@wattsup:

Diode A

+>- 120.5vdc
->+ 102.1vdc

Diode B

+>- 69.6vdc
->+ 0.037vdc

According to these measurements Diode A is definitely fried, but Diode B looks good. 
Diodes typically have a voltage drop of about 0.7 volts, and to make a High Voltage diode they place many in series.
So having a voltage drop of 40 volts across a high voltage diode is quite normal. 
That small 0.03vdc leakage voltage is normal for the reverse voltage, a diode should block most voltage in reverse.
A little leaks by because of capacitance.
 

@Ghazanfar_Ali

I took the liberty  of removing the superfluous wires in your schematic as they are all sharing the same ground.

I probably do not understand the circuit well enough because I cannot see how the Q2 and Q3 mosfets can share the load if the gates are not receiving + voltage as they are grounded, so how can they open in tandem with Q1?


@ronotte and @AbbaRue

Thanks again for your appraisals. I have an extra good HV diode like the two others so I will put the set-up back together and continue on while I order more HV diodes (since I have a good record of frying components maybe I should open up an EE Bar-BQ restaurant).

I also forgot one major thing and that is the capacitor values of my string of capacitors it only 562pF which is way below the recommended 4000pF range so I am definitely not getting the HV punch per pulse. I have some more ordered and already in transit.

I built a new pick-up air coil that has three coils (see second image below).

1) One single strand (red) going left to right at first layer
2) One multi-strand (left white) wound two layers going 2/3 from left to right.
3) One Litz wire of 1650 strands of wound three layers on right.

This coil can slide into the Caduceus Coil so I can use it to see what is going on inside the CC under three different pick-up scenarios. One thing I am really looking forward to is when all three are shorted, what will happen to the output - hehe.

Anyways, more to come.

wattsup

Hi Ali,

Thanks for your answer. By now I have remembered where else I have already seen this circuit principle where a DC voltage source is connected across a coil wound onto a ferrite toroidal core by a very fast switch. Here is a link in English and next in French where Zoltan Szili explains how he got over a COP of 5 performance in a Spice simulation.

http://www.rexresearch.com/szili/szili.htm

http://quanthommesuite.pagesperso-orange.fr/circuitzolt.htm

As he explained he used the Jiles-Atherton Core Model for handling coils with nonlinear ferromagnetic cores, here is a link to such a model as used in HSpice:
http://www.ece.uci.edu/docs/hspice/hspice_2001_2-91.html 
(This model can handle core saturation. I think Mr Szili used it in the MicroCap circuit simulator.)

He also found it important in the simulation like you how fast the rise time of the input control pulse was (10nsec or less) and also the material of the core was 3E5 type from Ferroxcube, as well as the MOSFET type. (His coil had 40 turns on the specified core but he did not use any parallel capacitor like your 1 uF or now your 18uF.)

I wonder if your simulator (NI Multisim) considers lossy components like in real life a capacitor has an equivalent series resistance (ESR) or a coil has copper resistance and core losses too. IT would also be important to know if the inductance model in Multisim considers core saturation or hysteresis losses.   The model for the C2 is also important when you assign an initial charge (IC) to it, whether the capacitor model uses a voltage source for representing your 24V charge with a zero inner resistance or it is user defined. (I do not know, how it is in Multisim.)

I hope this Szili circuit addition helps you and Dole and others involved in this interesting setup. Keep up the good experiments.

Greetings,
Gyula

Hi Gyula
Thanks for your appreciation. I went through the link and read about Mr.Szili's circuit. It is of course similar to that of mine but u will notice that I am using only one voltage source which is driving the pulse generator (5V DC). Instead of other voltage source i am using C2 4700uF which is initially charged to 24V DC. Well interestingly I am using 18uF Capacitor in parallel with L1 to make it a Tank circuit well tuned to provide oscillation. By introducing 18uF C1 I am getting the phenomenon by which C2 never gets discharged and drives the whole circuit to produce high voltage across L1. The kick offs at this resonant state keep C2 charged. So I think its a lot of achievement. U just have to tune the circuit to its resonant frequency. One more thing as I wrote 1nsec is the minimum acceptable rise and fall time for the pulses in my circuit. Where as in Mr Szili's circuit he says that rise time should be less than 10nsec. I have tested my circuit for rise and fall time between 1nsec and 500nsec and best results are between 250nsec and 500nsec. i am attaching the performance at 500nsec. See the readings. Pertaining to your query about losses addressed in the pspice model, I am sorry I never went into that detail at simulation level. I know the results would be different at hardware level but not totally different.
Thanks once again for your response.
Kind Regards
Ghazanfar Ali

Input Power 5V * 13.397mA = 0.66985mW
Output power 200V * 457.656mA = 90.3312W 

Those calculations are conceptually invalid.
Multiplying average Volts and Amps yields average Watts only for Direct Current!

To properly measure non-DC power you must multiply the instantaneous amps and volts and then average out the results.  Reversing the order of these operations (averaging first and multiplying later) is simply WRONG.
Proper power measurement can be done with an analog multiplier (e.g. ADL5391) or with a multiplying oscilloscope that samples the current and voltage at high rate on 2 channels simultaneously.  In both methods, results of these multiplications need to be averaged out afterwards.

 

@Verpies
Thanks for your observation. I am attaching the schematic with wattmeter connected at the output stage across 150Ohms load. The average Wattage is around 42 W. Is this one correct now. This wattmeter of multisim calculates both for AC and DC signals.
Waiting for your reply.
Regards
Ghazanfar Ali

Wait another 40 minutes  and it should be ready


OK is ready
http://www.youtube.com/watch?v=Mn8TLBsR3r0&feature=youtu.be
Wesley