I've been looking at these machines in jewelry supply catalogs. They're expensive. And too much money for just my hobby budget. Anybody know how these work? I'm checking on Google today but any help is sure appreciated. Thanks, Eric R Snow
Based on zero firsthand knowledge about CD welding, I'd say you could probably build one fairly inexpensively. I did recently build a transient generator that produces 150-amp pulses for EMI testing.
claims 680,000 uF of capacitance and 110 joule output with voltage up to 18 volts. That's about $100 worth of caps if bought new from Digi-Key, but you might do much better at a surplus outlet. A supply to charge the caps would be dirt simple if you don't need high rep rate.
Switching could be done with MOSFET's or a big SCR. I'd use MOSFETs. IRF1404Z from Int'l Rectifier can handle 100-amp pulses with no problem; they're $1.64 each in lots of 10, and I'd use 10 of 'em in parallel for 1000-amp pulses.
I think lead resistance would provide some current limiting, but I'd still use a series inductance of about 220 uH to limit peak current to
1000 amps or less. That'd probably be an air-core inductor to avoid saturating core material.
I think it would be easier with fat Smart-FETS than with an SCR. Unless you're very lucky, you'll not find a gate turn-off SCR in that kind of capacity you can afford.
And unless you can gate the thing off, you'll have a tricky system to adjust the energy delivered to the work. With gating, you could just vary the pulse width, and keep the peak voltage constant. The IR smart-FETS are nearly indestructible, so the output stage would be easy.
A farad of capacitance at the 18v working voltage wouldn't cost much, for ordinary filter caps, but you'll need units specified for CD duty, or you'll fry the welds (or crimps) to the plates inside the caps at the currents you're asking them to deliver.
It is not quite as simple as just discharging some capacitors. The capacitors should have a low ESR [effective series resistance] or they might blow. That is why "photo flash" caps are made. Older small welders used oil filled caps at higher voltage and discharged them through a pulse transformer. If you ave access to used machinery stores look for Sippican [sp] or Hughes. They were used in the vacuum tube industry and we used them in the "old" days to make modules with discreet semiconductors.
A big SCR works fine, unless you need +/- 1% control of energy due to some standards organization. I build one of these things years ago. You adjust energy with a variable voltage power supply. I used a 100 A SCR, and applied a strong trigger pulse with a switch and R-C network to get the SCR saturated quickly. I used ordinary computer-grade caps, and it worked until the equipment was no longer needed.
If you aren't stupid enough to open them, I can't see any reason to care that isn't 100% pure paranoia.
Same applies to asbestos... "Abating" the so-called threat creates more threat than existed if it had just been left alone... (No, I'm not saying that asbestos is harmless - Just that it's harmless *WHEN LEFT ALONE*)