Ghazanfar Ali Generator - Utlilizing trapped energy

[forest]

There is some simulation program called Qucs under linux. Do you know how to set mosfet parameters to replicate exactly IRF260 ?

[poynt99]

If it does not already have a model for the IRFP260, then you can look at the SPICE model for it and change the parameters for your "default" MOSFET model in Qucs.

I have attached the PSpice model.

.99

[forest]

If it does not already have a model for the IRFP260, then you can look at the SPICE model for it and change the parameters for your "default" MOSFET model in Qucs.

I have attached the PSpice model.

.99

Thank you. I have no idea how to set some of them because in qucs they have single value while in Spice some of them have a set of values like : RS 8 3 0.026764
If  you can help me fix it I may run this circuit under qucs for an hour or so. preliminary test show 5 seconds no diminuition of oscillation, but that's not a proof of course


http://en.wikipedia.org/wiki/Quite_Universal_Circuit_Simulator

[gyulasun]

Thank you. I have no idea how to set some of them because in qucs they have single value while in Spice some of them have a set of values like : RS 8 3 0.026764
If  you can help me fix it I may run this circuit under qucs for an hour or so. preliminary test show 5 seconds no diminuition of oscillation, but that's not a proof of course


http://en.wikipedia.org/wiki/Quite_Universal_Circuit_Simulator

Forest:

In this line "RS 8 3 0.026764" the first two numbers 8 and 3 just after the component name (RS) refer to connection points in the schematic of a Spice circuit.  So those numbers are NOT a set of values.  The real value is only  0.026764 for RS.  So if you simply omit the two numbers just after the components names your simulator would probably be happy.

[duff]

 Here is my first test of the AliGen.

http://youtu.be/zI9xTK-hXO4

Component differences as noted in the video:
1. I am using IRFP260MPBF instead of the IRFP260.
2. A 17.4uF cap instead of the 18uF
3. I've added a FEP26JT diode pair as snubbers across the source/drain.
4. A 1mH choke is in series with the sig gen ground.

The Inductor was initially set to 377uH and then slightly tuned so that the LC circuit was resonant at 2KHz.

In my setup the peaks are located at the leading edge of the square wave pulses , whereas in Ali's sims the peaks are located at the trailing edge. I find this strange. I don't know what has caused it, unless perhaps the choke has caused the shift - I'll have to check it.