Kapanadze Cousin - DALLY FREE ENERGY

[zcsaba77]

The genuine TI-manufactured TL494, with careful circuitry, can be made to oscillate reliably at more than 800 kHz and with careful output stage pulse shaping can deliver a final power pulse train with the needed nanosecond-class rise and fall times. Don't believe me? See my "TinselKoil 2" videos.

I dont found it. Please link! Thanks

[TinselKoala]

I dont found it. Please link! Thanks

Here's the 830 kHz version. Later I made a new resonator and brought the frequency down to a less stressful 350 kHz or so. I found that when "overclocking" the TL494 the two complimentary outputs aren't so symmetrical any more, so I just used the + output and split-inverted that using a couple of 2n7000 minimosfets after the TL494, to drive an intermediate current amplifier stage before the main mosfet H-bridge. If you use a mosfet driver chip you can eliminate the intermediate stage but for overclocking I would still recommend just using the single 494 output and splitting/inverting externally to get your complementary square wave drive signals.
Also, if overclocking, make sure your chips are genuine Texas Instruments TL494.

http://www.youtube.com/watch?v=7BMRJoqdy6E

[TinselKoala]

(snip)
-----------------------------------
Side track project-
I have just assembled the 5mhz induction heater today but i'm using slightly inferior available mosfet IRFP450(higher Rds on resistance) instead of IRFP460.
http://4hv.org/e107_plugins/forum/forum_viewtopic.php?142264

Officially this the first time i replicated something that made in Russia and it works base on schematic.

This is really "cool" project, I'm starting one myself. I think that some improvements can be made, especially with layout. It looks like an effective, simple and cheap small induction heater. My wireless systems also work well as induction heaters but are too low-frequency to be very convenient for small items that I want to heat, and this one seems ideal.

I was able to get metal nail really hot but wasn't able to see glow on nail using variable 30....40volts 10Amps power supply.

I think he was using 50 V or so for his input. Did you also put a light bulb in series to see the current?

I'm using some excess silver-mica capacitor 1600volts 2000pf for this induction heater project which was previously purchased from Ukraine around 2 months back.
"This is to be used in the snubber circuit for the Don smith as found in page 260 of pjkbook.pdf."

For 1.5nf i am using available mica capacitors 560nf x 3 in parallel.

Uh... you must mean 560 pF  .... ?

Placing oscilloscope probe nearby the frequency seems started with around 5mhz then it slowly started drifting slightly higher to around 5.2mhz(I think this is due to the silver-mica capacitor is getting hotter)

Current consumption is around 40voltsx 1.6Amp and slowly increasing.

Or the mosfet itself. As it heats its Rdss rises and this will change the frequency too. I think also the coil and where it is tapped will be critical for good performance too. As the frequencies of our devices rises, our construction, layout and small differences in components become increasingly significant. I usually don't work much with frequencies in the MHz range for that reason. 5 MHz is pretty high for clipleads and breadboards.

[magpwr]

This is really "cool" project, I'm starting one myself. I think that some improvements can be made, especially with layout. It looks like an effective, simple and cheap small induction heater. My wireless systems also work well as induction heaters but are too low-frequency to be very convenient for small items that I want to heat, and this one seems ideal. I think he was using 50 V or so for his input. Did you also put a light bulb in series to see the current?Uh... you must mean 560 pF  .... ?Or the mosfet itself. As it heats its Rdss rises and this will change the frequency too. I think also the coil and where it is tapped will be critical for good performance too. As the frequencies of our devices rises, our construction, layout and small differences in components become increasingly significant. I usually don't work much with frequencies in the MHz range for that reason. 5 MHz is pretty high for clipleads and breadboards.

hi TinselKoala,

This  5mhz mosfet base oscillator is indeed a cool circuit.Later i will connecting the drain to the bottom of tesla coil to see if i can get any special effect using 5mhz

You will need to use light bulb for sure but later part i just remove bulbs and use multimeter which is set in 10Amp range.I'm using 22volts TVS diode(Planned to use IGBT later on) in parallel with 1uf ceramic cap to maintain a stable voltage."I have tested it in multisim but can't get 5mhz in virtual lab since i'm not using right coil".
I'm not using 10uf but i'm using 10,000uf 50volts capacitor.There is also 330nf 63v cap connected at the circuit power supply.

The bulb needs to be dim but upon inserting it needs to be brighter and not another way around."You will find out soon if the osc just stops if variable voltage is set too high"

Oops it was typo error suppose to be 560pf x 3 to get around 1.5nf.

I am using 14awg for the coil.total 8 turns ,3(center tap to +)5 turns.

All the pins needs to be soldered closely at this frequency ,breadboard at this frequency is out of the picture.

------------------

Latest update-1N4148 base AVR plug to white led does glow if placed around 1inch away from work coil.

[Marsing]

330 nanoseconds is not "nanosecond pulsing" it is "hundred nanosecond pulsing". You are still off by two orders of magnitude. Besides, it isn't the _length_ of the pulse that is critical, it is the rise and fall times. Certainly, short pulses must have even shorter rise and fall times, but if you can oscillate at 300 kHz _and_ make your rise and fall times be under 10 ns, you will be in the right ballpark.  The genuine TI-manufactured TL494, with careful circuitry, can be made to oscillate reliably at more than 800 kHz and with careful output stage pulse shaping can deliver a final power pulse train with the needed nanosecond-class rise and fall times. Don't believe me? See my "TinselKoil 2" videos.

Sure, 330 nS are in hundreds and i didn't yet consider fall times and rise times. that's just an example of calculation. and i don't think we need to make a signal with 1 ns precession , by now duty cycle is the key i guess.

what do you mean " You are Still off By two orders of magnitude" ?,   HA...    HA...

--------------------------------------------------

maybe we can make 1 % duty cycle.

if  F =  300 Khz    AND   Duty cycle=  1 % then    3,3us / 100 = 33ns
if  F =  1 Mhz      AND   Duty Cycle=  1 % then    1  us /  100 = 10ns   

now we have  PWM 3525 to generate signal up to 1Mhz by magpwr and TK 830 Khz.
...

edit....    '3825'    not  '3525',    sorry.