Voltage limiting in SSTC. / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

Voltage limiting in SSTC.

Move Thread

LAN_403

dmg

Wed Dec 29 2010, 11:12PM

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

I was messing around yesterday with the idea to shut off the driver of an SSTC after the input voltage to the bridge reaches a certain point. as often I accidentally input too much and it ends up frying my bridge.

I wanted to do this by using 2 non-inverting UCC drivers (set up as shown with the inputs inverting and non-inverting). and use a comparator with an inverted output to monitor the voltage, then when triggered. to turn off the UCCs.

Heres the schematic, I wondered if this is doable this way and I would appreciate feedback on it.

ScotchTapeLord

Thu Dec 30 2010, 02:31AM

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

The potentiometer needs to go from the ("Rset") to GND with the chip's pin tapping between the two resistors to make a divider. Looks like that was just a simple mistake, but the configuration you have would put the full voltage across the chip!

Also you'll have to switch the non-inverting and inverting inputs (+ and -) because with this configuration I believe it will not turn on until the + voltage is higher than the - voltage.

Looks good otherwise. You might as well design it to handle the full input voltage from the start because you know that's where you'll want to end up eventually!

dmg

Thu Dec 30 2010, 02:41AM

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

Sounds good. Ill do just that.

As for the potentiometer, yes, that was my mistake, I meant it as a divider to compare the voltage on the buss to the referance not a direct tap.

hvguy

Thu Dec 30 2010, 07:29AM

Registered Member #289 Joined: Mon Mar 06 2006, 10:45AM
Location: Conroe, TX
Posts: 154

This circuit is likely to oscillate about its set point. You should add some hysteresis or a one-shot on the out put to prevent any unwanted on/off cycling.

HV Enthusiast

Thu Dec 30 2010, 03:51PM

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

Your probably better off adding more margin with your bridge circuit as opposed to building a protection circuit. As i see it now, the protection circuit only disables switching - it doesn't prevent voltage excursions on the bridge itself, which i gather you are attempting to protect against.

If you truly want to protect the bridge from excessive voltage input, then you probably need to employ some sort of crow-bar circuit so that when voltage is exceeded on the bridge, the voltage is instantly discharged and latched off.

Also, your circuit will probably be quite prone to noise and spikes on the DC Bus. Even though its low impedance, you may still prematurely trip due to voltage spikes and/or ripple on the DC Bus. So in addition to hysteresis, you probably want some sort of filtering circuit as well.

However, in my opinion, i would just either ensure you prevent input voltage itself from exceeding a certain level (perhaps a mechanical stop on a variac if thats what you're using), or just build more margin into your bridge.

Just my 0.02

dmg

Thu Dec 30 2010, 07:08PM

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

EastVoltResearch wrote ...

However, in my opinion, i would just either ensure you prevent input voltage itself from exceeding a certain level (perhaps a mechanical stop on a variac if thats what you're using), or just build more margin into your bridge.

Id agree with this, as paying attention and taking some basic steps will prevent alot of heartache later on. but I was mostly just toying with the idea of having the driver itself do it for you.
Ive taken some of the ideas mentioned here and tryed my best to work them into the schematic, here is the updated version:

Once again, I love hearing feedback and suggestions.
Thank you.

ScotchTapeLord

Thu Dec 30 2010, 08:22PM

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

You can't have the IGBT up there as the gate needs to be charged through the emitter terminal, which is being blocked by your load.
You need to use a P channel device (which may be hard to find for your level of power) for this configuration...

You might be better off with zener diodes in series (to raise the voltage and spread the heating) to passively clamp the voltage, along with maybe an indicator light in series with the zener diodes to let you know when they are conducting. That seems like an easier idea to implement, and you can easily remove it from your circuit if you want to get rid of it.

HV Enthusiast

Thu Dec 30 2010, 09:12PM

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

That type of switch implementation is not going to do anything to protect your bridge against over-voltage. As i said before, the only way to protect against overvoltage is to:

1. Prevent your input (i.e. variac) from putting out more voltage that necessary - for example through the use of a mechanical stop

2. Build your bridge with margin to withstand whatever voltage your variac or input line can provide

3. Use a crowbar circuit to discharge all energy storage on the bridge as well as disconnect the bridge from the input line.

Why won't your circuit work?

1. No crowbar circuit. You have energy storage capacitors on your bridge, so if you have an overvoltage condition, those capacitors are already charged up and will damage your bridge (provided its build with such little margin to begin with)

The easiest implementation?

1. Use a double-pole relay. This is how almost all industrial over-voltage protection is employed. When the over-voltage fault is tripped, you just switch the relay off. When the relay turns off, the normally open contact (which is closed during operation) opens up and disconnects your input line to the bridge. The normally closed contact (which is open during operation) simply closes and connects a low-ohm resistor with the bridge to discharge any energy there quickly and safely.

A solid state device could replace the relay, but you would need two, and driving them just adds a few extra components.

dmg

Thu Dec 30 2010, 11:20PM

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

Okay, thank you again EVR, Ill use a relay method instead.

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