Plasmasonic II build / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

Plasmasonic II build

first previous 6 7 8 9 next

Move Thread

LAN_403

quinn

Thu Jan 07 2010, 01:58AM

Registered Member #2460 Joined: Thu Nov 05 2009, 12:18PM
Location:
Posts: 6

Dr.,Can i use Mosfets with 3 Pins just like IXFK44N50 for the plasmasonic system?Will there be anything wrong with it?

HV Enthusiast

Sun Jan 10 2010, 12:52AM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

quinn wrote ...

Dr.,Can i use Mosfets with 3 Pins just like IXFK44N50 for the plasmasonic system?Will there be anything wrong with it?

Sure you could.

However, you need to ensure you have VERY GOOD thermal transfer between those devices and the heatsinks. Also, the tab on those isn't isolated, so you'll need a thermal interface material thats electrically insulative.

Just keep in mind that unlike DRSSTCs, plasmasonics are CW type coils so they are running at very high duty cycles and dissipating LOTS of power. As long as you can keep the devices cool, you should be okay.

Also, you should make sure the TL494 chip is configured for 25% max. duty, not 50%. This will reduce the power dissipation of the MOSFETs.

Will.S.L.Liu

Fri Mar 19 2010, 02:45AM

Registered Member #2743 Joined: Fri Mar 19 2010, 01:59AM
Location:
Posts: 9

Hi there~! I am working on the Plasmasonic II board,and well....loved it ~!haha ~! Although I will be really happy if it doesn't make me crazy and starts working............
I am having some problems here....

First of all ......my board only have a frequency from 108KHz to 460KHz, which in the previously page, Moderator already said it should be ok~!so I think it's find too~!

next ~!....I have got the expected waveform before by using my 24AWG magnet wire with a in put voltage no more than 35VDC. the wire gets a little warm....and I think it should be fine .....then I just keep increasing the voltage to almost 70. Where the wire is start to burn......I realize that the skin effect for 24AWG only go up to 68KHZ....so I was thinking that may become the problem ....but I am still not sure....

I tried a couple of different wires: solid 27AWG; Standard 24AWG, magnet 28AWG, even solid 22AWG. and non of them give me the correct waveform....
the final wave form look should look like a positive pulse with a 0V and a negative pulse with a 0V within one period...but those 4wires that I have mentioned will not give me the correct waveform.....instead of that, my 0V is pretty much not there at all...it seems like it's trying to settle down to 0V but it doesn't ....and I end up with pretty much like a big positive voltage and a big negative voltage with no 0V......(MOSFETs is not turning off?! but when I use 24 AWG magnet wire, They turns of....x-()

I already changed my transformer for more than 20 times... 3 wire twisted or non-twisted I tried them all....my coil winding is really good, but they seem like very sensitive...even the 24AWG magnet wire which is the working one...I thought it a little ...it will change the waveform totally.....

And ...That's me ~! correct waveform with burning wire and working wire with incorrect waveform...

I really don't know what I am going to do now ...need you guys' help~!

Thank you so much ....

HV Enthusiast

Fri Mar 19 2010, 12:50PM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

Frequency is fine. You really want to have frequency less than 300kHz if possible, 350kHz maximum to minimize switching losses.

I wouldn't use magnet wire. The enamel on magnet wire is typically not rated for the voltage you need. Try some 24 or 26 AWG teflon or pvc coated regular stranded wire.

Can you post some pics of your waveforms?

The gate transformer is really, really simple and shouldn't be an issue at all. The fact you are having problems with that makes me thing something else is up somewhere.

BTW, how are you measuring the waveform? Is resonator connected? Are you using differential method for measurement (two probes?)

Burning wire is not good. Something is not connected up right, or you are measuring single-endedly.

Will.S.L.Liu

Sat Mar 20 2010, 02:29AM

Registered Member #2743 Joined: Fri Mar 19 2010, 01:59AM
Location:
Posts: 9

Hi:
There are some of the pictures that I took today ~!

This is my board design...The MOSFETs are the 3pin MOSFETs which the characteristic are very same as the 4 legs MOSFET. the big black box at the back is the four MOSFETs with heat sinks. MOSFETs are running through 18AWG PVC wires to the PCB.
SNV30417

The next one is my transformer winding by using 24 AWG stranded wire. I know I suppose to do a trifillar winding, but trust me , I have been trying different combination of wires ,either trifillar winding or non twisted winding, they don't really have too much differences...
SNV30416

The following pictures are the wave forms :(please ignore the frequency measurement on the picture...there is some thing wrong with the scope I think....but only this function.)

Gate waveform before adding DC voltage to the full-bridge switching power supply.(4wave form are the same picture. and they are either in phase or out off phase which matches the instruction book)
SNV30418

Gate waveform after adding DC voltage, the waveform changes(which i think the are not suppose to change.) : Q1 and Q3 are the same , and Q2 and Q4 are the same ....
Q1.

Q1 and Q3 is
SNV30421

Q2 and Q4 is
SNV30422

the final waveform by using differential method for measurement is (2.5us based)
SNV30427

and this is the final wave with 10us based.
SNV30428

I have changed a lot of different wires like I mentioned before to do the transformer, but unluckily, all of them give me pretty much the same waveform as I posted today.....

My MOSFETs are not heating up that much but the four TC4422 in the SMPS was heated up .....the 12V power supply is drawing 0.3A current at the Full-bridge increases to around 40VDC.

I already tuned my pulse IC to 25% duty for push-pull.....and I really don't know why they magnet wire 24AWG could give the waveform but overheated, and others doesn't heat up at all but not giving the correct waveform.....

Please help me ....I am going crazy........
I don't know why does the picture goes all to the bottom....but you should be able to tell right ~?!hehe ~!

Thank you ~!

HV Enthusiast

Sat Mar 20 2010, 01:13PM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

The gate waveform looks good and the TC gate drivers are supposed to get warm - they are switching lots of power to those MOSFETs continously..

Did you try running with a resonator yet?

What frequency are you running at?

Will.S.L.Liu

Sat Mar 20 2010, 04:17PM

Registered Member #2743 Joined: Fri Mar 19 2010, 01:59AM
Location:
Posts: 9

Those are the waveform without the resonator, and also with out the resistive load....
as you can see in the last picture , some of the MOSFETs are trying to close (which they should be) but they don't some how, and goes back to open.....

and the frequency of this is around 400kHz which is the highest frequency I can get ....and I tried to change to different frequencies such as lower frequency, the waveform at the end get even worse....the MOSFETs are not every trying to close any more.....so the waveform looks like a pure square wave......

and by the way, after adding voltage on to the Full-bridge, the waveform of the gate supposed to change?! why is that ?! as you can see...my fourth picture is the gate waveform of Q1 and Q3, and fifth picture is for Q2 and Q4. As you can see, the are very different from voltage level and waveform......is that supposed to happened?!

Wired right >?!

Steve Conner

Sat Mar 20 2010, 05:02PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Something is definitely wrong there. It looks like a lot of current is being drawn from your power supply.

Are you sure you wired up the MOSFETs correctly? (gate, source and drain terminals mixed up?)

Are you sure the gate drive to the MOSFETs is phased correctly? (if you get it wrong, two MOSFETs on one side can turn on at once, shorting the power supply)

The SOT-227 MOSFETs have two source terminals. Does Dan's board need both terminals to be wired?

Are you sure those toroid cores are made of ferrite, and not something else?

Will.S.L.Liu

Sat Mar 20 2010, 06:36PM

Registered Member #2743 Joined: Fri Mar 19 2010, 01:59AM
Location:
Posts: 9

well....I am pretty sure that non of the MOSFETs are wired wrong ....the MOSFETs are IXFK44N50.

I am using the same PCB as the guy named Quinn above...he is my friend...he already get the correct waveform....

the problem is I have got the correct waveform before which is using 24 AWG Magnet wire, the waveform comes out perfectly, but the problem is that wire started to get so hot after 50VDC and broke two of my toroids already....but the waveform comes out really nice....I checked my PCB for more than 10 times ....don't think any of the MOSFETs are wired wrong....

I know is so strange, I have already working this for more than 1 month with some of the professors in my school.... really don't know why .....

Steve Conner

Sat Mar 20 2010, 07:19PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

Edit: I changed my mind

You said you don't have any load on the thing... Try some load. This design turns all four MOSFETs off during the PWM dead time, so the outputs float, and I think that explains your screwy results. Just hook up a light bulb or something, it should light up and the waveforms should be OK.

Maybe your old GDT burnt up because you wired it wrong, or the insulation was bad, but as a side effect it loaded the outputs and made the waveforms look better. So maybe the new GDTs are good and that's what makes the waveforms "bad" :P

BTW I assume you know that two of the gate signals can't be scoped with power applied to the bridge?

first previous 6 7 8 9 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob