diy solar

diy solar

Weird bus bar idea

Leon

New Member
Joined
Dec 15, 2019
Messages
122
I am building a 3P8S pack from 150Ah Lishen cells. For various reasons, a good physical configuration is this:

cell layout 2.PNG

The supplied bus bars, simple plated copper links, will not work with this arrangement. Would it be a stupid idea to use large plates to make the series/parallel connections? Like this:

big bus bars.PNG
For material, I'm considering some 3mm aluminium that I have. For reference, my main fuse is 225A.
The lines on the top left plate indicate possible cuts to make the plates a bit flexible. Easy to do and I can't see any disadvantage. I have not shown the holes to the terminal posts. The cells have welded M6 studs like this:

DSC_0856.JPG

What do you reckon? Has anyone done something like this? I think I've seen something similar with laser welded bus bars, but not with bolted connections.
 
you’d also want 3 wires going from the top row to the bottom instead of 2. I would also recommend thicker aluminum. 6061 Al is about 40% as conductive as copper, so size accordingly.

the other issue is compression & expansion. When the cells expand, where do they go and what does that do to the bus bar?
 
you’d also want 3 wires going from the top row to the bottom instead of 2.
Agreed.
I would also recommend thicker aluminum. 6061 Al is about 40% as conductive as copper, so size accordingly.

Fair enough. A couple of sources I have seen indicate 3mm is sufficient but it's easy enough to beef it up.
the other issue is compression & expansion. When the cells expand, where do they go and what does that do to the bus bar?

The cells will be constrained by stiff panels along the long sides, connected with threaded rods through the spaces between the cells (vertical gaps in my sketch). As yet undecided whether to use springs. Only 6 cells in a row should limit potential expansion versus other arrangements. I also think that using relatively smaller cells rather than the more common 280Ah size reduces the whole issue somewhat since the cells themselves have better structural integrity. I guess the setup I have shown is no worse than any other using non-flexible bus bars.
 
Thanks! Doesn't mean it's a perfect idea but it's nice to see someone else has tried it.
 
To improve on the the exit terminals, use 3 short wires ganged together with an oversized lug. Slots like you've shown on the top left plate, would help to balance what's going on thermally, both temp and expansion wise.

The E fields that will develop using this parallel plate bus bar will be dog bone shaped X3. To maintain symmetry, based on the images you provided, enlarge the bus plate on all sides a little.

Welded studs will help to ensure well balanced connects, provided you have a good terminal cleaning technique and tighten to the manufacturers specs. To prepare connections, I use a light touch with a stainless wire brush on both mating surfaces, next wipe down with an alcohol pad followed up with thin coating of NOALOX(anti oxidant compound).
 
The only thing that comes to my mind is getting good contact across all the terminals due to slight variations in height... the cuts would help ... maybe alternate tightening the bus plates down and compressing the pack, loosening the plates and re-tightening, etc... so pull the cells all to the same height. Fortunately with welded terminals you can apply a fair amount of torque which should help create a good contact patch.
 
 
The currents between your cells will be VERY imbalanced and your bus bars don't lie flat so you have poor terminal connections. I highly recommend against this configuration. Cool idea ... but too many problems. It might work under light-load circumstances but I wouldn't want to support it, especially for long-term use.
 
The currents between your cells will be VERY imbalanced and your bus bars don't lie flat so you have poor terminal connections. I highly recommend against this configuration. Cool idea ... but too many problems. It might work under light-load circumstances but I wouldn't want to support it, especially for long-term use.
If they don’t lie flat they must heat but at 200 amps they do not heat at all. So where is the problem ?
 
If they don’t lie flat they must heat but at 200 amps they do not heat at all. So where is the problem ?

Cell balance and irregular/premature aging. The cells out the outer edge (left in your picture) will not take the same load as the cells closest to the battery +/- cables. The non-flat terminal contact means a high resistance connection which means a high degree of variability of resistances which further exacerbates the previously-mentioned problem. It doesn't have to be enough to create heat at 200 amps (and I presume that's 200 amps at the battery, meaning 200/8 or 25 amps per cell under ideal conditions). Like I said it might work for now but long-term you'll start having issues with various cells performing so much worse than others that things will get out of balance and you'll have trouble charging/discharging without triggering BMS LVD or HVD.
 
To improve on the the exit terminals, use 3 short wires ganged together with an oversized lug.
Yeah ok, that makes sense. Could a specially shaped solid plate achieve the same thing?

Slots like you've shown on the top left plate, would help to balance what's going on thermally, both temp and expansion wise.
I was only thinking of stress from physical misalignment, but those are good points. It's easy to do.

The E fields that will develop using this parallel plate bus bar will be dog bone shaped X3. To maintain symmetry, based on the images you provided, enlarge the bus plate on all sides a little.
It is this aspect that I wondered about and have no knowledge of. To clarify, if the plate was too small, the outer terminals would see less current flow, correct? Would the ideal plate be some specific shape or just infinitely large? I'll probably just go as big as I can without the plates getting too close in the middle.

Welded studs will help to ensure well balanced connects, provided you have a good terminal cleaning technique and tighten to the manufacturers specs. To prepare connections, I use a light touch with a stainless wire brush on both mating surfaces, next wipe down with an alcohol pad followed up with thin coating of NOALOX(anti oxidant compound).
Sounds good.
 
Last edited:
Why don’t you use infinite parallel pack as I do. I can add any number of new 16s packs I want at any time without any stop of the working system. What do you think?

View attachment 55264
That's an interesting setup. I'm having trouble understanding what I'm looking at. What sort of loads are you serving with that? I suppose with my idea, I could add more cells around the outside of the pack at any time by making new, larger bur bar/plates. No plans to make it bigger any time soon though.
 
The only thing that comes to my mind is getting good contact across all the terminals due to slight variations in height... the cuts would help ... maybe alternate tightening the bus plates down and compressing the pack, loosening the plates and re-tightening, etc... so pull the cells all to the same height. Fortunately with welded terminals you can apply a fair amount of torque which should help create a good contact patch.
Yeah, I hope that works out. The welded studs appear to have been placed on the terminals manually, as they are not all aligned exactly the same. I will have to drill each plate to mach a specific set of terminals, probably with slightly oversized holes. Still, I have way more contact area than the more common round-based posts that we see on the 280Ah cells.

I suppose annealing the aluminium would be good as it would make the material better able to conform to irregularities. I wonder what the electrical effect would be though.
 
Thanks for pointing me to your thread. I really like your aluminium terminal posts - I think they are the best available on these cells. I had not seen that type when I ordered mine.
 
That's an interesting setup. I'm having trouble understanding what I'm looking at. What sort of loads are you serving with that? I suppose with my idea, I could add more cells around the outside of the pack at any time by making new, larger bur bar/plates. No plans to make it bigger any time soon though.

You can see that each 16S is on 2 row and paralel is between each every 2 cells , please open the image and see the connections of batteries and also on the end are the inverters. Also for better balancing inverters will be moved at series ends. The setup is for long therm storage and laod is not too high. Also using 2.5mm2 balcing with WAGO allsows me to add up to 5 JK Active BMS to work in paralel. I consuider having all batteries as one pack and connecting4 or more inverters to the same pack better than using separate pack for each inverter (not possible with paralel setup also, so it should be same battery pack).
 
Cell balance and irregular/premature aging. The cells out the outer edge (left in your picture) will not take the same load as the cells closest to the battery +/- cables. The non-flat terminal contact means a high resistance connection which means a high degree of variability of resistances which further exacerbates the previously-mentioned problem. It doesn't have to be enough to create heat at 200 amps (and I presume that's 200 amps at the battery, meaning 200/8 or 25 amps per cell under ideal conditions). Like I said it might work for now but long-term you'll start having issues with various cells performing so much worse than others that things will get out of balance and you'll have trouble charging/discharging without triggering BMS LVD or HVD.

The is no heat.

You can see that each 16S is on 2 rows and parallel is between each every 2 cells , please open the image and see the connections of batteries and also on the end are the inverters. Also for better balancing inverters will be moved at series ends, now they remained from first 2 series of 200Ah and i just added 2 more series of 16s of 310Ah. The setup is for long therm storage and load is not too high. Also using 2.5mm2 balcing with WAGO allows me to add up to 5 x JK Active BMS to work in parallel. I consider having all batteries as one pack and connecting4 or more inverters to the same pack better than using separate pack for each inverter (not possible with parallel setup also, so it should be same battery pack).
 
JimStLaurent said:
The E fields that will develop using this parallel plate bus bar will be dog bone shaped X3. To maintain symmetry, based on the images you provided, enlarge the bus plate on all sides a little.
It is this aspect that I wondered about and have no knowledge of. To clarify, if the plate was too small, the outer terminals would see less current flow, correct? Would the ideal plate be some specific shape or just infinitely large? I'll probably just go as big as I can without the plates getting too close in the middle.
Let's use the water analogy for current flow. Your bus bar mounting hole is the drain hole. And the battery terminal the drain pipe. The water flows towards the drain hole. The water(current) will fall over the front edge of the drain( bus bar hole) and wrap around all the way around to back side, filling the pipe(terminal post).

Shown on your sketch, the 2 middle terminals look much more robust than the 4 on the corners. You can easily implement this design change to make all 6 terminal post Drain profiles match. The same applies to rounded corners, no current will make it into the hard 90 corner. But this would be unreasonable to do well with hand tools.

These are second order effects, but the key point is symmetry, its very important in high power delivery to minimizes system imbalances.

Note to Self: Make friends with a beer drinking plasma cutter owner.
 
Last edited:
Back
Top