Akula0083 30 Watt Self Running Generator.

[Vortex1]

Try tuning by varying the gap ever so slightly with shims while observing scope.

330 pF would probably be the correct value for C1 and C2.

Try a little bit of biasing with a PM.

Regards, Vortex1 / ION

[TinselKoala]

hi Tinselkoala,

There is a small problem with 1n5817 rated 20volts if the peaks is hovering somewhere below 100volts.

Maybe a UF series diodes of around 100volts ...200volts rating would solve the problem of ou if your lucky on a Good Friday.

Let's hope it was not a Russian version of "DR Lirpa Sloof"

Heh... I think that Dr. Lirpa Sloof has been the mentor of these folks for quite some time...

Yes, I thought of that too. So I changed the diode to UF4001. The only change I can see so far is indicated by the circle on the scopeshot below.

[TinselKoala]

Try tuning by varying the gap ever so slightly with shims while observing scope.

The above shots are done with no gap, clean faces in contact, core halves squeezed together with rubber band.

For the TestPoint 1 trace, as I increase the gap the voltage goes up, to a  maximum of about 90 volts in the first peak, at about 1.6 mm gap. I'm just putting shims in on the "round" side of the flyback core and letting the rubber band pull the halves together, so the "square" side is still in contact.
This happens without any observable change in brightness of the LEDs with S1 "on" but _reduces_ brightness when S2 is "off". There is very little change in frequency of either the main pulsation or the double-peaks.

330 pF would probably be the correct value for C1 and C2.

Try a little bit of biasing with a PM.

Regards, Vortex1 / ION

I'll change the capacitors for the next tests.

Bringing a strong magnet into contact with the core, in combination with 1.6 mm spacer,  has a pretty strong effect on the double-peaks, turning them into a single peak. Max voltage (90 v) and main pulse frequency are about the same as without the magnet. The LEDs are visibly dimmer when I do this.

Still no tendency to keep on running when power is removed, though.   

[Ed morbus]

thank you for sharing this with us

[TinselKoala]

You're welcome.


Well. Going back to no spacer in the core:

Capacitors marked "330" are 33 pF. But just in case 330 pF is meant.... I did some experimentation with 330 pF. Changing C2 to 330 pF didn't have much effect except raised the peak voltage at Testpoint 1 a bit. Changing C1 to 330 pF made a big difference, slowing the main pulse down and other waveform effects. By playing with core spacing I was able to get a single peak within the main pulse, similar to what the video shows.

BUT... Fooling around in this mode I have managed to blow 6 LEDs and 2 MC34063 chips, and I don't have any more spare chips, so I changed back to 33 pF for both capacitors. The chips and LEDs seem to blow when switching S1 from "off" to "on" while power is applied, or just after running with switch "off", removing power and switching "on".

With 33 pF as C2 and C1: with S1 "off", applying 8.7 V power lights the LEDs well. Then when switching S1 "on" with power applied, the LEDs dim a bit. Removing power, they go out. If power is removed when S1 is "off", LEDs go out. Then switching S1 "on" makes the LEDs flash briefly and brightly once, as C3 discharges.

With 330 pF as C2 and C1, behaviour is similar except that when S1 is switched from "off" to "on" LEDs and chips tend to blow.

I'm going to try using a higher capacitance value for C3 tomorrow, if I can find some spare MC34063 chips at the dollar store. The LEDs I'm using are from a very cheapo "christmas light" string, I have a lot of them and they are really poor quality so I don't mind blowing them. Blowing the chip is a drag though.

I'm sure that the Eastern watchers are very amused by my efforts.