On 2013-07-25 5:43 PM, Sean Breheny wrote:
> I would second Russell's suggestion. I once tried to make shunts out of
> copper and found that they accuracy was very poor due to the high tempco =
of
> the resistance of copper. Just the self-heating from a moderate current
> would warm the wire enough to cause an error on the order of 10%

I wish I could get NiCr wire that thick. A lot of the stuff I build is
limited by the availability of parts. For me, most of the point of
building it is to avoid spending $100s or $1000s if I can do it for a
few dollars. If I start needing to spend much more ordering parts (+
shipping), might as well buy an off-the-shelf unit.

I can get copper wire at the hardware store for a cheap price. NiCr
wire, on the other hand, nowhere to be found in any decent thickness.
Paralleling some much thinner gauge and bathing in "transformer" oil for
cooling might work. If not, the 10% (or even 20%) variability in copper
is acceptable. And since the current will be pulsed for very short
periods of time, the oil should be more than enough to keep the copper
cool.

Another option is using a big transformer and a thyristor. No high
current parts needed, energy can be easily controlled and measured with
relatively low current parts.

But honestly, if I managed to get a decent weld with nothing more than a
battery, electrodes, and my own hands, a microcontroller and a MOSFET
array would be more than sufficient for good quality welds.
--=20
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.