F5 Turbo Builders Thread

buzz. i'm gonna have 0.6A bias pr fet. and wondering if the best way is to use external sinks or main sinks for the diodes. the main sinks will run pretty hot i guess. 125+W pr side on 2x 10.080" profiles, 6" long.
on the other hand. with external sinks. how will the transaction be, when you WANT the diodes to conduct. i'm way out of the 1A ranch. maybe connect the diode sink with the main sinks with 2 L profiles?
i can not crank up the bias with rail voltage of +/-50V:D:D
 
Sanity check time for me....

V3, 4 pairs per channel, biased to 3,0 amps per polarity, 50 volt rails. That gives 0,75 watts/FET, or 0,35 Volts. Total bias current is 6 amps.

Now total dissipation is 8 * 50 * 0.75 = 300 watts. That can easily be spread over 4 regular sized F5 heatsinks at 75 watts per heat sink -- each heatsink carries one pair of FETS.

Efficiency is around 33% at full class A power. Does that sound about right?


As for Transformer VA, clearly 300 VA is needed just to run the bias. Now, add the load requirements. P=150 rms = 300 peak.

So, P=IV , V = 50 says I = 6 amps peak. That is another 300 VA to power class A operation. (That checks with doubling the bias current to come out of class A).

So far, the transformer has to be 600 VA *bare minimum*.

What do you spec for Class A/B overhead? Do you limit the VA to provide 50% of the steady state operating current of the output transistors?

Also, how do the diodes impact the heatsink size?
 
Last edited:
Gentlemen,

I am helping a forum member with his F5T -- I will save you the details and ask;

Would there be any problem building a v3 with ~ 42v rails and only 2 pair of output transistors?

Thank you,

-Jim
it will only be a high voltage V2. with output of 55W class A at 8ohm.
max safe bias will be about 1.9A. that will give you dissipation of 160W over 2 pairs. so 40W pr fet.
 

6L6

Moderator
Joined 2010
Paid Member
Or maybe put a Zen-style regulator in the PSU, referenced to 32v, There is at least a 4v insertion loss, I think... so would have to burn up 6-7v at ?200w?... That would take too many transistors. Nevermind.
 
Last edited:
If each Fet can handle 12 amps full load, then derating to 60% for longevity gives you a max of 7.2 amps/FET. That's and additional 1.2 amps x 4 fets *the other 4 are off*. Which is 4.8 amps or an extra 240 VA.

70% is 8.4 amps/FET => 2.4 * 4 = 9.6 amps or an extra 480 VA. That is a 1080 VA transformer.

Derating to 80 % gives your 9.6 amps/FET, or an extra 4 x 3.6 = 14.4 amps or 720 VA for a total of 1320 VA.

Looking at this backwards, if you have a 1500 va transformer, you've got 900 VA above 150 watt class A. 900 / 4 = 225 VA /FET = 4.5 amps. Which means that each FET can push 10.5 amps steady state, which only derates to 10.5/12 = 87.5 percent.

I'm not so sure this is a good idea. It is looking like a 1KVA transformer is about the max you want to supply to this sort of monoblock.
 
I'm struggling with this. Counting is hard....

If each Fet can handle 12 amps full load, then derating to 60% for longevity gives you a max of 7.2 amps/FET.

600 VA is used for 150 watts class A into 8R0. That's 6 amps of pure class A power. That is 1.5 amps over each FET, for a total of 2.25 amps/FET

You've got 5 amps/fet left to go, or 20a amps total. That makes for 1600 KVA total transformer size.

Is that about right?