Yes, that all rings true and I've had similar experiences with large transformers. The answer, for transformers built into equipment, is a soft-start circuit with a beefy series wire-wound resistor, shorted out by a relay or contactor after a suitable time delay - or a series NTC thermistor if the transformer's going to see a continuous load. For transformers in fixed installations the problem can usually be averted by suitable choice of the overcurrent protection - Type C or D MCBs or appropriate fuses, such as BS 88 gM motor fuses.
I find my 3kVA yellow tool transformer will take out a 20A MCB about 30% of the time on switch on. Never done it on a 32A one though. (that's not a torride though!)
I did some work on a medium sized (10kW) broadcast amp once that used this trick on its HT transformer[1]. The switch on sequence in that was done in software - primary "step-start" contactor was closed first, and then 200ms later the "start" contactor was closed to short out the series resistance (and prevent it going puff!).
[1] This had mains on the primary and stepped that up to several kv on the output. It was about 3' tall and had wheels on it. I seem to recall it took a ramp and two people to get it into the case! In fact, IIRC there is a piccie here:
formatting link
transformer being the brown bit, bottom right. To give you some scale the whole cabinet is about 6'6" at the tallest bit.
Yep. Unfortunately, the only control we had over the over-current protection was to put a warning in the customer documentation, so we knew that would come to nothing. I tried various sizes of thermistor first, but they took too long to cool down, so the breaker still tripped if the transformer was switched off for a few seconds. I ended up with the big timed resistor option but it was too late to cook up a suitable timer and we had to buy pay-through-the-nose timing relays. Not too elegant but at least none of the support calls had anything to do with tripped MCBs.
Apparently not if you post from Gurgle Gropes, I know that gmail allows you to create various aliases, presumably it's picked up a temporary one and is using it permanently?
Seems so. I looked at all the news access settings when it happened, but doesnt seem I can change it. I dont use gmail. Ah well, not the end of the world.
I think this is only accounting for half the potential toroidal inrush current.
The problem is that the steel used has very high hysteresis (probably an unwanted side effect of finding steel strip with suitable properties to wind into a toroid shape), and that there's not even a micro air-gap in the core. The high hysteresis means that you'll most often have high residual mangetism left in the core from last switch-off (assuming it was sudden and not dimmed down). When you apply voltage, and current starts to flow, the core now very likely saturates which would result in the current flow you mention above. However, the residual field also collapses generating an EMF from the transformer, which can add to the mains voltage. If it collapses at the same rate of change as normal field changes, I think you have a worse case of almost 240V * 2 across the primary resistance. However, I think why the rate of collapse is related to the normal rate of change of field in normal operation, so it might be even worse if the collapse is faster.
30 years ago, I fully understood this, having had similar problems with a toroidal transformer in a project I was building, and I went to ask one of our physics professors. However, I really can't recall all the details now. The other thing is that it's not a quarter or half cycle surge, but can take several cycles to get the hysteresis curve in the steel core back centred on the graph axis cross-hairs, so the primary current draw can be high for some cycles, but dropping back to expected levels.
I have one set of 10 x 10W halogens on a 100W toroidal transformer on my lighting circuit. I built that with an NTC thermistor to limit the inrush current, and it never tripped the 6A Type 2 MCB I used to have (nor the Type C I now have). I didn't try it without the thermistor. If you do this, bare in mind the thermistor runs hot, and they can explode if there's a fault. Mine is in a metal box for these reasons.
Other things that might help in theory (but I haven't tried): Use a dimmer with soft start (or better still, both soft start and soft off), although you don't want to run it dimmed except during on and off cycles. Dimmer might not survive driving the transformer though. Degauss the core before switchon. Leaking a tiny AC current through the primary whilst switched off would do this, but safety concerns about having a circuit which is supposed to be switched off still slightly live probably rules this out. Still thinking on this one, not 100% sure if it will help, but switching the transformer off when it has no load may make this worse. So if you can get it to come on once without tripping, and you ensure it always has a load connected, you might find the problem goes away, but this is really guessing. Would be interested to know if you can prove/disprove it though. Note that a zero crossing switch will not help.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.