Half or Full Bridge? / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

Half or Full Bridge?

previous 1 2 3 next

Move Thread

LAN_403

magnetomotive

Wed Jul 18 2007, 01:34PM

Registered Member #267 Joined: Mon Feb 27 2006, 09:44PM
Location:
Posts: 46

EDY19 wrote ...

There- Here is the new plan for a Half Bridge with a voltage doubler using 600V 30A diodes (which will be on a heatsink) and 2460uf per "leg." The Mosfets will be on a terminal strip, with leads bent 90 degrees so the mosfets are oriented vertically. 15V zener diodes will be across the mosfets too, I'm just going to put them where i can fit them after I build it

What is the purpose of the 6uF capacitor and the 10 Ohm resistors?

Also, what power level are you running this at?

ragnar

Wed Jul 18 2007, 02:17PM

Registered Member #63 Joined: Thu Feb 09 2006, 06:18AM
Location:
Posts: 1425

magnetomotive:

If you want to instantaneously open a switch and expect current to flow, you need to provide a source for it to come from. Foil or MKT capacitors tend to have a lower inductance and lower internal resistance than electrolytics. Therefore, the capacitors provide the current needed right on the switching transition -- the power supply or other smoothing caps can then take over after that, if that makes any sense.

The 10-ohm resistors are connected to the gates of the MOSFETs. Inductance of the gate-driver circuit, combined with the input capacitance of the MOSFET, can form a resonant circuit. On each switching cycle, the voltage on the MOSFET can overshoot and oscillate (imagine a decaying sinewave).

Because this oscillation might even be strong enough to let the MOSFET fall out of saturation, gate resistors are used to slow the switching time, and to lower the Q of the system. This means voltage spikes / oscillations have less magnitude, and less effect on the switch (and more importantly, the current flowing through it).

magnetomotive

Wed Jul 18 2007, 05:05PM

Registered Member #267 Joined: Mon Feb 27 2006, 09:44PM
Location:
Posts: 46

Thanks, that does make sense.

So basically you don't want the eletrolytic capacitors to supply that current because they would get too hot and would have a slower response?

CT2

Wed Jul 18 2007, 06:37PM

Registered Member #180 Joined: Thu Feb 16 2006, 02:12AM
Location: Ontario, Canada
Posts: 187

The electrolytic capacitors aren't "fast" enough, because they have a higher inductance and resistance current cannot flow as fast out of them. The smaller capacitance( 6uf) capacitor, is the opposite. It has low inductance and can supply alot of current very quickly, it drains really fast because of it's small size but by that time the big electrolytic caps can kick in.

Sulaiman

Wed Jul 18 2007, 07:59PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

I'm not sure how much use a 6 uF capacitor, no matter how high quality, will be
I've no idea what frequency you intend to operate at, for example 250 kHz
at 250 kHz the impedance of a 6 uF capacitor is just over 0.1 Ohm
I have electrolytic (inverter grade) capacitors that have an effective impedance
(capacitance, resistance, inductance) of less than 0.05 Ohms from 10 kHz to 1 MHz
and only 0.006 Ohms at 100 kHz.
So to save space and money you could just use electrolytics.

Check the specifications for your electrolytics fist though.

In any case the addition of the 6 uF capacitor will do no harm.

Steve Ward

Wed Jul 18 2007, 10:15PM

Registered Member #146 Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055

I've no idea what frequency you intend to operate at, for example 250 kHz
at 250 kHz the impedance of a 6 uF capacitor is just over 0.1 Ohm

Consider the speed of the switching may have content in the few MHz range. In this case, the 6uF is rather low impedance, and the lytics look worse because what ESL they do have looks pretty big now. But, i might agree that simply using high quality lytics for inverter use is probably enough. Most lytics scrounged from old electronics are probably NOT inverter grade and should be bypassed with a film cap. Its the same deal as logic design where .1uF caps are placed near each IC.

Sulaiman

Thu Jul 19 2007, 01:44AM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

I think it's better without the film cap because;
the actual desired output current (SSTC or DRSSTC) is composed of repetitive half-sinewaves at the output frequency
switching transient currents and 3rd 5th 7th harmonics of the output frequency are undesirable anyway.

I have no data to support this, just a vague opinion.

I don't have your experience so just take it as a thought experiment - not an argument

Steve Ward

Thu Jul 19 2007, 02:18AM

Registered Member #146 Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055

I think you are mostly right, Sulaiman, but keep in mind that the current is not always soft switched. In that case, there is high di/dt across the device, so you do have high frequency harmonics produced. There is no way around this, and without the snubber cap there will be a voltage transient produced by the inductance of the half-bridge loop wiring and the di/dt. On the other hand, i did notice on my large IGBT h-bridge running many hundred amps, that the film caps didnt help at all. I used high quality (and very large) inverter grade electrolytic capacitors on the bridge and had a laminated bus structure for minimal inductance.

Steve Conner

Thu Jul 19 2007, 10:18AM

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

When I was making my DRSSTC, I found that certain values of film capacitor would resonate with the stray inductances of the bus structure at harmonics of the operating frequency. The 1uF film caps that I tried first turned out to be exactly the wrong value, tuning the resonance to the second harmonic. With soft switching, the majority of the ripple current is, indeed, at the second harmonic. So the result was a huge mess, several times worse than without any film caps at all.

But with no film caps, there were nasty spikes, due to too much ESL and, I suppose, imperfect soft switching. (or stray capacitances, tail currents or whatever.) I experimented with different brands and values of film capacitor, and eventually got it looking good.

EDY19

Thu Jul 19 2007, 03:56PM

Registered Member #105 Joined: Thu Feb 09 2006, 08:54PM
Location:
Posts: 408

I just found that the SCR that I have would work wonderfully (I think) for the two diodes in the doubler circuit-this way I can turn on the doubler with a small voltage instead of using a large switch and having the contacts burn out periodically from the near short circuit of the capacitors across them. My SCR is rated at 90A RMS and 1950A Itsm, so this should be fine, right? Or I suppose I could make the soft start circuit shown here:

And in that case, +V would be taken after a rectifier on the hot 120V line, and the drain of that mosfet would be where the neutral terminal would connect to the doubler.
Advantages/disadvantages to either?
I'll be starting a project thread soon

previous 1 2 3 next

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