Tripping supply

[Heptagon]

We are having trouble in the theatre we hire with some distribution unit channels tripping on power up.

There is plenty of capacity for the wattage of the luminaires (the channels are running at less than 1/3 capacity)

The lighting power supply is not daisy chained - it is all fed from a tree of piggy-back plugs.

Is there any electrical advantage in daisy-chaining - e.g. does each luminaire impose some sort of delay on those further down the chain - which could prevent tripping?

 

 

[Ynot]

What lanterns are on the circuits that are tripping?

Discharge luminaires in particular have a high inrush current draw that will most likely exceed the rating of the circuits at startup, and you should be able to arrange that startup sequence so they don't all fire up at the same time.

However, what surprises some people is that LEDs also have a higher than you'd expect inrush on power-up so bear that in mind when planning your power distro

 

[themadhippy]

Quote

tripping on power up.

Sounds like in rush current,power your rig up in stages instead of all at once if possible

Edited August 29, 2023 by themadhippy

[adam2]

"A tree of piggy back plugs" is a bit concerning. Use of these plugs permits of multiple plugs being powered from one wall socket. This is unwise in the case of appliances with substantial running currents such as incandescent theatre lanterns, or other appliances with high inrush currents such as HID or LED lanterns.

[kgallen]

What’s tripping - an MCB or an RCD/RCBO?

MCB/RCBO could be inrush. What’s the curve type? B/C/D/K? B will be very sensitive to inrush.

If an RCD/RCBO, do you have lots of ‘leaky’ LEDs on the one supply?

Edited August 29, 2023 by kgallen

[sleah]

3 hours ago, adam2 said:

"A tree of piggy back plugs" is a bit concerning. Use of these plugs permits of multiple plugs being powered from one wall socket. This is unwise in the case of appliances with substantial running currents such as incandescent theatre lanterns, or other appliances with high inrush currents such as HID or LED lanterns.

To be fair, it's not much worse than daisy chaining. Although much depends on what is protecting the first in line, trip / fuse etc.

Piggy-backing or daisy-chaining won't make any difference to the problem you have as they all come from a single plug.

I'd suggest piggy-backing or daisy-chaining in smaller groups connected to separate power outlets (that are fed from separate trips), and if not possible to connect to separate trips, a staggered switch-on as already suggested.

Edited August 29, 2023 by sleah

[Jivemaster]

Happening at power up says two things to me; first it's likely to be a huge current surge at switch-on and secondly you are powering up with one switch.

Assume that every unit takes a switch on surge of 4 times it's running current for 4 cycles of mains (some will be more some longer) plug them all in then switch them on one at a time. 

See whether the problem is obviously caused by one fitting or type of fitting. Check that you have the suggested fuses and breakers as the manufacturers instructions. 

[Don Allen]

Do your power distros have individual RCBO on each channel or do you have three RCBO for the distro, one per phase.

The most common tripping of power distro channels appears to be the leakage of the LED SMPS. You can get away with about ten 200W LED fixtures, for reliability I recommend six. I have a Metrel MI3309BT PAT tester and it can measure leakage current which is a useful feature that not many users make use of. I have started measuring leakage current on LED fixtures to see how they compare.

A 30mA RCBO must not trip at 15mA leakage but must trip at 30mA, a typical value is 20mA leakage. Leakage current is not stated in most light fixture datasheets.

[Heptagon]

Sorry - again I haven't given enough info.

The lighting is all LED, but the actual units on each circuit are not documented. We don't have access to the luminaires at present. 

The distribution is via 3 LSC APS units (12 channel each but only about half of the channels are used (the bottleneck is the cabling to the patch points - there are very few, but more than enough for the steady load).

Switching on is by a Crestron unit which causes the APSs to power up, and they do the sequencing of the individual channels (but not, of course the lights on a single channel).

Switch off occurs when no DMX signal is detected for a configured time.

The first occurrence was a large number of channels, but I didn't see which ones - they had been reset before I got there.

I have only been able to document one case (out of 3 occurrences), but if there are any more I have asked the ASM to note the channel(s) he has to reset.

We have no control of the set-up, but I was hoping to be able to provide a recommendation.

If the luminaires provided a one cycle delay in the daisy-chain hardware, that would solve the problem. 

The reason this seems possible is that some luminaires have a statement that only luminaires of the same type should be daisy chained.

Thanks again.

[Tomo]

Fixtures that have an "in" and "thru" for their power do not create any delay at all.
It's just a convenient way to connect several luminaires to the same socket, electrically the same as a Y-split extension lead or Aus/NZ piggyback plugs.

As with all extension leads, the plug, socket, internal wiring and fuse (if present) have some maximum rating that must not be exeeded, and it's fairly easy to accidentally put "one too many" on the daisychain - or a stack of plugs for that matter.

Many manufacturers say "No more than N luminaires of the same type may be daisychained" simply because it's an easy and quick way of checking that a run is within the maximum safe loading of that first luminaire in the chain.

As you've guessed, the problem here is 'simply' that there are too many luminaires being switched on at once for the inrush rating of the breaker.
The solution will be to spread them more evenly across the circuits of the LSC APS unit, or remove some fixtures.

The most likely actual cause is that a few of the fixtures have been rearranged or added, increasing the inrush of one of the power-on sequence steps.

It won't trip every time because the actual inrush depends on many factors, including the exact timing of the switch-on, the charge in the capacitors of the loads and of course whether you're in a hurry with a fractious full house.

This is becoming an increasingly common problem these days.
- System designed for six fixtures, someone adds "just one more" and it seems fine, but tomorrow it trips on powerup.

[p.k.roberts]

Those nice chaps at ETC have an explainer video here .....

 

 

[sunray]

One of the issues I've encountered several times is tripping the supply breaker rather than the individual circuit.

For example a 24 channel dimmer pack (I think a Strand Wallrack) powered by a B63A MCB... yeah I know a bit iffy but it ran for years with 500/650W fittings without issue. As fittings were changed to LED, circuits were being reconfigured without the dimmer components and many fittings were being added to each circuit, as the lighting expert said 'I can put 20 on a circuit, they're only 30W each'. They were picking up cheap s/h fitting of all denominations and had something like 50-60 fittings on several circuits and suddenly switching on the local isolator would trip the supply MCB.

Hall owners electricians refused point blank to alter the supply so we added some isolators for individual circuits, users had to switch on main isolator then add the LED's in stages. Access the dimmer pack was restricted to 'key holders' of which the drama group were but any other lights users were not granted access and given a printed sheet with instructions.

That was all about inrush, RCD is another problem.