• Have you tried out dark mode?! Scroll to the bottom of any page to find a sun or moon icon to turn dark mode on or off!

diy solar

diy solar

What is the best combination for diy battery protection, class T fuse/battery disconnect/dc beaker?

Read this about parallel batteries - the resistance of the copper in the busbars throws things off.

The point of 2 pole breakers is that both poles are switched at the same time.

@timselectric is the expert for all things NEC

 
Last edited:
I choose a simple diagram where the total resistance for each battery pack is the same on purpose, so it is easier to explain what is the factor left out that causes the imbalance problem OffGridForGood was pointing out.

The resistance is not the same as you posted it.

The thread discussing this was linked, read it.
 
Correct - if you have 4 equal IR & Capacity batteries, with equal length and ga cables.

Why do you think the 4 pack cabinet of mine has the 4/0 attached at the 2 points mentioned? :ROFLMAO:

I ran 20 months this way, never a problem with pack balance. The large busbars don't hurt any but each cable from cells to busbar is the same resistance too.

The 4 packs are not equal IR or capacity. Watch the video, 280N's on top 2 shelves, half 280K (B grade) and half 280N on third shelf and bottom shelf is Grade B 304 cells. I had extra cells and thought, why not do it and see what happens.

I guess that blows your theory above. :)

But a similar set up with 6 packs will not work out this way, nor will 7, 8, 9 10 packs and beyond.
I learned about this from having two 100Ah packs mixed with 4, then 5 then 6, 280Ah and 304Ah packs.
@Solarod (retired EE) has a very interesting post on this - to compute the tie-in points for main pos and main neg he needed to use a computer and iterations of calculations. It is not as simple as just adding up.
The super flat voltage range of the cells - typically 3.0 to 3.4volts means that very very low resistance differences have pretty large effects on pack balance.
I added the 5th pack in the Luyuan box with a busbar and surprisingly it has held balance well with the 4 pack cabinet. The Luyuan box is always a few tenths of a volt higher simply due to the busbar it is attached to but it does have smaller cable and the cables are actually slightly longer so there is some balance due to resistance. . If CV is reached during charging, the packs will fill to basically the same capacity. One may need to occasionally balance the Luyuan pack to the other 4 but it will be some period of time before it is needed.

If loads.charging rate are light and very low C rate, the differences in resistance are minimal. High loads/charging rate and C rates, resistance will have an effect. With the larger banks of 8 or more packs, resistance won't have the same effect as a bank with 4 packs.
 
The resistance is not the same as you posted it.

The thread discussing this was linked, read it
I appreciate your answer, but without wanting to get contentious, could you point out exactly what is wrong with my simple calculation. I would like to know where my simple math fails. That is the whole point of my post.
I have been reading that discussion but unfortunately i cannot find a clear explanation of the math behind the results of the different battery arrangements, all i get is that that forum member is an expert and there has been some software use to come up with those results.
 
You really need a YR1035 meter to be able to measure and work out the real world impedance from the bus bar perspective through the pack and back. Measure each pack while connected to bus bars and then disconnect it and do the same on the others one at a time - adjust cable size/length/crimps to get the impedance as close as possible.

See the second link my my signature line for all the math behind placing the cables for parallel batteries.
 
I appreciate your answer, but without wanting to get contentious, could you point out exactly what is wrong with my simple calculation. I would like to know where my simple math fails. That is the whole point of my post.
I have been reading that discussion but unfortunately i cannot find a clear explanation of the math behind the results of the different battery arrangements, all i get is that that forum member is an expert and there has been some software use to come up with those results.
All resistance in the circuit has to be in the calculations including the internal resistance of the battery, the resistance figures used were from this page.

There is some good information in that link, they just didn't go far enough with the calculations. Any difference in circuit resistance, even minute will affect the current flow thru that circuit.
 
All resistance in the circuit has to be in the calculations including the internal resistance of the battery, the resistance figures used were from this page.

There is some good information in that link, they just didn't go far enough with the calculations. Any difference in circuit resistance, even minute will affect the current flow thru that circuit.
Ok i am testing this shit. I'll post results.
 
You really need a YR1035 meter to be able to measure and work out the real world impedance from the bus bar perspective through the pack and back. Measure each pack while connected to bus bars and then disconnect it and do the same on the others one at a time - adjust cable size/length/crimps to get the impedance as close as possible.
Correct - that is what I have been doing.
each DIY battery pack is slightly different IR and where necessary the cable connection point can be adjusted - towards or away from the main cable to vary the total loop to match the others.
How far down the rabbit hole you go is up to a user, but I know from my original set up, that a near balance is 'good enough' at some point.
This time of year even having a 100kWh + ESS I can drain down to 15% SOC at times. Keeping all the packs in near balance keeps them all working equally hard - no slackers allowed on the homestead.
 
I appreciate your answer, but without wanting to get contentious, could you point out exactly what is wrong with my simple calculation. I would like to know where my simple math fails. That is the whole point of my post.
I have been reading that discussion but unfortunately i cannot find a clear explanation of the math behind the results of the different battery arrangements, all i get is that that forum member is an expert and there has been some software use to come up with those results.
See this post: https://diysolarforum.com/threads/c...ing-battery-currents.37937/page-3#post-574248

Details in post #4 of that thread.

Full tilt boogie here: https://diysolarforum.com/threads/propellorhead-circuit-analysis-math.38972/
 
Last edited:
While i am learning more and more i come up with new questions. In a 16 cell 48v lifepo4 battery build, what hardware would you think is best for protection?
Right now i am putting just a 160 amp dc breaker. Could a class T fuse in combination with a simple battery disconnect switch be better?
Skip the Chinese MCCB's like the ones that are shown in your photographs. A 160A MCCB does not disconnect at 200-240 amps, thus it does not provide any over current protection (maybe short circuit protection). These MCCB's only bring more resistance = heat to the circuit.

I would purchase the most expensive (low resistance) disconnector my money can buy. For me that is an SE INS 100 4P 28925. The four poles are used in 2 times 2P mode: thus continuous current is reduced to 160Amps DC. Schneider Electric does not have any paralleling option for this 30 mm pitch disconnector. The larger SE INV disconnector has such a parallel option, however that is pitch 35 mm. Which leads to making one or connecting 2 wires.

The path resistance of such a Chinese MCCB is measured at 0.45 milliohms.
One contact of an SE 28925 measures 0.33 milliohms at most. With 2 poles in parallel that resistance is half, down to: 0.165 milliohms.

The cheapo battery breakers can be thrown away within 10 times switching. The resistance of the cheapo disconnectors increases every switch. You might increase your switching count when inrush current is limited.

Ps Your spring compression is well done.
 
Last edited:
Skip the Chinese MCCB's like the ones that are shown in your photographs. A 160A MCCB does not disconnect at 200-240 amps, thus it does not provide any over current protection (maybe short circuit protection). These MCCB's only bring more resistance = heat to the circuit.

I would purchase the most expensive (low resistance) disconnector my money can buy. For me that is an SE INS 100 4P 28925. The four poles are used in 2 times 2P mode: thus continuous current is reduced to 160Amps DC. Schneider Electric does not have any paralleling option for this 30 mm pitch disconnector. The larger SE INV disconnector has such a parallel option, however that is pitch 35 mm. Which leads to making one or connecting 2 wires.

The path resistance of such a Chinese MCCB is measured at 0.45 milliohms.
One contact of an SE 28925 measures 0.33 milliohms at most. With 2 poles in parallel that resistance is half, down to: 0.165 milliohms.

The cheapo battery breakers can be thrown away within 10 times switching. The resistance of the cheapo disconnectors increases every switch. You might increase your switching count when inrush current is limited.

Ps Your spring compression is well done.
Interesting, so the chinese MCCBs are janky... I am not surprised. If they dont disconnect at the rated amps they are only good as a dc switch, but you talk also about resistance. I never considered this factor.
Does the resistance of the MCCB make a big difference?
In my system the smaller ones under the inverters have 100 amps going through, at 0.45 milliohms they should dissipate (P = R × I²) only 4.5 watts.
 
I have purchased a number the different Chinese MCCB. I've take them apart and examined them. It really depends on the brand as to how well they are put together.

The TOMNZ are well made... the DiH00l are also well made....as are the TAIXI breakers....the Chintz, BAO, and I forget the other are built similar, but lack lubrication at the moving joints and are less well put together. They also lack parts of the arc chutes that the TOMNZ and good ones have...

I am just about setup to test them to see the trip at the right points. And will take pictures of all the brands I have in pieces.
 
I have purchased a number the different Chinese MCCB. I've take them apart and examined them. It really depends on the brand as to how well they are put together.

The TOMNZ are well made... the DiH00l are also well made....as are the TAIXI breakers....the Chintz, BAO, and I forget the other are built similar, but lack lubrication at the moving joints and are less well put together. They also lack parts of the arc chutes that the TOMNZ and good ones have...

I am just about setup to test them to see the trip at the right points. And will take pictures of all the brands I have in pieces.
Update this thread when you do!
 
Since I have 30 of the Tomzn in my system already and my main ESS Disconnects are the DiHool - I hope they are ok ! LOL.
interested to hear back after your testing.
I just use the breakers as switches, since every place you see one, there is also a class T beside it.
They don't heat up under sustained 80% load on the whole solar plant, so I don't worry about them.
 
In my system the smaller ones under the inverters have 100 amps going through, at 0.45 milliohms they should dissipate (P = R × I²) only 4.5 watts.
This calculation is correct, per pole. Your MCCB looks like a dual pole model.
Thus 9 watts in total for the positive and the negative pole.

At 150 amps, the total heat losses will become 20.25 watts.

With the SE disconnector at 150A the heat losses would be reduced to 7.425 watts.

I'd wish a configurable MCCB (the blue rotating dial, bottom right of the action switch), then I would dare to skip the fuse (0,47-0,48mOhm = 28W@200A).
4P-stroomonderbreker__MCCB-stroomonderbrekers_Kortsluitbeveiliging_Overbelastingsbeveiliging_f...jpg
I've tried Amazon sellers BreTecak, VoeNes, Walfront Store (twice). No Amazon seller actually delivers the UCB 100 R MCCB, they all come with excuses like package lost in transit, etcetera.
 
Last edited:
@Aridom82 I wouldn't mount a fuse in between cells. Your photographs showing them after the MCCB is better in keeping the fuse resistance generated heat away from the cells.

One more (contact) resistance optimization that can be done, and is usefull when transporting currents > 100 amps, with these multi level (multi dimension cable shoe hole) connection terminals, is to use a lower (larger hole) contact level.

I don't know the shape of your terminals. In my case with a square surface on the lowest level, the three level surface areas are:
  • M6 (upper level, red) ≅ 65 mm² contact area
  • M10 (middle level, orange) ≅ 135 mm² contact area
  • M20 (lower level, yellow) ≅ 670 mm² contact area
What is the best combination for diy battery protection, class T fuse_battery disconnect_dc be...png
 
@Aridom82 I wouldn't mount a fuse in between cells. Your photographs showing them after the MCCB is better in keeping the fuse resistance generated heat away from the cells.

One more (contact) resistance optimization that can be done, and is usefull when transporting currents > 100 amps, with these multi level (multi dimension cable shoe hole) connection terminals, is to use a lower (larger hole) contact level.

I don't know the shape of your terminals. In my case with a square surface on the lowest level, the three level surface areas are:
  • M6 (upper level, red) ≅ 65 mm² contact area
  • M10 (middle level, orange) ≅ 135 mm² contact area
  • M20 (lower level, yellow) ≅ 670 mm² contact area
Thank you for your answer. In the end i discarded the idea of putting the fuse between the cells, not for technical reasons but just because is easier to put them after the MCCB. It was just an idea that seemed convenient to me at the moment but it is not in reality.
I considered briefly also doing what you suggest with the battery terminals to get more surface contact, but then i tested a pack at 150 amps with a thermal camera and i decided to leave it as it is for the moment. Again just because is less work, but still worries me that at certain point one of the contacts could start generating a lot of heat, so i may do it in the future. I was thinking of using some copper washer with the mid level inside diameter and the thickness of the step until the upper level so when i screw the busbar down i am making contact also with the mid level too, but of course the thickness has to be exact.
Thanks again.
 
One thing to keep in mind with fuses - they are almost never structural. This means they are not made to be mounted without a fuse block to hold them.

A class T, if you grab both ends and twist they will come apart without a lot of force. The ANL, CNN, Mega (32v version) are the same - put one end in a vice and pull the other and twist and they will come apart ... some are pretty strong when not blown but they come apart easy when they are.
 

diy solar

diy solar
Back
Top