I have constructed a spot welder from a very large capacitor charged by a car battery, that capacitor is shorted across a pair of copper welding tips to fuse nickel strips to 18650 cells. I am currently controlling that shorting by means of a switch formed of 8 IRF1404PBF, they each can handle a hundred amps, so between 8 of them I'm able to dump a few hundred amps from the capacitor for a few tenths of a second.
I switch on the IRF1404PBF, by means of separate 12V battery which is operated by a foot switch and is applied to the fets thus dropping their resistance to tiny amounts and allowing the violent welding current for a moment.
The issue is that I can't control how long the short is allowed to occur for very accurately. With a momentary tap on the foot pedal I can see (on oscilloscope) anywhere between a couple of microseconds to half a second and the weld goes from nothing, to good, to vaporized in that time!
What would be great would be to trigger an arduino with the foot switch and then some code respond, hold an output high for a controlled number of microseconds and thus 'turn on' the fets and pulse the welding current in a controlled way.
Trouble is I need about 12V to do this. Would a non-inverting opamp circuit work to step the arduino output up to the required voltage?
I drew a circuit diagram and attached, I think. Kind of. Apologies - its probably nonsense!!
Top is working if inconsistent welding circuit. Foot switch driven by small battery puts 10V via a voltage divider across the GS of the Fet and hence allows the capacitor to discharge briefly across the welding tips.
Bottom is me suggesting replacing the foot switch and voltage divider with: (an unlabeled for some reason - sorry) arduino. The foot switch informs an i/o pin on the arduino (set as input with pullup turned on) that I require a pulse. Code would then set the i/o pin on the right high at 5v and the op amp amplify to 12V which is wired to the gate (and source).
Does that sound stupid / insane / foolish?
If not any recommendations on op amp chip?
Apologies if it seems very naive and if my diagram is unintelligible.
Opamp is possible but complex and not ideal can't output a lot of current). Because you want rapid turn on/off I would just go for a mosfet driver IC.
And about the multiple mosfets, do you use balancing resistors for them? Otherwise there is a change the current will just pick the mosfet with a tiny bit less Ron to do 90% of the work. Although a single IRF1404 can do 800A pulsed already
Thanks for the incredibly fast reply! I'm reading up on both mosfet driver ICs & on balancing the fets, although as you say they are capable of handling a lot and don't seem to have any problem at the moment, however I don't really want them failing after a few weeks of use either.
Tom, thanks, but I've already made the majority of this, and there will be only one, so I'm not looking to make the cheapest most repeatable thing, I'm trying to solve a problem with bits I know. That timing chip is probably an ideal way to provide a pulse of known width in a high volume production board, but I don't need to save money, I need to make just one thing work. I have lots of arduinos, I know enough to do simple things with them, and I can write software just fine.
Just literally working out how to setup that LM555 chip, how to then interface it to the mostfets, how to sit there and easily adjust the timing over and over again while testing welds etc. All of that would take some figuring out, wheras with an arduino the only unknown to me is how to get it to drive the mosfets.
Thanks for the suggestion, but you see whilst it might be ultra elegant, it wouldn't actually be good for me 'cos I just want to get something done. If you were to tell me that the LM555 can be adjusted on the fly easily and directly drive power mosfets, that might be different...
You're just kidding yourself if you think you can build a proper gate driver circuit from parts on hand, especially op amps. You've said the cost isn't as important as making it work so spend a Buck/Euro/whatever and buy a proper gate driver IC.
Further on down the road, perhaps you'll learn about ESR and abandon the capacitor and just use the car battery since delivery of current is all about the lowest possible resistance.
Oh, yeah, almost forgot to mention, this has been done before, more than once. Here's one with an Arduino Nano controller. Take note he's on his third or forth revision, still refining. The point being, you will not get it right the first time. Learn from the mistakes of others and don't repeat their failures.
avr_fred:
You're just kidding yourself if you think you can build a proper gate driver circuit from parts on hand, especially op amps. You've said the cost isn't as important as making it work so spend a Buck/Euro/whatever and buy a proper gate driver IC.
Thanks for the rather curtly expressed advice!! A rather friendlier chap suggested that and I've spent a little while researching and have ordered one Can I suggest a tad less coffee, or a bit more, or maybe a bun with some icing or the like? I find old episodes of Frasier cheer me up no end.
I'm pretty much facing the same challenge. I'm wondering... why not use a relay to send the voltage pulse that turns the mosfets on? That would be pretty easy to control with an arduino. Unless I'm missing something crucial? The relay would be switching a very low current so arcing wouldn't be an issue. Again, I think...
