EBME Forums Forums Maintenance Technical Assistance Isolated power systems in Group 2 locations

Any transformer will experience a high inrush current at switch on and the bigger the transformer the bigger the inrush. Almost all isolated supply will be protected by an MCB with a type B rating. Ideally, if large inductive loads are being used, a type C rating would be expected.

The problem is even if the transformer is fitted with in-rush protection this is usually via a NTC (high resistance when cold and low resistance when hot). So once the unit is powered on the NTC has a low resistance and is effectively bypassed. If now the equipment is quickly turned off and on then the in-rush protector resistance will be low so a high current will occur. This can happen when plugging equipment in or out, as the contacts in the plug may make and break rapidly. Remember it can take a few minutes for the NTC to cool fully and restore its effectiveness.

The same applies for most switch mode power supplies so these can also be a new source of problems!

Sounds like a good enough reason for a GFCI at every isolated outlet.

Although I'm now wondering if the real answer isn't RCBO:- "Residual-current Circuit Breaker with Overcurrent protection"*. I'm not sure if these are available for single circuits (individual outlets), however.


I wonder what became of Type A?

1) Source

2) Many others are available.

* An RCBO is a combination of an MCB and RCD and provides both overcurrent protection and earth fault current protection in a a single device.

Things are getting (even more) complicated now Mike!

OK, the in-rush limiter is basically a NTC thermistor. Where does it actually sit in the overall circuit (of the type we're discussing)?

Do we now need yet another component ... an "in-rush limiter cooler"? How about using a Peltier device? ... I'll get on it straight away! Ha, ha.

Maybe in the future we'll be looking at real-time monitoring of individual supply circuits (outlets) - via computer algorithms I guess - with all sorts of protection device options set at various tripping levels (and response times) depending upon specific applications (and, set up by program variables - or maybe just good old DIP switches)! Expensive though (and let's hope they also remember to include fuses as the line of last resort)!

By the way, "in-rush" may well have been the problem here (as well as in many other mysterious - and often "one-off" - faults on high-current (switching) equipment).

Clearly any NTC in this case sits at the input (primary) of the mains isolation transformer (or often any large transformer). It is also possible to perform this function by using a resistor that is bypassed by a relay contact after a time delay. It is also possible to have more sophisticated electronic devices that reduce transformer in-rush.

Please note that it is not allowed to install any such device in the primary of a medical IT system (IPS) transformer!

As for complex monitoring, there are moves by the manufacturers to try and make such individual supply monitoring compulsory for medical locations! The question no one can answer is what is a "normal" monitored level and what is abnormal? If a socket-outlet one day has a residual current of 1mA and the next 2mA is this dangerous or normal? Will such equipment enhance safety or simply drive up costs? Already we have lots of additional IPS units that may be doing nothing but often the infrastructure behind these units is not all it should be. For example I know of cases where all the hospital power went off for hours, even the IPS supplies! The doctors carried on via torchlight!

Good points; and good questions!

Maybe it would be a wiser course to have any complex monitoring incorporated (where need be) into the mains input side of the individual items of equipment themselves.

On your last point:- sounds like (yet) another case of lack of maintenance (which should include regular testing) of the standby generator(s).