Go Back   Home > Forums > Blogs > RJM Audio Blog

If I put my notes here, I might be able to find them again later!
Rate this Entry

A power amplifier.

Posted 19th January 2017 at 11:37 PM by rjm
Updated 22nd May 2017 at 02:26 AM by rjm

This is an unfinished design idea.


Another request.

The Sapphire4 and Unity (H) boards were based on the 47 Labs Treasure 0347 amplifier circuit, and the original was not so far away that I couldn't "de-modify" the Unity (H) board back to what is essentially a 0347 (but with neither the original transistors nor the original resistor values).

I changed just about all the resistor values slightly for one reason or another, and changed all the transistor types. I've run my own simulations to check my version basically works as it should. Output BD135(39),136(40) are shown on the schematic as being leftovers from my headphone design, they will work to play music but to meet high current output specifications something bigger in TO-220 or similar thermal package will be needed.

As I hope you can see from the circuit schematic, the design - for a power amplifier - is unconventional. There is no feedback loop around the output stage to reduce crossover distortion. What you get is what you hear. Instead, you have to rely on fairly well-matched transistors to get clean output once the circuit output moves into class B.

Of course the feedback can be rewired to the output for the conventional global feedback operation. Which is better is something of a polarizing issue. Trying to paper over crossover distortion with feedback is - problematical. It is better to minimize it before feedback, but after doing all you can (component matching, careful choice of bias current) is it then a good idea to put some feedback around the output stage to keep distortion low? The conventional answer is yes. My 0347 suggests maybe no.

R6 controls the output offset voltage. R15 adjusts the output bias current. (It should be set to about 30 mA.)

Q13,14 should be heatsinked (duh...) but it is sufficient to bolt them to the chassis if the panel thickness is reasonable (2 mm or more) and it is made of aluminum or similar.

Gain as shown is about 26 dB. Output power is about 30 W into 8 ohms (on paper, ignoring thermal constraints), though distortion at even moderate output powers (above 1 W) is already comparatively high.

Here's what else you need to know:

Runs off 100VA or higher power transformer, 2x15 VAC secondary, bridge rectifier, 2200 uF filtering or more on the power supply rails in addition to what is on the main boards.

The outputs go to a mute toggle switch and then a 2A fast fuse before connecting to the speaker terminals.

In the original, Q9 and Q10 are taken off board and glued to the face of Q13 and Q14 packages. This provides overtemp control, as the output transistors heat up the Vbe of Q9, 10 decreases - reducing the output bias current. It has been suggested to me - and I agree - that ThermalTrak transistors could and should be used for the output pair.

C2 was not present on the original. I include this compensation capacitor as an optional tuning option, though my own tests show it shouldn't normally be needed. R2 can be adjusted a little (x2 /2) to change the gain +/- 6dB.
Attached Thumbnails
Click image for larger version

Name:	pcb-unity47-10a-brd.png
Views:	270
Size:	16.4 KB
ID:	2235   Click image for larger version

Name:	Unity47-10a.png
Views:	222
Size:	185.2 KB
ID:	2236   Click image for larger version

Name:	pcb-unity47-10a1-sch.png
Views:	549
Size:	19.0 KB
ID:	2239  
Attached Files
File Type: zip unity47-10a.zip (97.3 KB, 94 views)
File Type: asc Unity 10a (H) 47 mod.asc (5.5 KB, 202 views)
Total Comments 2


  1. Old Comment
    Nice to see someone design amplifiers! But this one is gonna blow up under heavy load, as BD139 and BD140 output transistors coming in TO-139 package are specified with 1A continuous (and sometimes 2A maximum) current, while 15V*1.4/8R is yet over 2A. Also, matching output transistors does not reduce XO distortion -- only raising bias current of output stage or introducing feedback of output signal can. Happy listening and good luck!
    Posted 31st January 2017 at 07:54 PM by Grasso789 Grasso789 is offline
  2. Old Comment
    rjm's Avatar
    Hi, thanks for the comment.

    Yes, if you want to push more than 1-2 W peak through this, you'll need to upgrade the output transistors. See note above.

    On the second point, it depends on whether you want to define crossover distortion only as the sharp break in the waveform as one transistors "hands off" to the other, or also include the fact that the two haves of the waveform are reproduced to slightly different scale. For the latter, matched transistors might help. For the former, you are correct.
    Posted 1st February 2017 at 01:28 AM by rjm rjm is offline

New To Site? Need Help?
Copyright 1999-2017 diyAudio