60 X 27mm, batteries do fit in them as shown. Anyway, if you put in the legwork on the circuit design I'll do the layout and drill you an enclosure.  

You'll have to wait for any serious stuff on my part, cause the electronics competition is 3 weeks from now. After than I will dismantle it and do some experiments. The only difficulty I see is in making the self split work well. May have to rig up a gyrator in the filament chain to get enough AC impedance.

You gonna include a tone control?

[frequencycentral]

Good luck with the competition. The timing suits me, got oter stuff to build first.

You gonna include a tone control?

No, too much loss through a tone control if we're talking 100v or less.

[merlinb]

I suppose, although the power valves will be just as sensitive, rather than the lower-mu ones you use in the Superfly.

[frequencycentral]

Merlin, can you explain the heater circuit in your schematic on the previous page please, the two npn transistors. Also, do you think we should start a new thread for the 9v DM160 uFly?