10 watt Simple Power Amp

[danville]

Here is a simplified schematic of a small power amp that could supply 10-15 Watts. The idea would be to use it for tweeters, headphones etc.

The design is certainly not original. I have seen IC buffers that use the same idea. Chances are one of you makes a version of this amp.

R1 & R2 set the bias of the output stage. R3 & R4 are small
Assume a good op amp and appropriate current sources (probably the 2 transistor method).

My question is has anyone built this kind of amp and did it have reasonable performance? Somewhat better than a typical National power amp IC?

[danville]

I realized at lunch that the transistors indicated on the schematic are too small. I would probably use something in a TO-263 case or maybe a thru hole part.

[jaycee]

I briefly experimented with a similar topology in SPICE... using BC546/556 and BD139/140 for outputs. That was for a headphone amp. Haven't got round to trying it yet, but it should work. I'd probably still have a VBE multiplier to control bias.

[danville]

I thought about a bias variation as well, but not about a solution.

The bias voltage setting is fairly simple with the current circuit but will vary considerably over temperature. This will probably mean more allowable power dissipation than ideal.

I suppose I could add a Vbe multiplier between the Q3 and Q4 bases.

[Ipanema]

I have build sometime ago as a headphone amp. It can even drive my Tangband well. :-)

[danville]

I have build sometime ago as a headphone amp. It can even drive my Tangband well. :-)

Yours really is the classic op amp buffer version.

Did you like the results? Have you compared it to anything?[

[Ipanema]

I've only OPA2134 CMOY based amp for comparison, which I found lack of power. That's why I build this diamond buffer for added kick. Much better than CMOY. It has a slightly warm characteristic. The actual opamp that I use is LF353.

[jerluwoo]

I've built severaal of these simple amps with great success and good sound. I used the lm833n because i had them around. This circuit simmed a nice .00075% thd under multisim 8 pro and the sound reflects the low thd numbers ,although in a built circuit is much higher im sure, 100db sensitve speakers tell all.

sorry for the bad picture but you get the idea.

[CBS240]

It seems to me if Q1 & 2 have similar Vbe vs temperature variation as Q3 & 4, could you not just mount them together? Not saying it would work, but Q's 1 & 2 acting as the Vbe multiplier....theoretically.

[danville]

I haven't looked at heat dissipation issues yet. I have never been too satisfied with Vbe multipliers that are mounted on the heatsink since there is too much heat lag in a heatsink.

OTOH, if TO-92 transistors were used for the Q1 & Q2, they could be glued to Q3 & Q4 since they have leads. The mounting problem of another Vbe multiplier doesn't go away, so this may be a good solution.

I guess I should build one and see how much the bias varies. I like the simplicity of this amplifier.

I could reduce parts by using constant current diodes instead of the familiar transistor circuits. I'm not sure this is really necessary since the parts count is still quite managable.

Most of my designs tend to have lots of transistors, so this is a bit of a departure.