It has been awhile since this thread was active but I see it is public so I would like to post an observation I have made. I was going thru some of my old files and came upon the attached simulation. I do this from time to time to see if I may have missed something as it seems I did with this particular circuit. The idea here was to take the reflection that returns to the input of a shorted symmetrical transmission line and charge a capacitor to see if the resulting energy levels had a gain. Although the sim uses an ideal delay line, I believe a bench transmission line will result in the same effect but with some additional losses.
The first pix is the schematic.
The second pix is the plot of the entire charge cycle of C1 from 0v to 36.82v. We see from the plot math that the input energy taken from the 20v DC supply V1 is 29.444mJ . We also can see that the energy in C1 at the end of 563us is 27.113mJ . Disregard the label on the math plot calculation for "V(vc)*V(vc)*40e6/2J" of mOhm/s which should really be mJoules. Never have figured out how to change those labels! Anyway, if the plot math isn't believable to you, simply calculate the energy in C1 yourself. So, we see for the entire charge period to 563us, energy is conserved.
The third pix now shows an abbreviated cycle that starts at 400us and ends at 563us. We now compare energies and see that it takes 5.925mJ to recharge C1 from 17.786mJ to 27.113mJ ! This is a gain of (.027113.017786)/.05925 = 1.57 . We now see that in this time period on the charge curve of C1, there is a gain!
With the starting voltage across C1 = 29.821v at 400us and the ending voltage = 36.81 at 563us, we calculate the average to be 33.31v DC over the charge cycle of 163us. If we now connect the proper load to discharge C1 from 36.81v  29.82v = 6.989v in the same time as the charge time of 163us and assuming relatively constant current for easy calcs at this point, this would require a current of di = dE*C/dt = 1.715A . With an average output voltage of 33.31v over the 163us discharge time, this requires a load of 33.31/1.715 = 19.42 ohms. The output over 163us would be 1.715^2*19.42 = 57.1W but this is 1/2 of the total cycle so the effective power would be 28.5W.
With a total cycle of 326us, the pulsing frequency in the delay line would be 3.067KHz! Sound familiar???
I don't know how I missed this earlier but of course I could be missing something!?!
I've also included the .asc file below.
Edit: Thanks to Gyula, he pointed out my mistake in that the energy levels should be in mJoules and not Joules. Corrections are made in the text above.
Regards, Pm
« Last Edit: 20240228, 20:20:19 by partzman »
