Kapanadze Cousin - DALLY FREE ENERGY

[verpies]

I used your idea of a trimpot in series with the 100ohm resistor in the monostable, but instead of 10k, I used 1k. It works pretty well from 150 ns up to 1.5 us, with no jumps.

150ns might be a too much forward conduction time for some DSR Diodes. Try to make it adjustable down to 50ns.

I have sharp pulses at the 1N5408, but amplitude is about 120 to 150 volts. I have a limitation, which is the 40 MHz response of my oscilloscope, so it is possible that the amplitude is much bigger, but I'm not able to measure it. Suggestions?Verpies?

The cheapest way to measure the peak voltage is by a small fast diode in series and a small capacitor (e.g. 200pf) in parallel (after the diode). This forms a primitive analog sample-and-hold circuit. The good news is that multiple pulses have a cumulative effect (short the capacitor to reset it).

Also, multiple pairs of transistors in SR flip-flop configuration can be used to form a digital sample-and-hold circuit, with infinite storage time (not useful for measuring peak voltage though).

For observing the ns and ps pulse widths, you would have to build your own make-shift oscilloscope.
This is not as hard as it sounds since it can be done with a shorted transmission line and many of those sample and hold circuits placed along the line that can be read later at much slower speed (maybe even with a HiZ voltmeter).

This trick works because a short pulse fed to a transmission line will reflect from the shorted end, travel back and superimpose on itself.
...and because the superimposed amplitude is 2x larger than the initial pulse amplitude (see here), it can be observed and memorized by the multiple sample-and-hold circuits placed along the transmission line.

Even in a 0.66 velocity factor coaxial cable, a 1ns pulse has the length of apx. 20cm. so if you put those sample-and-hold circuits every 1cm along the line, then you will be able to resolve down to 50ps. 
That's like a cheap 20GHz scope !!!

[Black_Bird]

@verpies

As always, you have great ideas. Thank you. I'll work on it.
PS: I measured the amplitude with your diode capacitor s/h. It confirms the scope measurement, so I don't have the real thing yet. I'll work on the nano pulser to reduce the pulse width and see what happens.

[itsu]

Looks like the KT926A transistor is not fully saturating.
When a BJT is fully saturated, the voltage measured across its emitter and collector should be close to zero (almost a short circuit).

guess you are right, and the BJT is not saturated. It needs 500ns pulse or more to do that, but that's to much for my nano-pulser, so guess need to modify my pulser with a LS7402.

Can you confirm the pulse needed on the cathode of last DSRD is plus, see last part of video?

Video here: http://www.youtube.com/watch?v=gPsansVHs-0&feature=youtu.be


@Black_Bird,

So our pulses across the 1N5408 are similar (130V).

I guess that when you really had 1Kv pulses your scope would had shown you that by smoke signals :-)

Regards Itsu

[Black_Bird]

@itsu
My understanding is that the diode has to conduct during the 100-200 ns pulse then get reversed, for the dsrd effect to appear. So, I think that the anode should go positive during the pulse, not the cathode.

[verpies]

the BJT is not saturated. It needs 500ns pulse or more to do that,

...or more current into the base.

Can you confirm the pulse needed on the cathode of last DSRD is plus, see last part of video?
Video here: http://www.youtube.com/watch?v=gPsansVHs-0&feature=youtu.be

If that positive pulse, visible on the red channel of the scope at 3m43s, coincides with the positive pulse on the base of the KT926A transistor then the DSR Diode appears to be connected wrong (backwards).
During that pulse, the DSR Diode is supposed to conduct forward for approximately 100ns.
Later, the current through the diode reverses but this is not due to the current flowing between the emitter and collector of the KT926A transistor (rather this reversal is a result of RLC circuit oscillation).

It would help to record the current trace through the DSR Diode, since the mere voltage is too ambiguous to analyze the DSR effect.