Kapanadze Cousin - DALLY FREE ENERGY

[magpwr]

Hi all 

https://www.youtube.com/watch?v=_wKZ9ehCciU

something for the view here.
The effect as how it should be. was very hard to get on recording since it was hard to maintain effect.
but moments after resolving the values of caps needed and finally did a good one.
Tuning is very precise , the raw tuning with these variable resistors on the TL494 board need to be refined.
When the output or series resonance is shorted the amp meter goes almost 0 amps on scale when connecting more heavier load
the amps go down instead of going up.

Cheerz~

Hi GeoFusion,

Good work.
There is only 2 comments base on my quick observation.

1) I noticed the spark from the top of tesla is violet in colour but noted sparks at Antenna is orange in colour which reveal there is presence of higher current.Nice work.

2)I do hope you record down frequency of tesla coil and frequency of pwm along with the duty cycle.This way it will no longer becomes a trial and error and most importantly it will help you to save time.

[itsu]

Void,

in looking for better snubber arrangements, i found this pdf:
http://electronics.etfbl.net/journal/Vol13No2/xPaper_09.pdf

In there we see our RC snubber and also the Oleg snubber here named RCD snubber.
There are some calculations mentioned that perhaps better match the components to the situation (heat).


Also mentioned in this paper is an active protection setup which should be superior to the above RC and RCD snubbers.
I think i might give it a try, however, there are some unknowns like the used components and where the PNgx 1 and PNgx 2 outputs are going to (ground?)


Anyway, might be interesting for you to.

Regards Itsu

[Void]

Thanks Itsu. I will take a look through that document. I had already downloaded about
15 different documents on snubber design, and have been going through them. Yes, the
Allega snubber is a type of RCD snubber. There appears to be two main types of RCD snubbers,
'voltage clamp' RCD snubbers and 'rate of rise' RCD snubbers. I have been trying the 'voltage clamp' RCD snubber,
but I plan to also give the 'rate of rise' RCD snubber configuration a try as well to see how its performance compares.
I have been busy the last couple of days, but plan to experiment more with the snubber circuits this weekend.
There is also a type of snubber called a 'non-dissipative snubber', which I may experiment with as well, but
I may go with an RCD snubber if I can get it to work reasonably well.

P.S. Itsu, I took a look through that PDF document you posted a link to, and I also am not certain
where the PNgx1 and PNgx2 outputs are supposed to connect to. From the wording in the document
"... switching on both powerful switches in push-pull inverter through PNGx.1, PNGx.2.", it almost sounds
like they may go to the gates of two MOSFET switches which turn on and limit the overshoot spikes, but
it is not clear to me as well, so I am not sure.

All the best...

[NickZ]

   Yes, it would be good to try different snubbers, as all the combinations that I've tried still get too hot to let run for over a minute or two.

[Void]

Ok, I did more testing with the Allega snubber arrangement.
It is a type of voltage clamp RCD snubber. With the specified 1K Ohm resistor, it does not limit the
voltage peaks enough for my own setup, but with a lower resistance value it performs better.
The main drawback of this type of snubber is that the resistor dissipates a lot of power,
so it seems to be quite a waste of power to me. I found that it works quite well even when using
a 0.1uF capacitor, so 0.1uF is probably good enough.

Test 1
I had two 47 Ohm, 10 Watt ceramic block resistors available, so I put them in series
to make an R of around 94 Ohms. That gives me quite good results as far as limiting the overshoot spikes,
(see the first scope screen shot) but it produces a lot of heat (wasted power) in the resistor! I left it running for
about 4 minutes or so and both 47 Ohm 10 Watt resistors were already around 75C (167F).
If left to run longer they likely would have got even hotter. The two 10 Watt resistors in series should handle it
with no problems, but you would need to keep the resistor or resistors away from anything that can get damaged by heat.
A single 10 Watt, 100 Ohm resistor may be able to handle the heat, but it seems it will get very hot.

Test 2
I used an R of 470 Ohms, 2 Watts. The resistor started to smoke in about 30 seconds or so!
This value of R still gives fairly good overshoot spike limiting results however, but a 2 Watt resistor is not
going to cut it at R = 470 Ohms. (See the second scope screen shot for the results).

Conclusions
A resistor value of 470 Ohms still gives decent results but you would want at least
a 5 Watt resistor or maybe even a 10 Watt resistor. As a compromise, you can choose
a higher R value between 470 Ohm to 1K Ohm, but use at least a 5 Watt resistor. Even at 1K Ohm,
I think a 2 Watt resistor may get very hot (in my test setup) if running for more than just a short time,
but it of course depends on the exact circuit setup and tuning and the type of load that is connected.

A ceramic block power resistor seems to work fine in a voltage clamp RCD snubber.
The high frequency ringing you see in my scope shots is probably due to the fairly long wires
wires I am using, as my snubber circuits are breadboarded on a solderless breadboard for easy testing.
The results appear to be about the same when using the ceramic block power resistor (which I think
is wire wound) and when using a 2 Watt carbon film resistor.

I think what Allega did was combine two separate voltage clamp RCD snubbers into
one snubber circuit, so the resistor is doing double duty. I think you should be able to separate
the one snubber circuit into two separate RCD snubber circuits, one for each MOSFET, and then
the power will be divided between two separate resistors. In that case, you should be able to get
away with a lower power rating on the resistors, depending on the value of R you are using.

Nick, I am using an R of 220 Ohms, 2 Watts, and a C of 10nF in my series RC snubbers (Drain to V+)
and the resistors stay completely cool as long as I have the Allega snubber connected, so the voltage clamp
RCD snubber really helps to keep the resistors in the series RC snubbers cool.

Tomorrow I may try the 'rate of rise' RCD snubber configuration for comparison if I get the chance.

P.S. I was testing the Allega voltage clamp RCD snubber setup, which is a little different than the
voltage clamp RCD snubber schematic I posted in my previous comment. The difference is that the
Allega snubber configuration has one leg of the capacitor going to ground, rather than going to V+.
This configuration seems to give a little better results in limiting the overshoot spikes than when
both R and C both go to V+.   

All the best...