Few issues about SSTC / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

Few issues about SSTC

1 2 next

Move Thread

LAN_403

steven968

Mon Feb 07 2011, 11:31AM

Registered Member #3516 Joined: Wed Dec 15 2010, 10:40AM
Location: Australia
Posts: 32

Hi guys,
This is my first TC, sch is Ward's minisstc and do a few modify: IGBT full bridge(HGTG40N60B3), input voltage increase cause the voltage here is 250v and new secendary(twice larger than orginial), I just wanna know if the length of sparks increase or pri-sec arcing? do I need to add zener diodes as well? (the original sch don't have it) and there is a cap which is 420uf will I be able to change it to 470uf 450v?

cheers.

ScotchTapeLord

Mon Feb 07 2011, 01:17PM

Registered Member #1875 Joined: Sun Dec 21 2008, 06:36PM
Location:
Posts: 635

Spark length will depend on your amount of primary turns. For the same given turns, your system should produce longer or at least thicker sparks than its 120V counterpart. Primary-secondary arcing always becomes more of an issue when you are dealing with more power, just design accordingly.

As for the 40n60's, the UCC driver chips will not nearly be enough to drive their gates if the calculations I just ran are anywhere near correct regarding their turn on/turn off energy. You will need to utilize something akin to that of Steve's later gate-driver designs using a FET bridge (bottom of

) OR use switches with more modest gate capacitances, or, alternatively, you can stack the drivers and limit your system to very low duty cycle (though in that case you are seriously derating your switches).

As for the bridge, your rectifier (D9), capacitors (C12, C7, C8), freewheeling diodes (D1, D2), and switches (Q1, Q2) will all have to be rated for at least 400V. Some will say that your pseudo-switch capacitors only need to be rated for half that, but if something goes wrong you'll be thankful for the extra leeway.

And finally, 470uF/450V is a fine value for that capacitor. Make sure it's close to the bridge rectifier as well as the bridge.

Gregory

Mon Feb 07 2011, 03:37PM

Registered Member #2922 Joined: Sun Jun 13 2010, 12:08AM
Location:
Posts: 226

You can drive the 40n60 with UCC! But for full-bridge use two in parallel each side of the driver! and the better GDT core and relation that you can get. You need to know that IGBTs are a medium speedy devices.. run at high frequencys only in ressoant mode.

ScotchTapeLord

Mon Feb 07 2011, 07:54PM

Registered Member #1875 Joined: Sun Dec 21 2008, 06:36PM
Location:
Posts: 635

Gregory wrote ...

You can drive the 40n60 with UCC! But for full-bridge use two in parallel each side of the driver! and the better GDT core and relation that you can get. You need to know that IGBTs are a medium speedy devices.. run at high frequencys only in ressoant mode.

UCCs can drive 40n60s at cw? As for speed, the datasheet shows that the switching times for this particular igbt is acceptable for fairly high frequency.

Gregory

Mon Feb 07 2011, 10:07PM

Registered Member #2922 Joined: Sun Jun 13 2010, 12:08AM
Location:
Posts: 226

At CW?? hahaha, never man! or yes.. But with a lot of heat! For CW with high gates capacitances devices use a dedicated MOSFET driver to the GDT, like IRF540/9540.

ScotchTapeLord

Mon Feb 07 2011, 11:12PM

Registered Member #1875 Joined: Sun Dec 21 2008, 06:36PM
Location:
Posts: 635

I believe the schematic he is using has an interrupter that can be set to 100% duty cycle, and therefore, CW, and it would be a very bad idea to use UCCs here in that case. I think they are rated for about 500mA of average output current.

Gregory

Tue Feb 08 2011, 12:51AM

Registered Member #2922 Joined: Sun Jun 13 2010, 12:08AM
Location:
Posts: 226

His schematic works because he is using a low gate capacitance mosfet

ScotchTapeLord

Tue Feb 08 2011, 03:05AM

Registered Member #1875 Joined: Sun Dec 21 2008, 06:36PM
Location:
Posts: 635

I acknowledge that the original designs work, but just wanted to point out that just directly replacing the FETs with 40n60s would be a bad idea.

dmg

Tue Feb 08 2011, 04:11AM

Registered Member #2628 Joined: Fri Jan 15 2010, 12:23AM
Location:
Posts: 627

Well, this schematic has alot of shortfalls, but doing things like parraleling UCCs is not the way to fix things.

Parraleling UCCs (or most ICs for that matter) might let you get away with it (and I mostly seen it happen in pulsed mode where it may just work), but it becomes alot more likely to fail. The fact that you want CW doesnt help that much either on this regard.
If you wish to drive larger devices, then you will need to impelement some sort of high side driver.
High side in the end is the only way to go about using this circuit with IGBTs.. seeing most of them need higher gate voltage, and higher gate capacitance also.

Also for some reason, I dont know if this is only happening to me, but my inverting UCCs keep failing on me, yet the non-inverting usually survives.
to fix this I just taped from the center of the NOT gate pair and fed that into a non-inverting UCC (so 2 NI UCCs total, just inverted inputs.). This solved my problem, but im still unsure to why the inverting UCCs themselves fail so much on me lately.

steven968

Tue Feb 08 2011, 07:13AM

Registered Member #3516 Joined: Wed Dec 15 2010, 10:40AM
Location: Australia
Posts: 32

So UCC's don't capable for driving IGBT's?
But what if 3 UCC's in parallel?

1 2 next

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