Greetings and a small starter project (ignition coil / Jacobs Ladder) / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Greetings and a small starter project (ignition coil / Jacobs Ladder)

1 2 3 4 next last

Move Thread

LAN_403

Fulmen

Fri May 13 2011, 06:53PM

Registered Member #3883 Joined: Fri May 13 2011, 06:30PM
Location: Norway
Posts: 87

Greetings everyone. Fairly new to HV-stuff, so I thought I'd start small for my first project.

Basically it's an ignition coil w/driver powered by a 12V battery. That should keep power down into the fairly non-lethal range (5-10W seems to be max) and make for a nice mobile show-off gizmo. The driver is a basic astable multivibrator (more educational than a 555) and a 600V MOSFET, this worked fine until i started increasing the spark gap beyond 2-3cm. The addition of the C3 capacitor should keep the primary voltage down, right? I'm fresh out of MOSFETS as of now, so I'll have to order some more before I can finish this.

What I really want is a "one-shot" ladder, meaning it will produce a single traveling arc by a bush of a button and then shut down. I was thinking of using two caps in series for C3, keeping the voltage below arcing threshold and having the switch shorting out one of the caps to initiate the ladder. Am I on the right track with this?

Later on I plan on building a micro tesla, also powered by battery (and probably ignition coils with the same basic driver). And then I was thinking of a 1KW Tesla

magnet18

Fri May 13 2011, 07:41PM

Registered Member #3766 Joined: Sun Mar 20 2011, 05:39AM
Location:

Posts: 624

What fets were you using? Make sure they have a high current rating or you'll blow them faster than you can say metal oxide field effe...

Fulmen

Fri May 13 2011, 07:57PM

Registered Member #3883 Joined: Fri May 13 2011, 06:30PM
Location: Norway
Posts: 87

I used a 10N60 I happened to have, rated at 600V and 5.4A . Should have been more than enough, or am I missing something vital here? The coil primary has a DC resistance of 3.5ohms, so I should be well below the current rating especially since I'm pulsing it a bit to fast to fully charge it (2.5-3A according to the Spice simulation).

magnet18

Fri May 13 2011, 08:19PM

Registered Member #3766 Joined: Sun Mar 20 2011, 05:39AM
Location:

Posts: 624

Hmm, well, personally I'd use some higher current, lower voltage fets such as the irf644 or the irfs654, you also might want to scope your gate signal (do you have a scope?), the transistors might be making more of a sine wave than a square wave, which will keep your mosfet in the grey area between on and off, and they don't like that.
You might consider running that signal through an op-amp like the tl071 or a switching chip like the tc429 to make sure you have a nice square wave, I know the tc429 is good for switching fets, not sure how the tl071 would cope, but I know you can make things sing with it

Fulmen

Fri May 13 2011, 08:47PM

Registered Member #3883 Joined: Fri May 13 2011, 06:30PM
Location: Norway
Posts: 87

No scope (yet), I'm feeling the disadvantage already. As for the FETs I appreciate any pointers as I will have to order new ones anyway. 250V seems a bit low though considering the application, but that depends on how FETs handle pulse loads. I guessed they would handle currents better than voltage, but maybe I was wrong?

As for the signal quality from the vibrator I'm really only guessing as I didn't find a model for the 669's I used there (stop reminding me of how much I need a scope), I'll look into that. A better model (or better transistors, I'll need to order some anyway I guess), and perhaps modify the circuit to improve signal quality if possible (not sure how though).

But as for basic idea of triggering a spark by reducing the C3, do you think it'll work? Once an arc is formed it should be able to sustain itself at a lower voltage, and since the energy is constant it shouldn't affect the output, right?

Ash Small

Sat May 14 2011, 01:10AM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

magnet18 wrote ...

, the transistors might be making more of a sine wave than a square wave, which will keep your mosfet in the grey area between on and off, and they don't like that.

I'll be posting soon on an astable ferroresonant GDT driver that I'm hoping will give an approximate square wave output, but I'm still working on the concept.

I'm sorry I can't contribute more at this point.

magnet18

Sat May 14 2011, 02:47AM

Registered Member #3766 Joined: Sun Mar 20 2011, 05:39AM
Location:

Posts: 624

Right, I'm guessing that your transistors are probably the problem, since they aren't linear. Do you know how an op-amp works? you might also be able to make a 555 wired as a Schmitt trigger work, as long as you don't mind inverting the signal. (theres no reason it would really meas with things)
and the other chip I mentioned basically takes the weak signal and makes it a stronger signal of the same voltage (more current behind it). Input pin, output pin, and supply pins. Look up the datasheet, it can tell you more than I.

I'm not sure what c3 would do, other than protect the chip from the voltage spikes, and it's a bit big for that, 10nF would work better. The bigger the cap the more you start sapping the switching.
Considering that you're switching 12V, 250V should be more than fine. Unless ignition coils make bigger interference spikes than flybacks, I doubt that, but cant say for sure. Beyond using one as a brief stepping stone into plasma-speaker land, I haven't messed with ignition coils.

You might be able to get that idea to work, but I would use something different, like I said, sapping at your signal isn't good for any component.

Fulmen

Sat May 14 2011, 10:01AM

Registered Member #3883 Joined: Fri May 13 2011, 06:30PM
Location: Norway
Posts: 87

I'm not sure I agree with you here. I found a (supposedly) better model for the 669's, didn't change the output in spice. The simulation looks just fine, and the circuit worked just fine until I tried pulling arks longer than ~3cm.

1305366708 3883 FT115450 Coil Driver Output

I don't understand how an opamp should improve this, I'm already feeding the FET a 12V signal. If I understand FETs correctly the high Gate-resistance should make them almost purely voltage-driven. However, you made me check out a few things, and to my surprise the simulation shows an reverse current at the gate. Don't know if this is real or not (hard to simulate such things properly), but it looks like something that could do damage.

BTW, don't forget that the FET sees a LOT more than 12V as it breaks the primary circuit. As you interrupt the current through an inductor the voltage will increase to keep the current flowing, this can produce hundreds of volts (even thousands if it's interrupted fast enough). The C3 effectively reduces primary voltage by increasing the discharge time, just as it was done in old ignition systems.

magnet18

Sat May 14 2011, 01:02PM

Registered Member #3766 Joined: Sun Mar 20 2011, 05:39AM
Location:

Posts: 624

That was just mt two cents, take it or leave it.

Steve Conner

Sat May 14 2011, 01:24PM

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

The advice to use low voltage, high current FETs is bad. The ignition coil driver circuit needs quite a high voltage device to work: if you want high voltage on the output, a pretty big spike has to be reflected into the primary by transformer action, and the transistor has to stand that without breaking down. 600V is probably about right.

Switching C3 is probably a bad idea too. As anyone who worked on old mechanical car ignition knows, there is an optimum value for it. If you go far away enough to make it stop sparking, you may blow the MOSFET. It would be better to just switch the power supply to the circuit.

Well done on getting it to work with a transistor multivibrator, that is more fun than the usual 555 circuit.

1 2 3 4 next last

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