Loosing Power Supplies

[jcassity]

As a general rule , rectifiers (ie- your power supplies) wired with 480/277 3ph will require an isolated / dedicated N leg along with a ground.

If you only have 4 conductors, you probably just found your problem.

I will assume the power supply tech specs call out 480/277 with N and ground.

If you have a back up generator that kicks in during your power outtage, then the genset also needs to have dedicated N legs as well as ground wired back to the transfer switch.

At your plant, you have the the main power room and in your switch gear where your feeder breakers are, there is a common ground bar.

Here is what you need to do.............
open the unit , or research the factory spec wiring. I highly suggest you observe the pretences on 230v engineering because many german makers are very savy with US wiring but also are polluted with those not really understanding the need of a N and a G leg on the input to rectifiers.

Since you did not mention a back up genset kicking on, we can ignore that option.

Your primary problem is getting the rectifiers to fire back up and as a result of the power loss. The inrush isnt really the issue here but the amount of potential on the rectifier ground upon startup. DC is a totally different ball game and you need to have experienced people who know the science of integrating DC into an 3phase AC infrustructure.

Ill bet you will find that your N and G leg are coupled together somewhere on the input side of the rectifiers close to the source.

Ground is not "GROUND", You can make ground any potential you want. "GROUND" is a REFERENCE VOLTATGE. If your rectifier ref ground needs to be zero =-1% tolerance, then make it so.

in short, you need 5 dedicated wires coming to the rectifiers. the inrush will not effect your rectifiers any longer once you do this.

 

[jcassity]

you dont need any transformer if th equipent is designed for a 480/277input.

Is your site wired 480 WYE or DELTA,,HUGE Difference when you integrate rectifiers.

I will assume you are a 480 Wye from here forward.

1- you dont need any fancy loan testing equipment, just a clamp on meter to read the amperage on the AC input legs and N/G.

In addition, when dealing with rectifiers, it is very important that your side is not cheating with respect to grounding.

I will suspect if you open the specs or look at the wiring diagram on the input, you will see that there is a requirement for a N and a G. they are not the same. Your input to the rectifiers need to have a dedicated N leg to absorbe the excess potential of any phase running low. Your ground needs to be suppled as well.

All 5 conductors should be a continuous run back to a dedicated circuit breaker at the switch gear. You will have your three hot legs landed on your molded case breaker and you will have your N leg bonded to the N bar, and your G leg bonded to the ground bar.

If you see 4 conductors only, you just found your problem.

You never mentioned if a genset kicks on when power goes out but,,,,, if there is one and the genset N legs bonded back at the genset to G , and then brought into the sw gear ground bar, there lies another problem.

I have seen a lot of Previously groomed 480Delta systems converted to 480Y and the N leg is bonded to the G. As time marches on, Rectifiers are blowing up.

The rectifier is likely not the problem, your site wiring most likely is on the input to the rectifiers.

 

[beanland]

Overvoltage

Overvoltage

I suspect that the mains supply is running hot, especially when loads turned off, so that supply exceeds 500V. That power supply may be marginal in design at 500V so POOF. Traisnents from inductive spikes would take out the solid state front end of the supply.

I agree, 480-120V then 120V to DC. Install surge protection and/or transient traps.

 

[charlietuna]

We have run into this on many large chiller plants and also on some large elevator motors. we transformed the source voltage down to 120 volts , then installed a small ($100.00) power conditioner and then fed the control power transformer into the battery back up power conditioner. isolate the problem...

 

[jcassity]

no to be picky or nothing,, but you guys are really over complicating DC power.

Its nothing to get rectifiers to work and we do it all the time all the way up to the 10,000A range. You just have to wire the input to the rectifier according to what the spec calls.

 

[jcassity]

sorry for the double post,, edit button missing.....

Your probably best with a time delay circuit on the 480 in to the rectifiers. this way, all the inrush as already rushed in:wink: plus it might be a cheaper solution.

an isolation xfmr might be a solution as well although it only needs to be a small one, it may not fit in the equipment.

Is your service 480Wye or Delta,, and is it 480/277?
Most 480Wye service drops to a facility can opperate perfect with no N leg when you are only pushing motors , HVAC and the like. Rectifiers however are not so fogiving. If your cabinet input has 3 hot legs and one green ground, you are probably sharing the same green ground as your rectifier input ground source,, such that your ground has current on it as well.

Can the rectifiers be wired for a different input voltage since you said each cabinet has a power mgmt center that drops in sources as low as 120 ect. If thats the case, you have an isolation transform secondary at your service and being such a small rectifier, Im sure you have room on the secondary to power up these rectifiers but this is only an assumption.

I find it odd that your rectifiers modules are only capable of being wired with only one input voltage configuration. 480 only is rather high which gives you the bene of having a low output fuse,, but,, its such as mall sucker, it makes me wonder if it could not also be wired differntly to take advantage of your cabinets existing transformer secondary.

To sum all this up,, ill bet you find that the application of a time delay circuit and a dedicated ground from the rectifier back to the MGB (main ground bar) will solve all your problems. Most any good electrician can hand build a time delay circuit from radio shack parts costing less than 20 bux. If you have a Pneumatic source available nearby, there are pneumatic switchs that can be made up to transition this 480v into the rectifier based on a spring loaded diaphram working a contactor regulated by the amount of air coming in. The amount of air coming in would as as your timer perse' or have an "hour glass" effect. A water well pump switch works on the same principle and most of the time, they already have three poles to wire out. Get yourself a dedicated ground and your in like flynn.

 

[Billy_Bob]

Many people in Europe speak English. Might be able to pick up the phone and speak to the technician who is repairing your power supplies - find out what component(s) is getting fried. Perhaps get recommendations for solving this problem.

Also if they are repairing your units and sending the same units back (as opposed to swapping with another repaired unit), you might be able to see recent solder marks for a particular component.

Do you have a schematic of the power supply?

Don't forget the time difference!