Kapanadze Cousin - DALLY FREE ENERGY

[T-1000]

Video here:  https://www.youtube.com/watch?v=5-vilYPhNn4&feature=youtu.be

Regards Itsu

Hi Itsu,

I just replied to your video there:

It seems your push-pull driver gates signal are not fast enough to close before other side opens. My suggestion would be to use Oleg's PLL push-pull circuit from http://imgbin.org/images/26553.jpg with duty cycle control so you could adjust it to have right timing. This can be adjusted on DT pin4 >resistor>VSS pin3 of 14 pin IR2184 MOSFET driver.

Cheers!

[Jeg]

I just looked at your video. It looks like what d1 described earlier. Your gate opens earlier than it should. This is because it sees a negative spike at its drain. Negative spike is being produced due to the positive front of the other mosfet.  Perhaps a diode as I told you before between drain and source can help.

[skywalker66]

@ Itsu
Nice video ! as usual.
It seems you got an error in your setup. Your 2 x12 turns primaries look like being in bucking mode, which is wrong.
If that's the case please change direction for one of them, and your setup will perform nicely.

[itsu]

Itsu.
I have the same results with you. I mean that with allega's snubber, oscillations still exist. It is just that it cuts the first higher peak of the oscillations down to half value than without the snubber.

Your result using the snubber calculator is remarkable. I will try to do the same. Don't you think that a diode is missing from your drain? (cathode to drain anode to source)? It will ground the negative peaks as they can't be grounded through your tvs combo. I think.

Hi Jeg,

the linked website to the RC snubber only mentions the RC, so i already added the Bat 46 diode across the R.
I know what you mean by adding another diode across the drain source, but thought it would not be needed.

But i did anyway now (another bat46) from cathode to drain and anode to source (so parallel to the internal MOSFET body diode) but it did not change the signal nor the behaviour with the gate lagging the drain 

I just looked at your video. It looks like what d1 described earlier. Your gate opens earlier than it should. This is because it sees a negative spike at its drain. Negative spike is being produced due to the positive front of the other mosfet.  Perhaps a diode as I told you before between drain and source can help.

i don't see that "Your gate opens earlier than it should", the scope screenshot  shows that the gate is not open yet while the drain is already on (if you know what i mean) 
And i don't see on the scope this "negative spike at its drain".

So somehow the drain is being activated by something else than its own gate

Thanks for your remarks.


Itsu

[itsu]

Hi Itsu,

I just replied to your video there:

It seems your push-pull driver gates signal are not fast enough to close before other side opens. My suggestion would be to use Oleg's PLL push-pull circuit from http://imgbin.org/images/26553.jpg with duty cycle control so you could adjust it to have right timing. This can be adjusted on DT pin4 >resistor>VSS pin3 of 14 pin IR2184 MOSFET driver.

Cheers!

T-1000,

thanks for your info, i doubt the ucc37322 is to slow, it is specified at 20ns rise/fall times which seems faster then the IR2184.
It will take major rebuilding to change it over, but i will keep it in mind if anything else fails (i don't think it will be of much influence as the resulting resonance sine waves are clean, but i just want to understand what is happening in my setup).

Not sure i understand what you are saying about the pin / chip layout (pin 14 of the IR2184?)

Itsu