I want to make a micropower tube amp... calling tube gurus!

[earthtonesaudio]

I'd like to get the most "big amp" features (power tube distortion, inductor saturation, "sag," etc.) that are feasible to put in a "little amp" package.

The thing is my <0.5W Noisy Cricket is too loud for home use.  I have some 12A*7 tubes and an odd assortment of inductors, HV caps, etc.  So I'd like to make a tube amp with maybe 1-50mW output power.  No starved-plate design, just like a normal big amp but with a tiny fraction of the normal power.


I guess you could think of this as a pedal, just with really low output impedance (4-8 ohms).

[lunchie]

I would check ax84 web site - High Octane amp with low power option - 6n1p as power tube. You can connect just 1 triode instead of both for even less power output. But even than it will give about 1/2W.

Some time ago I built 30V sopht amp, which outputs about 1/8W and has nice bluesy crunch. But it's low voltage amp with not much gain. It sounds nice.

[CynicalMan]

I'm not an expert on tube designs, but can't you get a very low power amp by using a 12A*7 as the power tube in a single-ended triode design?

[raulgrell]

The murder one is an amazing amp designed by a fellow forum member (sorry man, can't remember your name!) I've made the PCB, but I can't find the preamp tubes where I live, but the sound clips are great...

[puretube]

Rick Holt, aka: "frequencycentral"

[markeebee]

Alex - I am on the same quest, and I'm sure many others are too.  Frankly, you've got far more chance of finding something than I have because I really have no idea what I'm doing, but here's some things I know.....

Murder One is a marvellous piece of work, like everything Rick does, but it's not strictly HV.  It's lovely, though.

Doug H's Firefly is a similarly marvellous piece of work, it does the 'big amp tone' nicely, and it uses 12A*7 tubes at higher voltages.  Only problem is that the output is something slightly more than 1W, so it'll be louder than a noisy cricket and it needs to be cranked to get the best out of it.

This looks like a decent candidate, less than 100mW using a EF86 push-pull at 177V:

http://www.pentodepress.com/receiving/EF86-power-amp/EF86-power-amp.php

It's near the top of my list of things-to-do-after-Christmas, along with checking out power scaling on my Vox AC4.  But surely Christmas is still months away??

My bestest quietest sounds have come from sticking an L-pad attenuator on an already low-output amp, because I haven't found a better way to do it although I'm sure it exists.  HOWEVER....I'm not sure if the pad was killing the tone, or if the speaker wasn't being driven hard enough, or whether it was just all in my head, but it didn't sound, um, right.  It was kinda artificial, like a bad ampsim.  I think you need that whoosh of air pressure, and that organic coupling with the environment (or something) to get an authentic valve amp sound.

If it's not loud, it'll never sound loud.  I think.

[raulgrell]

Rick Holt, aka: "frequencycentral"

Thank you sir! And Apologies to rick 

[petemoore]

 Finding something like this, '1/2 there' is tricky. Small-ish, ''Medium small'' to large are more common old tube amp sizes.
  Makes it trickier to dig out information on such items. Good links already though, more is known now than there was ''then''.
  That's nice in one way but in another way you have to find, somewhere:
  Coils....speakers and transformers.
  The rest of the stuff fits other size amps mostly.
  Transformer, I too wouldn't mind to have another tranny loose around here, Hammond apparently sells on with taps on both sides, making it a versatile 'hooka whatcha got up through here' unit.
  I'd start with 12au7, lower gain/higher current tube, +preferable I think to 12ax7 for speaker driving.
  They sure do start getting loud-ish quickly, I cobbled a 6sn7 [+2x 12a_7's] amp, starts crankin' and it's only supposed to be a fraction of X many watts [.7w?].
  Overkill on the output tranny didn't hurt at all, just a little large 'n heavy.

[earthtonesaudio]

I'm always "subscribed" to the frequencycentral channel.    But I got the impression his amps were at least as loud as a Ruby/Cricket type of miniamp, if not louder.  For me that is waaaay too much power.

Alex - I am on the same quest, and I'm sure many others are too.  Frankly, you've got far more chance of finding something than I have because I really have no idea what I'm doing, but here's some things I know.....

Murder One is a marvellous piece of work, like everything Rick does, but it's not strictly HV.  It's lovely, though.

Doug H's Firefly is a similarly marvellous piece of work, it does the 'big amp tone' nicely, and it uses 12A*7 tubes at higher voltages.  Only problem is that the output is something slightly more than 1W, so it'll be louder than a noisy cricket and it needs to be cranked to get the best out of it.

This looks like a decent candidate, less than 100mW using a EF86 push-pull at 177V:

http://www.pentodepress.com/receiving/EF86-power-amp/EF86-power-amp.php

It's near the top of my list of things-to-do-after-Christmas, along with checking out power scaling on my Vox AC4.  But surely Christmas is still months away??

My bestest quietest sounds have come from sticking an L-pad attenuator on an already low-output amp, because I haven't found a better way to do it although I'm sure it exists.  HOWEVER....I'm not sure if the pad was killing the tone, or if the speaker wasn't being driven hard enough, or whether it was just all in my head, but it didn't sound, um, right.  It was kinda artificial, like a bad ampsim.  I think you need that whoosh of air pressure, and that organic coupling with the environment (or something) to get an authentic valve amp sound.

If it's not loud, it'll never sound loud.  I think.

You may be right... but I still want to try.    At the very least I want to get the HV/OT saturation/PT distortion parts correct, so that if it still doesn't sound "right" I won't always be wondering "if only I had used high voltage..."

I'm not an expert on tube designs, but can't you get a very low power amp by using a 12A*7 as the power tube in a single-ended triode design?

I am of the exact same opinion!  But unfortunately I'm not an expert either.   

[markeebee]

Where have all the gurus gone   

On the topic of output iron, check this:

http://uk.rs-online.com/web/search/searchBrowseAction.html?method=getProduct&R=2106475

Good for up to 3.5W, multi-tapped on both sides.  It's not mega-quality but for micro-amps it's the kershizzle and, best of all, only £7.  Seven British Pounds!  At the moment that's about two dollars or half a euro 

[sean k]

Yeah reading the EF86 thing it brought out what I'd been thinking and that is that smaller outputs require much higher plate to plate impedances so 40k to 60k / 8 ohm transistions might be slightly hander to come by. But we're forgetting that voltage ratios are what count and it's not hard to find 10k/600 ohm transformers that are used on the inputs and outputs for tube preamps and once you do the math it might be that some usable iron is actually available for miniscule output tube amps. Another option might be inefficient speakers to soak up some power.

Hey Markeebee, thats the exact transformer I used to do the duties with some 6AK6 tubes in push/pull on the first amp I designed completely from scratch.

[PRR]

> I'd like to make a tube amp with maybe 1-50mW output power.  No starved-plate design

Output stage: use a REAL power-amp tube, but scale the voltage. Power will run roughly as square of supply voltage. So if 6V6 at 300V makes 5 Watts, the same 6V6 at 30V will make 5W * ((30/300V)^2) = 0.05W. The plate current will be -roughly- 4mA. This is using a linear approximation which is NOT accurate for a 10:1 scale-down. But if the reality is a 20:1 scale-down, it will still be useful. In fact you don't for-sure know that 50mW is your target. And 30V 4mA is a trivial supply to breadboard, and easily modded for 50V or 20V, whatever sinks your particular boat.

Input stage: if you don't shrink your guitar, you must stay near "normal design". You need to take 1V input peaks cleanly. If you stick with 12AX7, this means the usual Fender 100K+1.5K values working near 300V and 1mA. The 30V output stage will not need much drive, but not so much less than you can omit a stage. So 2mA needed.

Since your big supply is 30V 4mA, and your little supply is 300V 2mA, you have an odd situation. It may be easiest to just do 300V 6mA, and use a 75K 2W dropping resistor to dribble into your "power stage".

Alternatively, use low-gain preamp tubes under a smaller supply. 12AU7 with no bypass has gain like 8, can take pickup level while running on 40V supply. Stay with 100K plate load, fiddle cathode resistor to set plate at 1/2 to 3/4 of supply voltage. Guessing the sensitivity of 6V6 at 30V, you may want preamp gain of 90; two stage of 12AU7 gives 64 which may be close enough. You can bypass the second cathode to get more.

Use the Fender or Gibson 12AX7-6V6 "champ amps" for basic guidance, except much lower voltage.

At ~~30V it may be amusing to explore the table-radio tubes. 25L6/35L6 is a heck of a tube, cheap, and you can run the heater off your "B+".

Tube guitar power-amps are a lot about transformer nonlinearity. You really prefer something far smaller than a 3 Watt part. Haunt 5-toobe radio sites and suppliers to find the cheepest smallest OT. 5-tube radios often claimed 1W but usually broke-up long before that. Still the cheeziest iron may be "too clean" for 0.05W flavor. That's also the problem with using "small" tubes to get less power: quite different interaction with the iron. Yes, there are some very fine little amps using teeny-tubes and/or relatively generous (for their power) OTs. But getting anything like "big amp sound" at very low acoustic level is going to be a long extended experiment.

Big (50W amps) have bass. Small (5W) amps often shave bass to avoid embarassing their small speakers. Down at the 0.05W zone, you might have to reverse course. Put in all the bass you can hear at 77dB SPL. Maybe an actual bass-boost.

[sean k]

Actually I was completely off the mark with the idea about 10k/600 transformers... the voltage ratio is going the otherway and is 4/1 but then why not use a 10k/8 ohm transformer into a 10k/600 with the resultant impedance 80k/8 ohms!
Therefore an 8ohm speaker on a 10k/16 0hm into a 10k/600 ohm would result in 40k/8 ohms and all the losses might do alot towards getting the sound you want! Maybe those weird old metal can preamp pentodes would be fitting on such a fine time machine!

But after all that isn't a four ohm speaker on the sixteen ohm output the same as raising the primary impedance by four?

or is it the other way around? Good golly, I'm certaily showing my ignorance... but then again I'm pretty sure tubes are alot more able to still function and still do the things we want with a good dose of ignorance and pure hope thrown in.

[earthtonesaudio]

Alright now I'm gonna show a bit of my own ignorance.

Tube guitar power-amps are a lot about transformer nonlinearity.

This is the sort of thing I want to know more about.  I think you can put a magnetic offset in a transformer by running DC through it (no problem there if used as a plate load), but I don't know how this affects the sound in a "guitar amp way."  I imagine you would be swinging more toward one end of the B-H curve during normal operation, so saturation would only occur, or occur most, on negative (?) signal excursions, causing a bit of waveform asymmetry. 

For lower current operation, it seems that to get the same sort of magnetic saturation you'd want to use a different type of core.  I don't know what that would scale to, but I do know saturable-core reactors are fairly common in small packages (in computer power supplies for example) and I'm certainly not above winding my own toroid.


PRR, if you say lowering the voltage is all that's needed to reduce output power, I believe you... but how much would this change the sound (assuming that somehow all else was held constant)?  From what I've read from the low-voltage tube distortion pedal (Valvecaster, etc.) topics, low voltages give an "exaggerated" version of the tube's sound.  I don't really even know what this means, but it seems like lower voltages would change the response somehow.

[PRR]

What happens if you squeeze a 100 Watt amp's output through a "10W" transformer?

It depends. On many things. But generally: the bass comes out bent. In fact a hi-fi 10W OT makes a fine 20W guitar OT: the 50Hz THD is high, but guitar generally stops at 82Hz, and any stray distortion may be "good", add color to bass notes.

As an old hi-fi-guy, I think most guitar OTs are "undersize". For good reason: cost, weight, and because the bigger iron sounds "too mellow" on stage.

> want to use a different type of core

Nah. We use the same iron from 3W to 3,000W. The main reason it is hard to find iron below 3W is that it doesn't get much cheaper to go smaller; below 1W the cost tends to rise because the wire size gets too fragile for high-speed winders.

Yes, it does take some understanding to use tubes far beyond where the designer intended.

And it may be that these modern "TUBE!" products are deliberately MIS-biased "so you can hear the tube". The number of buyers who actually know what good tube sound is is not enough to support a mass-market product. And a good tube amp is not THAT much different from a good transistor amp. So the world is full of nasty bottle-amps, which may not be "good tube", but are different, and there is a market for "different".

[Renegadrian]

Here's the solution!




really, dunno how much loud can that be...So digital multifx and headphones are my late night companions...

[bancika]

yeah, I moved to headphone realm, don't have conditions to play loud in my apartment so I had to change something.
Nice tube preamp, palmer speaker sim and boss microBr into headphones are pretty good alternative.
I'm looking into building a headphone tube amp....

[amptramp]

If you are looking for output transformers, the most plentiful would be the 70.7 volt line - 8 ohm speaker transformers that are used for discos, department stores etc. and they are cheap.  Since they were designed to work from the AC-coupled output of an amplifier, they offer some good asymmetry that will produce even harmonics.  The impedance is usually selectable from the power taps with 1 watt being 5 Kohms.  Usually there are a number of taps so that 2 watts = 2500 ohms etc.  The impedance is V * V / P.  For the common 70.7 volt series, Z = 5000/P.  A 12AU7 should work well from the 2500 ohm / 2 watt tap.

Another decent tube would be the 12K5 which was designed for 12 volts on both the heater and plate.  This simplifies the design considerably and produces about 40 mW.  There are some other threads about that on this forum.

[Cliff Schecht]

I like the 6AK6 power tube in the microamp I posted in the submission thread. Running this tube with pentode operation gives a bright and punchy output that distorts nicely. Operating the tube as a triode would drop the power even more and would probably tame some of the brightness of the tube. I haven't tried a 12A*7 tube as a push-pull pair so I have no comment on doing this, aside from that it seems popular in designs like the Firefly.

Running any tube at a lower voltage will allow you to drop the power levels reasonably. But you should be careful about how small you really go, a 50mW amp won't sound much louder than an electric guitar with NO amp. You should probably be aiming for something more like 1/3rd to a 1/2 Watt of output power, anything else would be pretty useless IMO. The amount of power you get out also depends on the speaker size you're running into. An 8 Ohm 1W speaker isn't going to be anywhere near as loud as a 4x12 8 Ohm half stack.

Is there a certain sound you are going for with this amp? Are you looking for a clean amp or something that can get it's own distortion?

On the topic of output iron, DC bias does have an effect on the overall frequency response and saturation level of the output transformer. More DC bias will cause the OT to saturate quicker and give you that nice sounding tube compression and distortion quicker. I believe that DC bias makes it easier to achieve asymmetrical distortion which is desirable in some amounts in a guitar amp.  Also under higher saturation levels, the transformer has a frequency limiting effect that depending on the application can also be desirable.

I think that choosing a plate resistance as high as 40-80k is not going to give very pleasing results. In a single ended amplifier, you are going to get a LOT of odd harmonic distortion which is going to give you a very high gain, distorted sound. But this may not sound so nice, the amplifier may sound weak and, well, underbiased. The low end will be weak and you will always have a high end buzz that will sound more like a bad fuzz than a good tube amp. If you want that real tube sound, I recommend sticking to something more like 5k-15k.

[earthtonesaudio]

I like the 6AK6 power tube in the microamp I posted in the submission thread. Running this tube with pentode operation gives a bright and punchy output that distorts nicely. Operating the tube as a triode would drop the power even more and would probably tame some of the brightness of the tube. I haven't tried a 12A*7 tube as a push-pull pair so I have no comment on doing this, aside from that it seems popular in designs like the Firefly.

Thanks for the suggestion.  Would you be willing to suggest some mods to your switchmode power supply for variable voltage operation?

Running any tube at a lower voltage will allow you to drop the power levels reasonably. But you should be careful about how small you really go, a 50mW amp won't sound much louder than an electric guitar with NO amp. You should probably be aiming for something more like 1/3rd to a 1/2 Watt of output power, anything else would be pretty useless IMO. The amount of power you get out also depends on the speaker size you're running into. An 8 Ohm 1W speaker isn't going to be anywhere near as loud as a 4x12 8 Ohm half stack.

Is there a certain sound you are going for with this amp? Are you looking for a clean amp or something that can get it's own distortion?

I guess the sound I'm going for is the "non-master volume tube amp cranked up to 11" but at about the volume level of a typical acoustic guitar.  The reason I was thinking 50mW is because when I crank my LM386 (at less than 0.5W) amp, it's WAY too loud to play at home.  Based on the "rules of thumb" for power vs. loudness, I figure a factor of 10 less power than that, with the same speakers, would be about half as loud, which might be usable.  I know simply lowering the voltage will drop the power, but will also change the way the tube distorts.  I don't want it to sound like a starved-plate "tube distortion" circuit.  I want it to be as close to a "real amp" sound as I can get, minus the volume.

On the topic of output iron, DC bias does have an effect on the overall frequency response and saturation level of the output transformer. More DC bias will cause the OT to saturate quicker and give you that nice sounding tube compression and distortion quicker. I believe that DC bias makes it easier to achieve asymmetrical distortion which is desirable in some amounts in a guitar amp.  Also under higher saturation levels, the transformer has a frequency limiting effect that depending on the application can also be desirable.

Another reason to keep the supply voltage higher... or add a third winding to the output transformer for adding DC offset.

I think that choosing a plate resistance as high as 40-80k is not going to give very pleasing results. In a single ended amplifier, you are going to get a LOT of odd harmonic distortion which is going to give you a very high gain, distorted sound. But this may not sound so nice, the amplifier may sound weak and, well, underbiased. The low end will be weak and you will always have a high end buzz that will sound more like a bad fuzz than a good tube amp. If you want that real tube sound, I recommend sticking to something more like 5k-15k.

Thanks!