Backfeed relay resistance calculation in Wiring Unlimited

I’m debugging an sporadic imbalance issue within a parallel system and had a question on something within the Wiring Unlimited docs: 6. AC wiring

The voltage drop tolerance of a 100A backfeed contactor is about 20mV at 100 A. Therefore, the total cable resistance (input + output) should be larger than R = 60mV/100A = 0.6mΩ.

The first part of the sentence says voltage drop across the backfeed relay is 20mV, but the ohms law formula uses 60mV. Is this accurate? If so, I’d appreciate if someone could explain.

EDIT: For greater context, the AC wiring to/from each unit in the system has ~3.5milliohms of resistance, which I believe satisfies the recommendation in this section (~21 feet of #2 awg copper wire)

However, when testing with ~40A of AC current flowing through the parallel system (which should be ~20A per unit), we noticed the voltage drop across individual AC circuit breakers ranged between 25-200mV (protecting the AC input/output cabling of the 10k Quattros)

Doing a simple calc of 200mV/20A = 10milliohms – which mean the greatest point of resistance is the AC breakers (and could explain the imbalance, if I understand correctly). We are now debating switching to 100A fuses, rather than 100A circuit breakers.

AC breakers often have higher and inconsistent contact resistance compared to fuses. So switching to fuses could reduce resistance and improve balance, assuming you’re okay with the lack of reset functionality.

You are right, there is probably an imbalance because of them.

Do you have 100A breakers feeding each individual, or 100A main and then 50A/60A to each individual?

Here the throughput on ac input and output must be greater than that. (Not just through the feedback relay which is one point in the system)
So the simplified use of the figure is because of that.
The imbalance can come from either side or a combination of both.

Thanks @lxonline. This is a 120/240V split phase system, with a 200A mains feed (200A per 120V “leg”). We have a 2x2 configuration of 10k 120v Quattros (4 units, 2 sets in parallel). Each unit is protected with 100A breakers on the AC input/output.

Ah, I think I understand. So in our case, that recommendation is still satisfied regardless, in terms of the cabling (the 23 feet I mentioned above is split as 11.5 feet on input and 11.5 feet on output).

I’d prefer something resettable but I can’t see how to accomplish that in this particular scenario.

I’ve proven it is possible to run the system close to “perfectly” balanced for weeks at a time, and spent weeks trying to understand what causes the system to go out of balance.

This week I noticed knocking/flipping the AC breakers in just the right way does affect the balance significantly (which I’m guessing slightly adjusts the internal contactors).

It’s also possible that I measured the voltage drop at the wrong time – and the imbalance is caused by resistance somewhere else in the system (e.g., so more current was flowing through a given breaker due to imbalance somewhere else, causing a higher VD).

However, the fact that we measured a drop of up to 200mV across any breaker – even if it was 30A flowing through one breaker and 10A across the other – would imply the breakers are still the highest point of resistance and worth attempting to switch out to fuses so the cabling ends up being the highest point of resistance. Does my logic seem correct here?

Curious if you have any other thoughts.

I definitely agree that an ac breaker can cause issues in parallel set ups. And it gets worse if there has been high current breaking.

It would in higher current application be easier to balance (since it not variable contact) fusing since at 100A as now you do need to spend more on really good quality switchgear otherwise. Or switching to the big MCCB as small contact points are less of an issue? They are big and ugly - i do not know what kind of space or installation circumstances you have. So that part is really up to you as the installer.

Always a cause for concern as you don’t want unexpected thermal tripping.

Thanks again @lxonline. Yes, I agree (we tried to find some that would fit). We don’t unfortunately don’t have too much room (photo of AC cabinet below). The heavy guage wiring also makes things fun to work on.

Currently using these breakers: https://www.se.com/us/en/product/QOU2100/mini-circuit-breaker-qou-100a-2-pole-120-240-vac-10ka/?range=7230-qou-unit-mount-circuit-breakers&selectedNodeId=12367668540

It’s a bit challenging to find din-rail mountable 100A fuses, but did manage to find these:
https://www.eaton.com/us/en-us/catalog/fuses-and-fuse-holders/bussmann-series-class-cf-fuse-blocks-and-holders.html

I’m not as well versed in fuses… we noticed they make two versions. A fast blow vs dual-element time delay. Do you have any thoughts on which we should use with the Quattros?
https://www.eaton.com/us/en-us/skuPage.TCF100.html
https://www.eaton.com/us/en-us/skuPage.FCF100RN.html

I can order some today and follow up on this thread with the outcome in a ~week.

I have some experience but never used eaton.

What is your incomer size?
If you need it to be at 100A then fast but if there is a bit of wiggle room the other will be ok to use the other type, your cable can do it.

I have used mersen and some chnt. Mersen do a cylindrical fuse (22x58) with din mount holders. Or the blade ones. But it also may be a tight fit with the blade

What do your wiring standards ask for though? Regulations also matter.

Thanks! I ended up ordering the “fast” one as they matched the trip curve of the breaker a bit closer. Will report back here once all installed in a week or so.

I’ve had 3 fuse holders melt and almost catch fire. All eaton. Won’t touch their stuff now.

The same resistance for safety automation is really a headache. I solve it with one simple method, sorting. I order a large batch of circuit breakers, then I make a selection, everything that does not suit me I return to the supplier. For example, for this parallel installation I had to measure 50 pieces of circuit breakers to select 8.

Oi. Not a great experience.