F6 Illustrated Build Guide

Digikey lists this part number "IRFP048N" as "Obsolete". No longer manufactured... How long ago did you acquire them? The F6 parts kits have been out of stock for at least 3 months now. I'm waiting with bated breath for them to come back in. A consideration for substitutions might be to compare the "input capacitance" of the devices. If it is too high, it will require more current drive to overcome the input capacitance and may suffer some high frequency roll off internal to the circuit... The IRFP240 also comes up as Obsolete, no longer manufactured at digikey.
Story of my life, there is always some component to any project that is made of unobtainium...
See if this helps. I’m happy to get others if you need them.
I’m trying to get these from the center of the holes. The 4 sets of vertical mounting holes are 40mm apart. The 3 sets of horizontal holes are 80 mm apart. Each hole about 5mm from the edge of the board. Reading from left to right, the first hole for Q1 is around 99mm from the left edge of the board, 110.3mm to the third hole. Reading from right to left, the first hole for Q2 is 19mm, the third hole is 30mm from the right hand edge of the board.

one more thing, the Q1 and Q2 pads are about 1mm closer to the edge of the board than the board mounting holes.
Last edited:
  • Like
Reactions: 1 user
For reference, my experience starts with the Green LED bias network as captured by jreel in https://www.diyaudio.com/community/threads/f6-illustrated-build-guide.277850/post-7286651
That helps bring up the bias current to a stable value in a more timely fashion. I also like the (now discontinued) FQH44N10 Mosfets. They happen to like higher voltage and bias current. I have taken advantage of this. Power rails are +/– 26.5V and bias is 1.95A, as recently re-adjusted and confirmed.

The next part may muddy the waters more than some builders want:
The source resistors in the output stage may be split into two values – one above the connection with the 1000 uF capacitor and one below. This possibility is discussed in Nelson's original article on the F6, back when it was still using the mostly unobtainable SemiSouth devices.
I have set R1 to a sum of 0.40 Ohms, 1% below the cap junction + (0.1 Ω || 0.22 Ω) above the cap junction, or 0.468 Ohms total from the Source to the speaker output. It looks funny, but sounds very nice.
I set R2 to a single value of 0.40 Ohms, 1%.