X-Node inrush current - really large or what?

[pasdesignal]

Is anyone else noticing the huge inrush current on X-Nodes?

I have 7 X-Nodes sharing a 10 Amp power distribution board with a "C" curve miniature circuit breaker, and they trip the circuit breaker if they have been left off for any decent amount of time (> 20 minutes). I'm on nominal 230V power.

The normal current for these devices is <100mA at 230V.

I will endeavor to get an accurate measurement of the inrush current - but that is no simple task at millisecond accuracy. In order to build a system that will withstand power-down situations (in case a generator fails) and reboot fine by itself I need to address the issue.

Just wondering if anyone else is having the same problem or if Telos know of the inrush current figure from their own testing?

[pasdesignal]

I have managed to get some approximate figures of what inrush current the Analogue X-Node has.

I used a Fluke 43B which has a specific inrush measurement function, so its a fairly suitable tool. It had also just been calibrated.

Here is a link to a image of the typical result I was getting - about 40 Amps peak for the initial 5ms after switch on. That's on 230V/50Hz supply.

http://postimg.org/image/y2an493q5/

If someone else could verify this, it would be very useful to get an accurate figure. The Fluke 43B isn't actually that good for such short periods - evidently the 370 series would be more accurate over the millisecond time range. I don't have access to one.

If I have 7 nodes sharing a 10A distribution box, that's approx 300 Amps inrush current on a C curve breaker, within about 5ms. That still shouldn't trip the circuit breaker so I am suspicious that the real inrush current is significantly higher.

[Geoff Fairbairn]

Yes, we made this discovery when testing the power redundancy arrangements in our installation (the xNodes are powered by PoE alone if their main supply or PSU fails). We have two banks of fifteen xNodes each on one 16A C type breaker on a 230V 50Hz supply. Despite the XNodes being powered by PoE the breaker will not latch unless switched on twice rapidly (and the 30kVA UPS logs an output overload).

This would be extremely inconvenient if the xNodes were deployed at a remote location and tripped their breaker on restoration of power after each local outage. The issue is not specific to xNodes or even to Axia of course (we have some other gear with SM-PSUs that draw high in-rush currents so have the same issue when installed in bunches in racks), but whilst the PSU probably isn't an Axia design, it would much appreciated if a version with a controlled switch-on current were sourced for future versions and products.

[twolochs]

A simple cheap fix for this, which maybe should have been included in the original design, is an NTC inrush limiter - a small cheap component that has a high resistance at room temp, but warms up (not to much!) rapidly under steady load and drops very little power in the steady state.

For example, a CL-140 has a 50 ohms resistance at 25C, so would limit the inrush current to no more than 5A or so, but under the steady 100mA load would drop to about 10 ohms, dropping just 1V off the supply and consuming 0.1W.

It occurs to me that these could be fitted into a distribution block as an after-market fix for the problem for anyone with a problem installation, but they do need to be chosen appropriate to the load, so care might be needed that the distribution block wasn't prone to getting reused for other loads! The CL-140 I mentioned, for example, would be best protected by a 1A fuse as a precaution against someone plugging in a kettle!