NigelL

This is most likely caused by the grid voltage dropping below the normal working threshold for the UPS. 
The voltage on one of my phases often drops below 200V in the early evening.  If it drops below 195V, my Multigrid-II disconnects from the grid and runs off the battery. This usually only lasts for a few minutes and then goes back to normal. It does make a noticable clicking noise when it recovers!
Getting back to TTT's problem, it would be useful to know if the grid voltage was outside a normal range when his fault occurred,

Hi VisN, have you verified that the correct Voltage settings have been programmed using VEConfigure? What is the "DC input low shut-down" voltage setting? This can also be affected by the ESS "Dynamic Cut-off" feature if you have enabled this option.

In most of the new Victron MPPTs this is the case. In some of the older ones the current sensor was in the negative line. Because this is a hall-effect sensor you won't see it by measuring it with a DMM, but if you earthed the panels and the battery on the negative side you would bypass the current sensor and blow up the MPPT. I've seen it first hand too. But that was the old ones... the new ones have current sense in the positive line.

I printed this onto an A4 self-adhesive label sheet, so one just peels off the backing and it's ready to stick !  The lamination bonds to both the printed-side of the label and the peel-off backing. So one discards the back half of the label-lamination "sandwich" which exposes the sticky backing of the self-adhesive label.

I have found a fairly simple and effective way of making custom labels - e.g. for marking/labelling Distribution Board. See example below

 
I designed the labels in Visio and then had it colour-printed onto an A4 self-adhesive paper label (glossy type).
I then laminated the whole printed label, including the peel-off label-backing. The lamination protects the label and gives it a hard-wearing finish.  
The printing cost about R11/page and the laminating about R13/page at my local "Wizards" print/copy shop. 
 

I'll be happy to grab it from you! What a great cost! Do PM me your bank details good man!
Jay

As soon as load shedding hits, that thing is going to fly like it has wings... 🙂

The costs jump up a bit more when you have a 3-phase supply 😬!  Unfortunately I did not get everything at the same time otherwise I could have arranged a bulk-discount...

I wonder if the surge-protection circuit in the APC UPS may be causing/contributing to this problem?  I would try connecting the UPS loads directly to AC-OUT1 and see if this changes anything. Then repeat this test with APC UPS completely disconnected from AC-OUT1.

I wouldn't lose sleep about it.
Weather storms on our own planet reek more havoc than this far more regularly.
And once you're used to Stage 4 load-shedding, just think of this as a Tuesday.
 
 
 

I guess that I have just re-discovered one of the immutable laws of the universe, that applies to many different fields.
Actual_DB_Space_Required = Planned_DB_Space + n (where n >= 1). 😃  

Agreed! I had an Efergy unit installed next to my "proper" Landis & Gyr meter so could easily compare values - they were typically 10% out, but sometimes up to 20% different.
The effergy also cannot compensate for the Power Factor of the load, so will typically give a larger power reading than actual.
They are still usefull to get a ball-park idea of power use and specifically power use at different times of day.

Hi Jaco - thanks for confirming!
Famous last words 🙂

A common method is to join the wires using a barrel-type uninsulated ferrule. I am not an Electrician, so the following are just observations and may not be the proper/legal way to do this 😉 
Notes:
One must select a Ferrule that is just big enough to fit all of the wires (it should fit snugly before crimping). For crimping multiple wires together, insert all wires from the same end of the Ferrule, with the stripped section going all the way through the barrel. Crimp the ferrule on both sides of the middle notch with a proper crimping tool. Double check that all wires are properly secured and cannot move or be pulled out of the ferrule (give all of them a good tug) Cover the ferrule with a few layers of insulation tape.  This connection must be done inside a connector box with a separate compression gland for each cable.

If you are combining two wires a much cleaner method is to use a 2 wire ferrule like the one below. But it only recommended if both wires are the same size.


 

Hi Chris
I have a similar setup but have not experienced this problem.  I think this is due to having a one-way valve installed on the cold-water inlet to the Geyser.
The circulation pump gets the cooler water from the "bottom" of the geyser and pumps it through the collector and then back into the geyser - so should only fail if the geyser is almost empty?  It may also be worth checking that the mesh strainer/filter is not partly blocked - my filter needs to be cleaned every few months.

Great service, thanks for the info. They made it possible for me to buy IC's thats not even in SA yet at very good prices. 

I often buy components from Digikey - through their local authorised agents in Cape Town. They bring in a consignment once or twice a week, so you do not have to purchase any minimum order value.  Speak to Mark or Edith Smith at Nuvision.
Edith Smith
Internal Sales Administrator
NuVision Electronics (Pty) Ltd.
Tel: +27 21 554 0038
Email: [email protected]
 

I often buy components from Digikey - through their local authorised agents in Cape Town. They bring in a consignment once or twice a week, so you do not have to purchase any minimum order value.  Speak to Mark or Edith Smith at Nuvision.
Edith Smith
Internal Sales Administrator
NuVision Electronics (Pty) Ltd.
Tel: +27 21 554 0038
Email: [email protected]
 

See attached sketch showing what I think is happening in @Javi Martínez paulty panel.  A significant portion of the current generated by the good string is being shunted by the string with the shorted bypass diode.

See attached sketch showing what I think is happening in @Javi Martínez paulty panel.  A significant portion of the current generated by the good string is being shunted by the string with the shorted bypass diode.

Of course - silly question!
This topic reminded me of a persistent earth-leakage problem I had many years ago on a plugs circuit. Everything still tripped when all appliances were physically unplugged. I eventually found a very small and very dead gecko that had found it's way into one of the recessed plug mounting boxes.

That is a tricky one to pin down!
As you know, an RCD will trip if there is a defined imbalance between the currents in Live and Neutral (in your case 17mA). Although this imbalance is typically when Live "leaks" to earth, one can also get a trip where there is a current flow between Neutral and Ground. Neutral is normally at the same potential as ground (due to the Neutral-Earth link at the main breaker) - but I understand that the Multigrid's Neutral output can "float" if the chassis is not properly connected to earth. The internal relay connects the Neutral output to Chassis during a mains power failure (after disconnecting the Neutral input).
Basically I am wondering if your nuisance tripping may not be related to a Neutral-Earth problem rather than a Live-Earth leakage problem?