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

[frequencycentral]

I've been looking for another submini amp build, I've breadboarded a few of my own ideas, but nothing satisfied. Today I breadboarded a submini version of Doug H's Firefly using a 6112 and a 7327. I really like it. Very crunchy at maximum volume, and pretty clean with the volume rolled back. The 7327 tube I used is a rare beast indeed, it's a mil-spec replacement for 6111/6021 and capable of handling up to 300 volts on the plates. A 6021WA could be used as a suitable replacement, the WA suffix having higher volatge rating than the standard 6021. At a push I guess a standard 6111/6021 could be used in place of th 7327, though it would be operating outside of it's spec. The subminis are tough little @#$%ers though! If they can be shot out of missiles at 80,000 feet they can handle a few more volts!

I'm using the nixie SMPS to power the amp, which has been posted elsewhere a few times and is based on this design: http://www.ledsales.com.au/kits/nixie_supply.pdf I'm running it with a better inductor than the one listed in the diagram, kindly handmade for me by a friend. My B+ is currently 185 volts.

I'll be building this up into a tiny enclosure over the next week or two, at the moment it's nestling happily on my breadboard! Many, many thanks to Doug H for allowing me to re-draw and post his great design.

[SISKO]

Nice project to work on!
Why did you omited the cascode stage?
Any spundclips?

[John Lyons]

Nice work!
I'n my opinion the cascode is the magic of the firefly (V4).
With the cascode on there is a much better chime to the sound.
More gain for sure but you can turn the gain down and get a nice
slight overdrive that sounds great. For pedal use the non cascode is
better though...

John

[frequencycentral]

Thanks, but I just re-drew Doug's design, so all props to him.

I missed out the cascode because a third tube would have put too much load on my SMPS, and also have added hugely to the current draw, what with the extra heater and all, probably I would have needed to use a wallwart with double the ma to implement it. I'm running it from a 12 volt 500ma regulated wallwart, but I believe the wallwart puts out more ma than it says on the box. Also, Doug stated in the recent thread re. cascode that he thought it's more trouble than its worth. So this little submini version is 'stripped down', if I need boost I have a few pedals available.

I'll do some soundclips when it's built, as it's on two breadboards (one for the SMPS, one for the amp) with the 125A loose, so not easy to move to where I'd need it to be to record it.

[tranceracer]

Totally awesome Rick!  Looking fwd to the sound clips!!   

[frequencycentral]

Totally awesome Rick!  Looking fwd to the sound clips!!   

That should read "Totally awesome Doug!" All I did was to breadboard it with different tubes. 

[SISKO]

I'll do some soundclips when it's built, as it's on two breadboards (one for the SMPS, one for the amp) with the 125A loose, so not easy to move to where I'd need it to be to record it.

Does the SMPS work fine on the breadboard? I dont remember what the switching freq was, but maybe the intercapacitance of the breadboard could bring some troubles

[punkin]

Can't wait to hear how this sounds 

[kurtlives]

I dont quite get the output section of this design. Is this class AB or A?

There is clearly no phase inverter or splitting of the signal prior to the output section. So how does signal get to both output tubes? The second triode's grid is grounded, hmm.

Ideas to increase out...thinking out loud here....
Use pentodes
Use pairs of pentodes, a quad.
Add a bypass cap for that 440R resistor.

[snap]

selfinverting output: signal for lower tube comes in there through cathode and gets amplified non-inverted, while the upper tube inverts.

[frequencycentral]

Yeah, self split is very neat and efficient isn't it? There's a lot more power here than if the second triode were a phase splitter. I think it would be interesting to try another self split section in parallel with the existing one. More power output at half the impedance? I'll try a bypass cap on that 440R and see what happens. You're right, it would be interesting to try self split with pentodes.

@ snap: you're very knowledgable - why don't we hear more from you? Your posts are always very succinct. 

[frequencycentral]

Add a bypass cap for that 440R resistor.

I tried a bypass cap, a 33uF was handy, so I tried that. Didn't discern any real difference. And at high volumes it made the whole thing hum - any idea why?

[kurtlives]

Add a bypass cap for that 440R resistor.

I tried a bypass cap, a 33uF was handy, so I tried that. Didn't discern any real difference. And at high volumes it made the whole thing hum - any idea why?

That's weird. Did you notice a shift in frequency when you installed it?

[RedHouse]

You should have named it the uFly.
(micro-Fly)

[Johan]

Add a bypass cap for that 440R resistor.

I tried a bypass cap, a 33uF was handy, so I tried that. Didn't discern any real difference. And at high volumes it made the whole thing hum - any idea why?

That's weird. Did you notice a shift in frequency when you installed it?

if you bypass that resistor, you defeat the selfsplit. you might get more gain from the tube with the grid input, but the lower one, the one with the cathode input ( and grounded grid) is basicly silenced..
and besides, ..push/pull amps sound so much better than single ended amps to my ears...
j

[Ben N]

I missed out the cascode because a third tube would have put too much load on my SMPS, and also have added hugely to the current draw, what with the extra heater and all, probably I would have needed to use a wallwart with double the ma to implement it.

Use a transistor (like the CCS pentaboost, etc.).

[puretube]

Add a bypass cap for that 440R resistor.

I tried a bypass cap, a 33uF was handy, so I tried that. Didn't discern any real difference. And at high volumes it made the whole thing hum - any idea why?

That's weird. Did you notice a shift in frequency when you installed it?

if you bypass that resistor, you defeat the selfsplit. you might get more gain from the tube with the grid input, but the lower one, the one with the cathode input ( and grounded grid) is basicly silenced...
j

+1!

BTW.: the "Laptube"-amp does the auto-phase-inverting with its 2 EL84s
while the 2 ECC83s do a crossover-duty of what the E-H "Hot Tubes" & "Tube Zipper" do elsewhere...
and: the power supply is 2 suited back-to-back 230V:12V transformers... 

the auto-thingy-trick results in a little less output-power...

[kurtlives]

I wonder if you will show us a schematic...

[frequencycentral]

BTW.: the "Laptube"-amp does the auto-phase-inverting with its 2 EL84s

I read one of DougH's posts over at AX84 yesterday, which i think refers to the same thing(y):

"Maybe there's a way to make self-split more efficient. Personally, I like the idea of screen self-split using pentodes."

From this thread: http://www.ax84.com/bbs/index.php?id=333694

A bit of googling turns up these:

http://ax84.com/media/ax84_m141.pdf

http://geckoamps.com/magnatone109/

......which I haven't got time to study right now, but I think demonstrate screen self-split using pentodes.

[puretube]

I wonder if you will show us a schematic...

Why? haven`t you caught it from this so-called gentleman`s site yet ?:
http://analogguru.an.ohost.de/temp/EH_Tubezipper_factory.pdf