New Build with IRF640s

[silverspring]

Well, decided to do some experiments with Class E. I've built a test setup which is fairly similar to the FAT5 at http://www.shortwaveradio.co.uk. This uses 4 IRF640s in a push-pull configuration.
With 13.14 volts and 4.5A at the drain, I measured around 50 watts output. I reckon the efficiency is just OK at around 84% although this seems a little low? Also I'm not sure on the waveform at the drain. It looks to be sagging a little. What do people think of it? Here is a picture of the actual waveform on my scope:

[Gilly]

Hi

Have had that waveform myself at some stage...
Not sure what frequency you are operating on....
You are very close to being ALL GOOD...the efficiency is a little low.....
The position of your Tuning Cap seems low in value and the loading Cap is high in value....
I would position both capacitors to their maximum and start again reducing the Tuning Cap a little until a Class E type waveform appears and then reduce the load a bit..... this should give more power output etc:
Just play around a little with the tuning....I'm sure the waveform will straighten up a bit......
You could also try at the low end of the band and at the high end and see if one is better than the other....than maybe add or subtract from the coil.....
Also the loading capacitor looks around 900pF...maybe try a 3 gang type to give it a little more Capacitance.....and see how that goes....

Hope this helps some...

What are the values of the load and tuning capacitors... ?


Wayne

[silverspring]

Hi Wayne,

Operating on 3.6 MHz

You can't really see much in the picture, but I also have a 500pF capacitor in parallel with the tuning capacitor and probably running something like 850pF total. Maybe I should remove the parallel capacitor and try a lower tuning capacitance. The load capacitor also 1900pF in parallel with 2 * 500pF.

I think I will try as you say and experiment with the value. What is strange with this setup however is I can't get any weird deformed waveforms. There isn't really any ringing or any other forms of distortion. The waveform only really starts to fall apart when there is way too little tuning capacitance causing a large current draw.

Thanks

[Gilly]

Hi

Don't know your name....

This is a photo of my 80M TX...



It might give you a few ideas etc:
Your coil seems rather small .... have you measured the Inductance ?
OK on your Capacitor values...they seem about right.......850pF for the tuning cap is high so to achieve a Class E waveform you would have to reduce it quite a bit......
On my photo its hard to see but when things are all going well...you should have a high C for the tuning....
Your timings not quite right.......causing the right side of your waveform to take more time to discharge to zero...
Looks too as though the voltage of the waveform is around 40Volts which is good .... 12V X 3.5 = appox 40V so your Shunt Cap are OK....
I would check again the inductance of the coil ... you are very close but just not quite there......

HAS THE PHOTO COME OUT OK ...IT SEEMS TO HAVE CUT THE RIGHT SIDE OFF ?

Wayne

[Gilly]

Hi

Had a look on the FAT5 web site and your coil dimensions seem to be very close to what they have....also they state an efficiency of 85%....so maybe your setup going by their design is is similar / same.....
Most guys on this Forum would not be happy with 85% efficiency
To experiment I would increase the voltage to 15Volts but no more as the IRF640's are only 200V devices...
Increase the inductance of the coil and start the Tune Up process again....

Wayne

[silverspring]

Hi Wayne,

Your TX looks nice. I like how neat your coil is - it's perfect.

The inductor in my TX measures around 4uH, same as that on the FAT5 site. I've just moved the scope over to the other set of FETs and the waveform is perfect as below (actually, it just seems too perfect). Making adjustements to both capacitors only affects the output power and current drawn. I don't get any distortion unless I really wind down the tuning capacitor to too low value.



So clearly there is a timing issue on that part driving the first bank of FETs. Hmmmm.

Dean.

[Gilly]

Hi Dean

The Waveform does look nice....
Read Steve's WEB site as well http://www.classeradio.com/tuning.htm
You will now have to check the driver waveforms.....its all good fun really and you get to know what the waveforms are like in more detail...and gives you more things to talk about on air
Remember they will not be identical nothing is....
Check using your CRO the input to the drivers TC4422 and output without any Drain Voltage applied....to the FETs...
In this testing phase I usually disconnect from the Gates of the FETs as well...
The inputs should be around 4 to 5Volts and the output not less then 8Volts using IRF640......also check that the Power Supply to the driver is good too....power to the Driver IC's has to be perfect...
The waveforms should similar maybe one with more ringing doesn't matter really...
Then make the connection to the Gates of the FETs and check again.....the waveform at the gates should be more loaded now......but similar.......
Like here .... http://www.classeradio.com/driver.htm
Check more then once being sure all is well......

See how that goes.........

Wayne

[kf1z]

I don't think I'd expect much more efficiency from IRF640s especially at 12 volts!
lower voltage will equal lower efficiency.


Also, believe it or not, peak efficiency does not come with a perfect "Class E" waveform!
I have found that the leading edge of the waveform will be flattened vertically a bit, at peak.

I didn't see any of your pics... did you do the output transformer configuration just like FAT5?

What about your shunt capacitors... what are those?
They play a KEY role here.

Bruce
KF1Z

[Diego-CX4DI]

Hi Dears,

It's not possible get more efficiency than 90% using ferrite transformers.

BR
Diego
CX4DI

[kf1z]

Hi Dears,

It's not possible get more efficiency than 90% using ferrite transformers.

BR
Diego
CX4DI

I'm curious about that statement.