Earlier Self quote: pulled 14A again with a very degraded (empty) crucible. So I'm back in production with the 1,000W unit.

Figured I would post a bit more technical detail while I'm thinking about it; and can test things again. Here is a look at my collection of new and degraded crucibles. I get at least 20 full heating cycles (>2,000F) out of graphite crucible with a ceramic fiber insulation mostly insulating things (seems like the more the better). Those heats take between 1.5 minutes and 3 minutes depending on how used the crucible is. I can get closer to 50 total heating/cooling cycles (before the crucible turns to dust) if those are split between full melt temperature and lower temperature annealing cycles. I've read that a paint on ceramic coating might help prolong the life of off the shelf pure graphite crucibles.

The 2 oz. crucibles are on the right; and a selection of cut 40mm (largest), 6 oz. and 3 oz. on the left. Due to wall thickness and density differences; the cut 40mm, cut 6 oz. and uncut 3 oz. all weigh between 25-30 grams; while the 2 oz. is only 13.5 grams (when new).

This particular setup draws around 19A with a 30 gram graphite crucible partially inside a 3 three turn 2 layer 45mm I.D. coil. I could likely draw more current if I fully submerged the crucible; but I'm already near the upper limit of a 1,000W unit; so I tune current draw by moving the crucible up and down. I have tried an uncut 40mm x 40mm crucible at 44 grams; and it draws far too much current this way.

While these are small; the whole range of crucibles all used 20Wh of energy to hit the >2,000F metric. It's a bit less than 20Wh with a warm graphite crucible; and can be a hair more with a larger load of metal. While the lighter 13.5 gram 2 oz. crucible draws less current; it still required 20Wh to hit melting temperature. Clearly data isn't going to hold up with larger melts; but it's a starting point in terms of my observations this far. I rarely go over an ounce in terms of metal weight with this setup; so that is very much negligible in terms of heating power/time for me.

I'm curious to see what I can do with kilogram sized crucibles; but a quick check indicates that a 1kg crucibles weights 140-160 grams, 2kg 240-330? grams, 3kg 340-360 grams, 4kg 447 grams, 5 kg 690 grams. Those were the best ranges I could find in terms of weights from product listings at least...

Unfortunately I can't tell what frequency my current furnace is operating at; or what my coil inductance is. The LC100 meter I bought years ago arrived broken (displays an error when attempting to measure low range inductance). Waiting on a replacement to show up now that I actually need one. Still need to build a frequency counter display when I get around to it as well.

Obviously Pete's work has helped me a quite a bit in getting this far; thank you very much (again) for all that. I'll try to condense some of data collection to post up as I have time. Here is a quick look at a melt cycle for 2.900 grams of non-ferrous (group 11). Took only 15Wh, and 1:35 seconds even in a heavily degraded crucible. That old wiring might have hobbled some of my early data (too much energy usage for same work); but even so, breaking 2,000F in 1 min 30 seconds is totally decent either way in my opinion. Here is a look at the meter pre-melt (from cold temp), cold furnace; first start wattage/current, then furnace color and final Wh count.

Just before my first near-catastrophic failure. There was a bright light in the wiring of the induction coil feed cable. Should have known better with those fat copper tails hanging around on top of the bad connection. Fact is; I am not sure what to use for a PCB connector with this type of 8 ga. stranded wire. This wired setup does help to isolate the coil/work heat from the PCB. Unfortunately those hookup parts that came with this 1000W unit were always junk.

Been running this for quite some time... At 24v briefly, then 48v; with a factory new chopped off 20mm (height) x 40mm (diameter) dense graphite crucible floating in the coil; it can draw in excess of 18A at 48v. As the crucible degrades it drops down as low as 10A. Still getting little more than 20Wh consumption when heating the larger crucible to melting temperature (though it's quite short and small in volume).

Tough to see; but the HAZ on those copper strands is widely varied. Some of them were glowing red hot after a very short heating session. I had to cut it off at 7Wh when I smelled burning. Time to refactor; not sad that I got a lot of milage out of this already (7,500 Wh and more than a dozen 20-40mm diameter crucibles).