Q17 - an audiophile approach to perfect sound

Isn’t this a Class B amplifier? I thought such machines were supposed to be comparatively efficient and cool-running.

HRDSTL have fun. :)

160VA per board for the single version (Q17-Mini) or 250VA per board for the P2/Turbo is in my opinion quite sufficient for normal use. 50VDC for the single, 60VDC for the P2/Turbo but work very well at 50VDC (x25) and is more secure If mains power supply has large variations.

Switching power supplies also work well, but I find that the low frequencies are too dry and fast.

By tightening well, you can put a lot of things in a "small box". ;)



  • IMG-5832.jpg
    544.6 KB · Views: 92
Last edited:
  • Like
Reactions: 1 user

I just discovered a problem with the positioning of the BD139 (Q12) on the Q17-Turbo.

If I put the BD139 on top (photo 1) everything works fine. If I put the BD139 below (so that BD139 is on the heatsink for better thermal management), I have an oscillation (see picture). The problem is present with or without signal (the scope picture is without signal).

I tried on my two boards I built and I got the same result. On top ok, below not ok.

I had never noticed the problem because until now I had never installed the transistor underneath during my various tests for practical reasons.

So beware of those who build the Q17-Turbo with my PCB. Don't put Q12 under the PCB.

For the moment, I have not found the reason but if anyone has an idea, I am interested. It would also be nice if someone else could try to see if they have the same problem (you need a scope).



  • picture1.jpeg
    186.1 KB · Views: 65
  • picture2.jpeg
    451.7 KB · Views: 65
  • oscillation.jpg
    66.6 KB · Views: 67
  • Like
Reactions: 1 user
Joined 2014
Paid Member
Is your heatsink connected to ground?

I used a fully packaged BD139 and have been playing the amp for the past couple weeks without any issues. It sounds quite good actually. Try replacing the BD139 with an insulated version and make sure your heatsink is grounded.
I have not measured with an O-scope though.
The BD139G is isolated from the heatsink with a plastic film. The aluminum plate on which the Q17 board is screwed is for the moment just placed on the workbench.

I will try lifting the BD139 slightly so that it is no longer in contact with the heatsink at all to see if that changes anything. I'm going to rummage through my boxes to see if I find an all plastic BD139.

But hey, the only big difference since the BD139 is isolated from the heatsink is the length of the legs.

I tested the transistors before installing them. The transistors are installed in position on the heatsink before being soldered. No mechanical stress.

I found two plastic BD135-16. It's a bit caviar for pigs in this position but it should work. Will try.

I will also connect the ground/earth with the BD139 to see.

It also annoys me not to find the cause, especially since I used the BD139 instead of the original 2SC2240 to have a good, easy-to-install contact on the heatsink.