Tube sound from a transistor amp, it's possible

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I would like someone's input on what kind of amp this is. This amp I've had for a long time. Looks like a piece of junk, tiny power transformer, only 10 watts per channel, made by mayfair (the tiny tape recorder company). I work on amps for a living and have heard hundreds, and my shop is in a music store, so get to hear LIVE instruments on a daily basis and I'd like to think I have a good idea of what sound should sound like. But this amp has something that Is so interesting and cool, it sounds like a tube amp. I've demonstrated this to several people "mostly musicians" and they think the same, I've even tricked people by hiding this amp, a couple cables, a decoy tube amp head turned on "standby". They reach for the controls and "whoops" they don't do anything, because the little mayfair is wired up instead. They are surprised, or confused, but they all say "that's a darn good sounding amp", or they think the mayfair has tubes in it.

I use it for music, not guitar, and it just sounds so good, very nice in the midrange, sounds like you can pick out each instrument in a song on it. The tradeoff is it's not that good in the bass. I'm more of a repair guy, not a designer, I know there's people on here that could give me some insight on the circuit, it's very simple. Only four transistors per side, outputs look like big TO92's, and are pretty fast in the Mhz and are complementary. I think it's a type of complimentary - symmetry circuit. It's 41 years old and all original and it can crank, giving that's a small room and it's hooked up to a couple of efficient speakers. I've driven it hard, but it doesn't clip, it starts to compress the sound if pushed too hard, exactly like a tube amp would. What in the circuit makes it do that? What is the 200uf and 10ohm on the 2sa628 for? Or any other details would be appreciated.

Thanks,
Keith

An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.
 
Administrator
Joined 2007
Paid Member
The 220uf, the 10 ohm and the 3.3k are the feedback components. Is it 10 ohm or 100 ?

The amp is a simple single ended input design (which is a major part of its success sonically) together with a simple class b fixed bias output stage. Increasing the value of the electrolytics in the signal path would improve the low frequency end dramatically. 470 uf for a speaker coupling cap nowadays (modern speakers and sources) is to low by a factor of at least 5. Also 40 year old caps can't be up to snuff today.
 
Looks like a 1960s circuit, even if it was actually made in the 1970s. The relatively low amount of feedback will mean it clips less aggressively than modern designs, but below clipping it will have much higher distortion. Crossover distortion may be a problem with such a crude bias mechanism. It certainly won't sound like a decent tube amp (whch won't sound of anything).

I would regard it as a historical curiosity, as it appears to be neither a guitar amp or serious hi-fi.
 
Hi,

Seems very typical for the low power amplifiers built into tape recorders,
and the restricted output cap value probably relates to the small tape
recorder drivers fitted, no point sending bass they can't handle.

Would have to model it in the free TinaTi as I don't
intuitively understand the circuit and whats going on.

Feedback gain seems far too high, unless I've missed something.

The JLH and DoZ are the accepted "valve-like" transistor designs.

This one has much less power and needs a thorough analysis.

10W per channel is impossible, more likely around 3W.

rgds, sreten.
 
The choice of transistors in it are what baffles me. Either they used what they could get cheapest or they are leftovers, but there would have easily been better choices available.
2SA628 25V(vce) 100mA(ic) 100Mhz(ft) hfe 100-800 depends on suffix
2SC711 25V 50mA 75Mhz hfe 300 max
2SC1209 20V 500mA 75Mhz hfe 35 min
2SA695 20V 700mA 75Mhz hfe 50 min

DF96 these are surely Japanese watts since the data for these transistors are only available in Japanese as well.
 
Last edited:
My mistake, the diode is supposed to be two diodes in series. The 10 ohm resistor in the feedback circuit is right. The 47K on the capacitor is 47,000pf. The amp has alot of gain, has some ringing on square wave test, four bumps on the leading edge. This leads me to believe that the neg feedback can't be increased much. It does oddly manage to have almost no crossover distortion (can't even see it on the scope). 3db points are 80hz flat to 50khz, with usable bass down to 55hz (possible predistortion in preamp). Power supply is around 23v. There's no way it's only 3 watts, has to be at least 7 per side (it can shake the floor). It's not a serious amp power wise, it's just the only transistor amp I've ever heard that sounds like a single ended tube amp. When under 2 watts it just sounds so clear, it kind of baffles me how such cheap components, weird "or poor" design choices can come together and sound really good. I just posted it to see of the topology of the amp could be built into a more powerful amp. I just figure that if such poor components can sound good, then better parts and fine tuning of values might lead to really great sounding amp. There's something to be said for a design that has a gradual increase in distortion with power. But I do consider this amp to be hi-fi, it's just not going to rock the house. Most of the time I listen to music under 5 watts anyway, if I want loud I use a more typical transistor amp. It can go loud, then it's bass gets really sucky.


Keith
 
Last edited:
Presenting part of the amp is only half the cake ...

behind it there is a preamplifier that is made 70's style also .. that will mean plenty of gain and raw tone controls adding an unreal but dear other wise very dear sonic signature ...

In the repair world there is many of people that preserve and repair amplifiers like that just to have the specific sonic result .

Obviously this has nothing to do with today's circuits when it comes to speed , linearity ,and accuracy .

kind regards and happy new year
sakis
 
You could be right. That would make more sense. It would drop the gain down to a more sensible value. The extra feedback would help cope with all the crossover distortion.

Mind you , a gain of 330 is quite possible since it would match
the line output level of "hifi" sources of the time , generaly about
100-200mV , it was common even with 70s japanese amplifiers ,
although with gains generaly being close to 100.
 
I just figure that if such poor components can sound good, then better parts and fine tuning of values might lead to really great sounding amp.
I wouldn't use the words "tube sound" simply because it's undefined, but can accept that it sounds good for guitar, why not?
And "poor components" might be part of the "magic".
Dear Repairguy, I'll ask you a personal favor: please drive it to *just* clipping, and then raise the signal 10X (20 dB)
Post both scope screenshots, I think it will probably show a quite unsymmetrical waveform, or one with a quite unsymmetrical duty cycle.
That would explain a lot, at least in the Guitar world.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.