For the testing of circuit breakers, we need to generate high current pulses of several cycles at up to 10x or 12x the continuous rating of the breaker, some of which are as large as 6,000 amperes. So we need at least
60,000 amperes into the breaker, which has a measured impedance of something like 100-150 uOhms, which means an applied voltage of at least 6-9 volts. We have built transformers that produce about 10-12 VAC and rated at 4000 to 6000 amps continuous, and their internal impedance has varied from about 90 to 150 uOhms, depending on design. So at best, a transformer that can produce 12 VAC, with 90 uOhms impedance, could push 63,000 amps into a 100 uOhm breaker, but only 57,000 into the 150 uOhm breaker. And most of the transformers are not that good. But recently I measured the internal impedance of a 3000 amp transformer, with about 10.5 VAC output, and it was about 70 uOhms.This transformer consists of three separate E-I cores, 9" high and 12.5" wide, two of which are 4.75" thick, and one is 2.5" thick. The entire transformer is 19.5" long, with two side-by-side 3/8" x 4" bus bars bent in a loop 4.62" apart. The smaller section is connected to a continuously variable source of 0-520 VAC, and the second and third section primaries are either shorted or connected to 480 VAC in such a way that 5 discrete overlapping ranges of output are produced. We connected a solid busbar across the output, and produced 2500 amps at about 20% of the adjustable source. The voltage across the output of the transformer was 0.207 VAC into this load, and 0.380 VAC at the same setting with the load removed, so I determined the internal impedance as (0.38-0.207)/2500 or 69 uOhms. The load impedance would be 0.207/2500 or 83 uOhms. So the maximum possible output current would be 10.5/(83+69) = 69 kA, and into a 150 uOhm breaker it would be about 50 kA.
There are larger transformers that are made with heavier busbar and more sections. But the best I have seen had about 90 uOhm impedance and maybe 12 VAC output, so it would not do much better, although the larger size is necessary for long-time overload testing at 3x rated current.
We were going to try a toroidal design that looked promising, but the project was abandoned because of management issues. This design used two pieces of 1/2" x 2" bus through the holes of two sets of toroidal cores placed in horizontal stacks. But the toroidal geometry meant that the bus bars had to be separated by about 7", and it seems that the wider separation increases inductance and hence impedance. The ideal transformer appears to be wide and low, and not excessively long.
We have now received some transformers made with C-cores, with approximately the same shape as the one described above, but the bus separation is 5.12", and the overall length is about the same, yet this one has only one small section and one large section. There is a considerably longer extension of the bus bars out of the end of the transformer to the connection to the load. So we get an internal impedance of about 70 uOhms, but only a 3.75 VAC maximum output. We actually use it as a 2000 ampere transformer by using a series connection which provides 7.5 VAC but at 280 uOhms, so we would expect about 30,000 amps into a short. We plan to add a second section which will add three more large cores and be used in parallel for 4000 amps continuous output.
These C-cores are constructed in the usual way, with a single winding on a spool around the double wide portion in the middle leg, and then the bus bar is wrapped around it, which gives a minimum spacing between the bus bars of about 5". I wanted to try using two smaller primary windings, one on the top C-core, and one on the bottom, and then wrap the secondary bus as close as possible around the middle section, which would allow a spacing of about 3.75". I think this would make an improvement, with an impedance about 75% of that with the present design. The overall size would be larger, but there is room, and this design might also provide more space for air cooling of the bus, which gets quite hot in the present design. I think I have seen some transformers wound this way, but none with a single turn of bus for high current. Transformers are not my strong point, so I would like some opinions by those who are better at magnetics theory and applications.
Thanks,
Paul