F5 Turbo Builders Thread

Well, heck, yeah!!!! Since we to "direct couple" the stages of our audio systems, why not "direct couple" our power rail requirements directly to our (local) power mains. Just think--eliminate the cost and weight of the toroid....

Hhhhmmmm--wondering about the safety considerations.....:eek:

For all of you "just dialing in". My comment above was in jest. I'd never design or build a circuit without a mains transformer--just for isolation from the mains, if nothing else.

Please don't try what I mentioned.....or if you do, please don't come looking for me afterwards.

I used to own a "universal" AC/DC transformerless shortwave radio (National Radio NC-60). It's amazing how I ever survived playing with that thing. Don't try this at home.....!! :headshot:
 
Suppose worst dissipation at 40V , 40 x 40 / 2 = 800W.
Max die temperature is 150C, 125 above 25C ambient.
On an infinitely large heatsink, 1 output device with a mica insulator could handle a little over 90W.

So 6 pairs would sufficient, with a heatsink that does 0,05 C/W.
Easy peasy with fan blown tunnel.
It's the cooling that's vital, not the output devices, with MOSFETs.
 

6L6

Moderator
Joined 2010
Paid Member
For what it's worth, Nelson has talked me out of using 65v rails on an F5Tv3. The voltage is too high and It would be running too near the edge.

Of course, Fearless Amplifier Builders can try it, but know that it's going to be a bit of a beast.
 
Think output level of 35V in 2 ohm, with 75V rails.
40V drop between drain and source, 17.5A current going through the mosfets : 700W dissipation.

Worst load for an output stage occurs when the product of output current times voltage drop is highest.
A sine curve shows that this happens below 50 percent of max output.

Pushing out full power is relatively harmless for an output stage : high output current, but low Vds.
 
I would be grateful for help confirming the following calculations as I haven't made a balanced amp before. I understand the calcs for a single ended design but I'm not quite sure exactly about the balanced amp figures.

Here's my plan to get a balanced F5 monoblock giving 100W class A.

Use 2 Mosfets in parallel, so 8 total. Each biased to 1.3A. This should give me 5.2 x 5.2 X 8 = 216W (into 8ohms), which is 108W average.

Use 25V rails. So 25 x 1.3 x 8 = 260W dissipation.

A heat sink with 0.1c/W leads to 26C rise in temp.

The following heat sink PS275 will be cut in half to make a tunnel with a fan to give some safety margin.

http://www.gdrectifiers.co.uk/upload...le_devices.pdf
 
Last edited:
Use 2 Mosfets in parallel, so 8 total. Each biased to 1.3A. This should give me 5.2 x 5.2 X 8 = 216W (into 8ohms), which is 108W average.
Correct

Use 25V rails. So 25 x 1.3 x 8 = 260W dissipation.
Correct

A heat sink with 0.1c/W leads to 26C rise in temp.
Correct (but you have to pick a heatsink that does 0.1C/W at ~50C)

Link doesn't work, and the site doesn't show a PS275, unfortunately. :clown:
 
f5 turbo version 50w 8 ohm, what case hifi2000 4u 400mm heavy is enough? Or which version of case in silicon ray?

50W Class A is about 120W dissipation per channel for the F5T.

HiFi2000 only has one 4U-400 model, either with grey or black front panel.
4U is for a 165mm high heatsink, 400 is for two 200mm wide heatsinks side-by-side.

For 50C heatsink temperature at 25C room temperature, you need ~0.20C/W per channel (25C / 120W)
Each of the heatsinks of the HiFi2000 case would have to be < 0.40 C/W.
For a stereo build, a better choice would be the 5U-400 model, imo.

Largest Siliconray case has 280mm wide x 150mm high heatsinks, 28 fins.
Not enough for two channels in one box.
 
Thanks Jacco, that's great. :)

Here's another go at the link:

Heatsinks and Hardware - Heatsink Extrusions - Heatsink - Stock Extrusions | GD Rectifiers

The particular seat sink can be found on this page.

My current F5 uses a transformer with 18VAC secondaries and I measure 25V DC on the amp when in use. So I will look for similar transformer but higher VA rating as it will need to idle at 12A comfortably :eek:

There's also no need to cascode the input. So which boards to use????
 
Use 25V rails. So 25 x 1.3 x 8 = 260W dissipation.

A heat sink with 0.1c/W leads to 26C rise in temp.

Correct (but you have to pick a heatsink that does 0.1C/W at ~50C)
apply the de-rating factor appropriate to deltaT ~30Cdegrees.
I'd guess at DF~ 1.2 to 1.3

This results in actual temperature rise of ~ 26 * 1.25, say ~32Cdegrees above ambient.