[snakebit95]

I am having a problem powering my new power supply reliably. I am using a 1.5 kVA dual-coil transformer; which was initially tripping my 20A single-pole circuit breaker. I have since upgraded my circuit breaker to a high magnetics version; which is designed for starting motors, large appliances, e.t.c. This resolved the breaker issue. However, now I am blowing in-line 20A time-delay fuses in my controls box. I am considering installing a 20A thermistor directly after the fuse-block to slow the in-rush current. I know the in-rush is the problem, because I have had the system power-up a couple of times without blowing fuses. The transformer coil is HUGE, and I am guessing it is pulling some serious current at start-up to build the magnetic field. Has anybody used a thermistor to limit current on their power supply inputs?? I searched the site, and found very little information.

Thanks in advance,
Wayne

[Al_The_Man]

One simple thing you could try is power the Transformer first, and get over that inrush, then power the output of the bridge, This way you do not get the double whammy of Transformer and Electrolytic inrush.
Al.

[snakebit95]

Hey Al,
Thanks for that suggestion. That is the way I already have it wired. I have the transformer in-front of a contactor, which is activated through the E-Stop reset button. I plan to power the transformer first, and then charge the capacitor by resetting the E-Stop. Just the transformer alone is killing the 20A fuses. I thought about installing a larger fuse........but then what good is the fuse when I am running #12 wire for standard 20A service?

Theoretically, a single 20A service line should be good for roughly 2.4 kW; which is well above the transformer rating. If I can just get past the in-rush. During my research, I found that my transformer could be pulling as much as 4X its rating at start-up; which equates to about 50A......albeit for less than 1 second. Any other ideas??

Thanks again!!

[Al_The_Man]

I would think there is some other issues there, there is no way a 1.5kva transformer should trip a 20amp breaker on its own, unloaded.
Transformer inrush is not 1 sec, more like 1/2 a cycle.
Check the transformer current when running, unloaded.
Al.

[pminmo]

You probably could help yourself by using a zero-crossing switching arangement. http://www.futurlec.com/Relays/SSR80A.shtml while I haven't switched 20a, I suspect it could help. The beauty of zero crossing ssr's is you are always starting at zero crossing, where a mechnical switch you can start at the phase peak or anywhere in the phase cycle, plus the mechanical contacts bounce.

But, I can't say I think it's wise to run your load on a 12ga, 20a circuit. A 30A circuit with 10ga wiring and devices is probably the real solution.

[Haydn]

I used an NTC thermistor in my supply to limit inrush and it seems fine. In my circuit I put the thermistor across a time delay relay (rated 30A), when I power on the thermistor is in the circuit for about 1 sec or less, the thermistor is then removed from the circuit when the relay switches over. If I hit e-stop the thermistor will be cold and will work properly when I restart the system. This can be done using a high wattage resistor as well, I have seen people do this but if the relay fails the resistor stays in the circuit and messes with the supply before eventually going pop or catching fire (unless you use a resistor about the size of a brick).

Here is a pick of the ntc in my controller.

[snakebit95]

Al,
Are you sure only 1/2 a cycle on the in-rush? The couple of times I did get the transformer to power-up without blowing a fuse, I could hear a very faint high-pitched hum that quickly faded-out (maybe 2 seconds long). I am guessing this could be the wires/coils resonating while the field was building? I did check the output when it powered-up, and it checked within a volt or so of the advertised output (40V @ 37.5A)......so I know it works. If it is of any help, I have a Hammond transformer part #182V40. Please let me know what you think.

Phil,
When I started this project, I would have never guessed that I would have needed more than a 20A circuit to run this power-supply. The general calculations just did not show it. Of course, I had no idea that the in-rush would be so large. You know what they say about hind-sight. Going to a 30A circuit creates other problems, such as finding electrical plugs and receptacles that are adequately rated to handle 120V @ 30A.

Haydn,
That is one very nice looking layout!! That is also a great idea on the time-delayed relay. I ordered some thermistors today to try, and may go the same route you did to allow time for the thermistor to cool.....and to also eliminate the slight voltage-drop. I want to see that the thermistor will resolve my problem before I throw any more money at it.

Thanks for everybody's feedback!! I am ready to get this machine cutting chips!!!

[Haydn]

Thanks for the comment Snake, I don't think i'd try and fit it all in a 4U rack case again! Good luck. :wave:

[Al_The_Man]

This is a fix from PowerTronix, I would imagine if you put the relay on the secondary instead, it should provide a bit more delay, it looks like the delay is the pick up time of the relay.
So far I haven't had this problem to worry about, I mainly use Plitron, I don't know wether they have slightly different characteristics.
http://powertronix.com/html/in-rush-protection.html
2 secs. seems a heck of a long time.
Al.

[ViperTX]

SSR's are not the answer.....most are not true zero crossing.....they are basically triacs....so, they will be turning on and off at near every AC crossing cycle.

It is not the transformer that's causing the problem....it is most likely the capacitors on the output.....when discharged they look like almost a dead short....that is the reason that you are sometimes able to power up without any problem....the cap's are partially charged.

[Al_The_Man]

It is not the transformer that's causing the problem....it is most likely the capacitors on the output.....when discharged they look like almost a dead short....that is the reason that you are sometimes able to power up without any problem....the cap's are partially charged.

Post #3 says it is on transformer power up only.:drowning:
Al.

[Haydn]

SSR's are not the answer.....most are not true zero crossing.....they are basically triacs....so, they will be turning on and off at near every AC crossing cycle.

It is not the transformer that's causing the problem....it is most likely the capacitors on the output.....when discharged they look like almost a dead short....that is the reason that you are sometimes able to power up without any problem....the cap's are partially charged.

The relay I used is not an SSR and it is on the secondary, reason being as Viper says, the inrush is almost all caused by the dead short of the filter cap when its discharged. I know inrush is a short lived affair but setting my time delay relay to 1sec isnt a problem for me, switch on, blink and everythings fine (it may be set lower I cant remember).

Looking at the pic it looks like its set to about 0.6s

[Haydn]

Try this link

http://www.bearpwr.com/pdfs/BEAR_Tec...ushCurrent.pdf

[ViperTX]

*chuckle*...Al same post.....the contactor is after the transformer...

snakebit95...are those AC rated cartridge type fuses that are blowing?

[pminmo]

SSR's are not the answer.....most are not true zero crossing.....they are basically triacs....so, they will be turning on and off at near every AC crossing cycle.

A triac is a triac, not an ssr. SSR's typically contain triac's and while I won't argue an SSR doesn't have to be zero crossing turn on, I haven't seen one that isn't, and I've used them for 25 years. The whole intent of SSR's were to cleanup the mess mechanical contacts caused. Anybody that has designed around sensitive measurement enviornments won't hot switch with mechanical contacts. They may isolate with mechanical contacts before or after a power on/off cycle, but not hot switch.

[snakebit95]

*chuckle*...Al same post.....the contactor is after the transformer...

snakebit95...are those AC rated cartridge type fuses that are blowing?

Sorry for the delayed response. My hard-drive developed the "clunk of death", and I had to buy a new one. I had to reload EVERYTHING!!!

Yes, they are AC rated cartridge type fuses (Buss FRN-R-20 to be specific).

[Al_The_Man]

These are 10sec time delay at 500% rated current!
It would be interesting to see the normal off-load current of the Tfmr.
Al.

[snakebit95]

SSR's are not the answer.....most are not true zero crossing.....they are basically triacs....so, they will be turning on and off at near every AC crossing cycle.

It is not the transformer that's causing the problem....it is most likely the capacitors on the output.....when discharged they look like almost a dead short....that is the reason that you are sometimes able to power up without any problem....the cap's are partially charged.

As mentioned earlier, I have a contactor between the transformer and the capacitor. The transformer is the only item that receives power when I turn-on the main power switch; which is when the 20A fuse blows. I wish I had a scope so that I could graph the in-rush current over time to see exactly what is occurring. I am afraid to try it with my cheap volt/ammeter <chuckling> It would probably go-out in a blaze of glory.

[snakebit95]

These are 10sec time delay at 500% rated current!
It would be interesting to see the normal off-load current of the Tfmr.
Al.

Al,
I can tell you that the fuses do not even last 1 second before they die. Do you think it is possible I may have a transformer that has a short-circuit that randomly occurs? This is still beyond my comprehension that the tranny alone would pull so much current at start-up.

[Al_The_Man]

If it lasts any appreciable time, you could put a very low resistor in the primary lead ~.1ohm and measure the AC voltage to obtain the current, this will place no stress on the meter.
A moving coil/analogue meter would have a faster response, if you had one.
The electronic version have a bit of a slow update.
If you have a high end meter like a Fluke etc, they also have a bar graph line below the digits that updates faster than the digital display.
You wouldn't have any kind of mechanical mounting around the outside of the core, by any chance?
Al.