4 common configuration on Bridge / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

4 common configuration on Bridge

1 2 next

Move Thread

LAN_403

zzz_julian_zzz

Wed Jul 04 2012, 03:44AM

Registered Member #3964 Joined: Thu Jun 23 2011, 03:23AM
Location: Valenzuela City
Posts: 332

-Half Bridge with Non doubled supply
-Half Bridge with doubled supply
-Full bridge with Non doubled supply
-Full bridge with doubled supply

Which of these bridge configuration stays the best considering the following:
- 1200v igbt brick / 300 Icm
- low peak current on switches (500Apk)
- mains 230 VAC
- Resonating frequency of 70khz range
- ON time

my bridge is currently set up using Fb with non doubled supply..
but my bricks are capable for doubled mains, I am worrying on the current that will flow in it, for the same "ON TIME" as it were on non-doubled supply system. However, I think that the power inputted into the bridge will be the same since P=VI,.
so if V is doubled, I is halfed (good for bricks) BUT for shorter "ON time" here is another thing to consider, a friend told me that decreasing the ON TIME will not let the air warmed up for a long sparks to achieve. =/ any ideas what is the best solution???
Thanks in advance.

EDIT: I will have 400 Amps / 800 Icm IGBT module with 1200 volts but will take a month to be delivered on my house.

zrg

Wed Jul 04 2012, 07:57AM

Registered Member #4762 Joined: Sun May 06 2012, 05:59PM
Location: Russia
Posts: 93

Last. The higher is supply voltage the better coil works, both DRSSTCs and SGTCs. At least that's what I noticed during experiments.

zzz_julian_zzz

Wed Jul 04 2012, 08:49AM

Registered Member #3964 Joined: Thu Jun 23 2011, 03:23AM
Location: Valenzuela City
Posts: 332

-Half Bridge with Non doubled supply (long ON time)
-Half Bridge with doubled supply (shorter ON time)
-Full bridge with Non doubled supply (long ON time)
-Full bridge with doubled supply (shorter On time)

HI zrg,

you recommend the last, but also with shorter burst???? produces the best output?

Steve Conner

Wed Jul 04 2012, 10:17AM

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

wrote ...
if V is doubled, I is halfed (good for bricks)

If you wanted to keep the power constant, when changing from a half bridge to a full bridge, you should also rebuild the primary with twice the inductance and half the capacitance.

That is in theory. In practice, the non-linear impedance of the streamers complicates things. It means that increasing V doesn't increase I quite as much as you would expect. I design the whole system at once using computer simulation, to get the spark length I want, and then I don't mess with it too much afterwards.

Burst lengths for DRSSTCs typically range between 150 and 300us, though I have run as long as 10ms.

If you design for too short a burst length, you'll find that you need a low-impedance primary and extreme currents to achieve the power output you want. You will also be limited in the spark length compared to the secondary coil length, before you start to suffer from racing sparks.

Too long a burst, and you'll be making the spark brighter and thicker instead of longer. However that can be useful in demonstration work, to make a spark that shows up well on camera.

zzz_julian_zzz

Thu Jul 05 2012, 01:56AM

Registered Member #3964 Joined: Thu Jun 23 2011, 03:23AM
Location: Valenzuela City
Posts: 332

When simulating with LTSPICE, when i try to increase DC supply of the bridge, (from 330VDC non-doubler to doubler 700VDC) the current also rises on the tank, and I have to decrease the burst length to maintain the current flowing on the IGBTs because it is over on its rating. I cannot remake the tank for it is already built and have impedance of 9.9 ohms(high)..

"It means that increasing V doesn't increase I quite as much as you would expect."

-I expect it based from the simulation in LTSPICE, So in reality, the simulation in tank current of SPICE does not really compliment the actual?

another thing, is it better to just build a halfbridge w/doubler of cm400 rather than fullbridge w/doubler of cm200? considering the output spark length.

zzz_julian_zzz

Thu Jul 05 2012, 02:01AM

Registered Member #3964 Joined: Thu Jun 23 2011, 03:23AM
Location: Valenzuela City
Posts: 332

Sorry for double posting,
here is the Pictures of what I am getting from LTSPICE
We can clearly see that the current in the tank boost as twice when supply is doubled from 330 to 700VDC, Thanks

Steve Conner

Thu Jul 05 2012, 06:53AM

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

Well, do you have a dynamic non-linear streamer model in LTSpice? No, because nobody has developed one. So of course your simulation won't give the same results as the real system.

In grossly simplified terms, C4 gets bigger the more power you pump into the system.

zzz_julian_zzz

Thu Jul 05 2012, 11:16AM

Registered Member #3964 Joined: Thu Jun 23 2011, 03:23AM
Location: Valenzuela City
Posts: 332

In that case, I'll just ignore the current in the tank =) I'll just do it and forget everything hahah =),, I'll give a try to a FB doubler systems =)..thanks

Steve Conner

Thu Jul 05 2012, 11:52AM

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

The way I have approached this in the past is:

Use Spice's parametric simulation feature to do several simulations for different streamer loads. Say a load corresponding to 1 metre of streamers, 2m, 3m and so on. Terry Fritz's empirical 220k + 1pF per foot will do for this.

For each simulation, get Spice to calculate the power consumed by the coil. Plot this data as a graph of power vs. streamer length.

On the same graph, plot John Freau's empirical relation between power and streamer length: streamer length in inches is equal to 1.7 times the square root of the power in watts.

Where the two lines cross is a reasonable estimate of what spark length your coil will actually give. Also, the nearest simulation to the crossing point is a reasonable estimate of what your currents and voltages will be. (you can interpolate from the two points on either side)

I design my coils using this and FANTC, and they come out pretty close to simulation. As I was feeling cocky, I built my last one without an adjustable primary, and it performed to spec the first time it was powered up.

zzz_julian_zzz

Thu Jul 05 2012, 02:57PM

Registered Member #3964 Joined: Thu Jun 23 2011, 03:23AM
Location: Valenzuela City
Posts: 332

All I can say is WOW!.. =) I am clearly away from this topic/understanding hehe., actually, my software for t.coiling is just LTSPICE and Javatc =)) thats all I have hehe.. All I know is just design, not mind much of the plotting, and hope that the coil will work and produce what I expect ... tonight I just lost half of my bridge.. cm100(brick) from a very very small spark=') don't know the why IGBT failed, very very sad night for me, anyway, another thing, I dont know if i drive the bricks enough, that it is fully turning on and off, (UD v2.0 s.ward's) but slightly smaller GDT and only 12 amps(P-channel fet) and 18amp n channel fet driven by discrete transistor and enhancement mode mosfet.. hehe... the driver works for my smaller coil with IRG4pc50ud Igbts. Thanks for sharing Steve.. =)

1 2 next

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