SOT: A little, old-fashioned amp. Wow.

[brett]

Hi
For a while I've been interested in vintage transistor amps, and in particular whether they share any of the benefits of vintage valve amps.  At first, you might think that solid state and valves have little in common, but not so.  Vintage germanium amps used to have output transformers, and ran at low currents because early transistors couldn't stand much heat.

Vanessa passed on a couple of classic schematics (including a little 1W amp used by Queen), and I set about seeing if there was any magic in these old critters.  The first design I can up with was this, which sounds great, but only manages 1W clean:



In order from left to right, the transistors are part of a buffer, a voltage gain section, a phase splitter, and two for the output.  As you can see, it uses a transformer in the phase-splitter as well as for the impedance matching at the output.  I re-wound wall-wart transformers to make the transformers.  It took about 200 turns on each of the primary and secondary of the phase-splitter, and 160 on the primary and 60 on the secondary of the output transformer.  I've previously posted a photo essay on making these transformers (ttp://www.diystompboxes.com/smfforum/index.php?topic=53786.0 ).

The amp in this setup is only low to medium gain.  To make it distort substantially requires a few dB of boost from a pedal.  The distortion is always very pleasant, even when driven way over the top.  I'm not sure whether this is due to the transformers or the push-pull output section.  On the scope, a sine wave gets softly clipped with mild overdrive, and a little "overshoot" appears on the leading edge as the overdrive gets harder.  There's never any squareness, perhaps due to the limited frequency response of the output section (-3dB at about 13kHz).

I'm also designing a version with some modern treats, like a tone section, and a bit more power.  I haven't built the whole amp yet, but the new output section produces 4W clean output into 8 ohms.  It now draws about 0.7A when going flat out (I power it from a 12V 1.5A switchmode power supply).  I'm also considering fitting some resistance to the power supply line to induce "voltage sag" during high current draw as experienced with many valve rectifier amps.

Here's the schematic.  All  suggestions are welcome.  If all goes well, I'll post something about the results in about a week.
cheers

[bancika]

very interesting, you've got my attention

[MartyMart]

Looks great Brett .... now I need some TIP31's  !!
A soundclip would be superb ??
Question :
Could I somehow use LT44's which are 1K/20K small transformers, (Titan boost) or somehow
adapt the circuit as just a fuzz/preamp with these OT's ? ....

MM.

[John Lyons]

Looks nice! Plus it fills a hole between tube amd and pedals. Would be a good inbetween for people who want to build an amp but are not quite ready for high voltage DC.
If you could find a way to make a clip or two that would be very informative.

Thanks
John

[wondergibbon]

Vanessa passed on a couple of classic schematics (including a little 1W amp used by Queen)

Would it be possible to post that? I am interested because many of Brian May's AC30s were modified by John Deacon (as electrical engineer I guess he is qualified! ) I would be interested in seeing its parentage and how it was used.

Overall, yes, I agree that what you are doing is opening a very interesting path for us all. Thank you for your efforts.

[IGR]

Interesting amp with similar topology http://www.belza.cz/audio/zes100.htm. Unfortunately in czech. It is aged custom build, to be operate with car battery, 100 W sine output! Author describes sound character as very pleasant, musical sounding, probably because of saturated OT core.

[Doug_H]

Looks interesting. I wonder about the bias network of the buffer on the first one though. Those 1M resistors might be starving too much current from the base. What voltage are you getting on the emitter?

[DDD]

I venture to recommend to insert two "gain controllers" instead of the existing one: pre-gain and master-gain. It will give a possibility to arrange a required rate of pre-amp and power amp distortion.

[brett]

Thanks for all of the positive comments.

now I need some TIP31's

For the 1W version, you could use transistors with lower ratings.  I used BD139s and they were hot, but not excessive.  The orinal designs used things like those AC128s in the big metal prism and OC81s with "tag" heatsinks.

I wonder about the bias network of the buffer on the first one though.

OOPS!  Thanks for pointing that out.  Those should be 220k resistors.  From memory, the emitter voltage is only about 2 and a half volts, which is slightly low, but I figured it's ok for the input stage.  I don't know why I drew the schematic with 1M, coz I use that 220k buffer quite a bit.  (RA Penfold's books show 220k resistors and 250k pots in inputs quite a lot)

I am interested because many of Brian May's AC30s were modified by John Deacon

Indeed, the little amp is known as the "Deacy" amp, and the story is that it was scavenged from a rubbish bin or something.  The base circuit is something like this:


Although this amp uses negative feedback to lower distortion, I found that my amp was better without it (and I kinda don't like to complicate things).

I venture to recommend to insert two "gain controllers" instead of the existing one: pre-gain and master-gain.

A very good idea.  I guess I hadn't thought much about the available distortion from the gain stage.  It should get quite high if pushed hard, so the pre and post controls might give a nice range of flavours.

Thanks everyone.

[brett]

Hi
RE:

I wonder about the bias network of the buffer on the first one though. Those 1M resistors might be starving too much current from the base. What voltage are you getting on the emitter?

and my response that they should have been 220k resistors...

While 220k resistors would be better in many respects, it seems that I did a version with 1M resistors.  That was the version that I did many electronic and audio tests.  My notes (that was an unusually organised day) indicate the emitter voltage was a lowly 0.8V.  The amp gives the rated output at about 100mV p-p, so I probably didn't notice any clipping in the workshop.  But it seems likely that driving it with a booster would cause clipping in the input buffer. 

The input signal and buffer output are both in-phase with the output of the driver stage, so limiting the negative swing in the buffer also limits the negative swing of the driver signal.  Maybe that's a happy coincidence that improved the sound of the amp.  At some stage I'll try switching between 220k and 1M input stages and see whether it makes much of a difference.

cheers

[brett]

Hi
I have some sound clips of the amp.
Not great, but they'll give some idea of what it does.
Can someone host the clips for me?
They total about 1 and a half Mb I think. 
(I probably should have used a lower sampling rate)
thanks for any help

[MartyMart]

Hi
I have some sound clips of the amp.
Not great, but they'll give some idea of what it does.
Can someone host the clips for me?
They total about 1 and a half Mb I think. 
(I probably should have used a lower sampling rate)
thanks for any help

Brett, you can upload mp3's here now in a gallery, just PM aron for one if you dont have one
or email 'em to me and I'll post them up here .

Marty.

[MKB]

Anyone tried to build such a beast with germanium power transistors?  I've never heard of this being done in a guitar amp, although some early SS models might have been built with such devices.  And apparently Ge power transistors are still around, I recall seeing some on sale recently.  Wonder what that would sound like....

One good thing about building power amps with transistors and output transformers is that if you're so inclined, the output transformers would be very easy to build yourself, due to the low impedances involved (less windings required).  This might be a good thing as there are fewer choices available for transistor power amp OT's.  There were some posts in the last few weeks detailing winding OT's that would be a great help.

Also, some of the finest high end stereo power amps made today, by McIntosh in Binghamton NY, are still made using transistors driving output transformers.

[brett]

Hi
There are substantial problems with keeping the heat away from Germanium transistors.  This leads to problems with biasing and thermal runaway.  The Oldie that I built uses TIP31 transistors which are available for $1 and are tough as nails.  The 1W version doesn't need any heatsinks and the transistors don't seem to get too hot (ie you can touch them for a short while, so I'd guess the case is 70 to 80C, and the junction is about 100 to 110 C.  That's well below the 150 C limit.)

I also doubt whether the germanium transistors would have any tonal advantages, as they wouldn't be stretched wrt gain, and wouldn't be clipping the signal.

Also, I need to clarify the thing about the input resistors.     The 1M resistors only work if there's a buffer present.  I've realised that I was playing through a buffer the other day (one invisibly hard-wired into my wah).  When I bypassed the buffer, the top end was totally dismal.  Without a buffer, 220k resistors should work in place of the 1M resistors.

I have some sound clips, recorded with a Yamaha RGZ using a blend of the middle single coil and bridge humbucker.  Speakers were a pair of Celestion Heritage 30s.  They were 16 ohms in total, so the amp output is reduced from 1W to 1/2W, and the amp "broke up" at lower volumes than with an 8 ohm speaker.

Clean, with guitar volume on 5.
http://www.esnips.com/doc/bdfd29e5-b7b6-4717-af15-3866f8e61eee/Clean

Guitar at full volume.  Despite the clip name, I don't remember using a booster on this.
http://www.esnips.com/doc/1f745bf4-ee35-4488-9dd7-cfa948aeaad1/Simple-blues-boosted

Solo with slight boost (Stratoblaster).
http://www.esnips.com/doc/53524024-1a17-420a-8e2d-75b77103c8e4/Full-volume-solo

A bit of rowdy half-watt fun (full volume, no boost).  I added just a little echo to fill out the sound of this clip.
http://www.esnips.com/doc/5997f5fc-6c07-4cb1-bd45-233002f0a109/Bit-of-half-watt-fun

Please excuse the quality of the playing....I build so many pedals and amps that I hardly ever get to practice.

All feedback welcome.

[MartyMart]

Sounds quite cool Brett, though a little "Transistor -y" it's a pleasing tone
Somewhat like a bright "fuzz" at max gain clip
Clean has a nice warm tone
MM.

[Dragonfly]

This is a bit tough to read. I'll "clean it up" if I get the chance. Anyway, its a Kalamazoo Model 3, which is a cool little solid state amp...nice bright, compressed sound...instant twang   really neat amp...i love mine.

AC

[brett]

Hi

Sounds quite cool Brett, though a little "Transistor -y" it's a pleasing tone

I was feeling rather "high gain" that day, so I was giving it heaps.  With the tone rolled back from 6 to 5 and the volume from 10 to 9 it changes character quite a lot.  A bit like the distortion channel compared to the clean channel on my Marshall.
cheers

[slim_blues_boy]

bumping very very very old thread.

been playing with small transformers and germanium transistors amp for a while.

all the pictures has gone, does anyone still have some archives regarding this thread?

thanks