F5 Turbo Builders Thread

In the F5T article NP says that diodes conduct at between .3 and .6v depending on temperature - hotter meaning lower - also I would be planning on 4 ohm loads so they would seem to be likely to do _something_ in that scenario.. albeit further into the conduction of the output devices than with the F5Tv2. Whether that something involves letting smoke out of things or making a horrible distortion or do something useful is beyond my feeble newb reckoning.. 😕 I don't even know really how to sensibly choose between the DOGC-H and F5T v2, other than I have a chassis with two toroids in already that will do for either (big with +-36vdc rails on caps) but included heatsinks are a bit wimpy (pretty chunky but very short fins) for the F5Tv2. That and my friend has a quad 606 that I like the sound of.. so this current dumping/error correcting feed forward stuff with low heat is appealing.. that and that the output devices don't need matching. Downside seems that not too much build info is about on the DOGC-H in English.. and there's buttloads for the F5 variants from all kinds of helpful people and it seems sonically very well regarded.
 
Last edited:
Resistors: (All are 1/4W unless otherwise specified)
R1 – 10R / 1W
R2 – 10R / 1W
R3 – 2.2K
R4 – 2.2K
R5 – 100R / 3W
R6 – 100R / 3W
R7 – 100R / 3W
R8 – 100R / 3W
R9 – 4.75K
R10 – 47.5K
R11 – 0.47R / 3W
R12 – 0.47R / 3W

This is from the partslist, I read it as R9 and R10 is 1/4 Watt. I understand that they logically should be 3 Watts, but it´s confusing to read: Resistors: (All are 1/4W unless otherwise specified)
 
Setting F5T v2 bias.

I am likely to miss something here - have nearly completed 1 ch of v2 with 2 OP pairs and with the diodes, 2 source resistors per FET (2 x 1 Ohm), rail V is about 34V. Tried to bias it yesterday, and was able to reach 0.555V (while did not ran out of steam on the multiturn pot). Sinks feel like ~50C (pretty hot, but can keep my hand on it indefinitely - that's after about an hour or more of running the amp).
The bias seems to be stable at that point (ambient ~20-21C).
Question - is not this V value a bit high (I expected bias to start running away in 0.350 - 0.4V interval)?
 
bearwik, I guess you are talking about this: http://www.diyaudio.com/forums/diyaudio-store/227933-diyaudio-f5-build-guide.html
then it does not match the one in the F5T article, hence different values.
Link <---> article
R1 – R3
R2 – R4
R3 – R5
R4 – R6
R5 – R9
R6 – R10
R7 – R11
R8 – R12
R9 – R1 (not exactly)
R10 – R2
R11 – R7
R12 – R8
etc.

If you are targeting F5T, stick to the schematic in the article - that will allow easier transition from F5 to F5T, and it's more up to date.
 
Some picks. Somehow I think they are not a great help 😀
Can anyone recognize the thick white PCBs? :rofl:
 

Attachments

  • Picture 56.jpg
    Picture 56.jpg
    426.9 KB · Views: 411
  • Picture 53.jpg
    Picture 53.jpg
    689.7 KB · Views: 410
  • Picture 50.jpg
    Picture 50.jpg
    547.9 KB · Views: 402