[Amps] QST, and class AB/F MOSFET amps

[Manfred Mornhinweg]

Hi Mike,

> I have already worked out details for my liquid cooling scheme.  I
> found a good source for a cold plate and recieved it.

I would be interested in where I can buy such a cold plate! I googled 
for them, and found hundreds of models for CPUs and video chips, but 
nothing large enough for the power level I need, much less if all that 
heat comes from a few small spots!

I do have some machining ability, and a small lathe which I can 
configure as a milling machine, but copper isn't an easy material to 
work with, so if at all possible, I would prefer buying this piece 
ready-made.

>  I had a guy
> help me in USA (I am overseas in Japan) with adaptors.  The reason
> why I need adaptors is because the water cooling pump that I ordered
> has different connections.

That would be no problem for me, thanks to the lathe. I can make 
whatever custom adapters I might need.

>  If you need details, I can provide some
> but it sounds as if your water cooling scheme is figured out. 

Any help is welcome! All I have worked out so far is the basic equations 
for heat transfer from the silicon to the water. I have not started 
designing the cool plate. I think with the information I have now I can 
safely design the cool plate, but information about any practical 
example would be most useful, even if it is only to check the validity 
of my formulae!

 > I used
> a radiator that is close to what you would find on a car but is for
> industrial applications and was found on Ebay.  The fan is 12vdc but
> is so powerful that the stainless steel blades are flatted out on it
> when power is connected. 

I imagine how noisy it must be! Beware! I couldn't hear the DX! :-)

No, for my amp I don't want any radiator nor a fan. Eliminating fan 
noise is the whole point of water cooling! I intend to use a simple 
bucket of water under the desk. That has enough heat capacity for a 
pretty long operating session, and can cool off naturally during the 
non-operating time. If I feel generous, I might actually buy a water 
container that has a lid!

In the meantime, I ran some design exercises. One was about a class AB 
linear amplifier using 4 modules having each a pair of IXZ210N50L 
MOSFETs in push pull. With power combiners and all the stuff. Very 
conventional technology. It turns out that assuming a pretty good water 
cooling system, this approach still falls a bit short of delivering 
1.5kW safely in high duty cycle modes. For SSB it would be OK.

It seems that using the dual version of this MOSFET, or using two in 
parallel on each side, can easily lead to power distribution problems.

Then I calculated a situation with just one pair of DE475-501N44A 
MOSFETs in push pull, in class AB/C with envelope restoration. Class E 
or F seems quite hard to do, because I don't want to have to tune the 
amplifer! That's a detail I had not realized before... those classes are 
so critical in the proper resonance, that tuning to the specific 
frequency and with the specific antenna impedance is mandatory. Well, in 
class AB/C with a modulated power supply I can develop 1.5kW in 100% 
duty cycle with a single pair of MOSFETs, but the cold plate design (and 
specially its making!) become rather difficult. The same project, but 
using ARF1500 MOSFETs, is easier, because they are much larger.

All the above is assuming that the whole mounting surface of each 
transistor is thermally active! I don't know how far that's from the 
truth. The farther it is, the more difficult cooling becomes. And the 
more difficult cooling is, the more useful it is to run class C or above!

I also passed the information I have to a good friend, who is not a ham, 
but is pretty good at mathematically solving problems. I asked him to 
look into the differential delay issue, and into the phase modulation 
caused by the MOSFETs. Let's see what he comes up with! I'm pretty 
confident that the distortion from differential delay is acceptable for 
ham SSB use, but I really have no idea now how big the phase modulation 
effect is.

I was thinking about making a small scaled down model amplifier, using a 
pair of cheap small MOSFETs at 80 meters or so, running at perhaps 30 
Watt output, and measure distortion there!

I just need time.

Cheers,
Manfred.

----------------------------
Visit my hobby website!
http://ludens.cl
----------------------------