hwy17's Orion battery build

[sunshine_eggo]

5-6mV difference at 3.40+ correlates well with top balance.

 

[hwy17]

5-6mV difference at 3.40+ correlates well with top balance.

It would be extra satisfying to whip cell 7 into perfect alignment, but I know we're actually just looking at this super zoomed in. If it could be zoomed out, it would look like full consolidation. And, when it's doing this under load .007 is already going to be a tight target.

Cell 7 was low in the initial balance too, so I think it may just be it's original state of charge still showing through, and not a problem with the cell.

 

[sunshine_eggo]

It would be extra satisfying to whip cell 7 into perfect alignment, but I know we're actually just looking at this super zoomed in. If it could be zoomed out, it would look like full consolidation. And, when it's doing this under load .007 is already going to be a tight target.

Cell 7 was low in the initial balance too, so I think it may just be it's original state of charge still showing through, and not a problem with the cell.

Given the scale of the deviations, you are picking fly turds out of pepper...

"perfect" is often hard to attain. If you stay < 50mV @ peak absorption, you've won.

 

[hwy17]

Reseating and tightening the bus bars on cell 3 did solve the high resistance numbers I was seeing for that cell.

Now it's actually got a lower resistance than every other cell. I'll tighten the other now highest ones, but try not to get carried away until I snap a stud.

 

[hwy17]

The Orion comes with a big sticker on it that says DON'T RECONFIGURE BUS BARS WITH WIRING HARNESS CONNNECTED.

In the manual, every other page says DON'T RECONFIGURE BUS BARS WITH WIRING HARNESS CONNNECTED.

In the software, it gives you pop up warnings: DON'T RECONFIGURE BUS BARS WITH WIRING HARNESS CONNNECTED.

So what do I do when I go to reseat bus bars on 3? Forget to disconnect the wiring harness of course.

I got away with it this time, I hadn't fully lifted the cell tap by the time I noticed. It's just too easy a mistake to make. Would be best to have a physical interlock that prevented me opening the cell box without disconnecting the harness first.

 

[hwy17]

2300 watt grid input turns out to be inadequate. It can serve our 24 hour average requirements if the system is set to an 80% charge target but it's not enough to be able to timer off the grid during 4-9pm and then recover all the 24h consumption in the 19 hours of on time.

Also, once the solar gets up I'll want the charger to hold the battery at 20% to leave room for solar storage, so then I'll need peak output input to reliably backstop that.

Gonna probably get another 2300 watt.

 

[spitfire6]

Greetings,
I am very familiar with Orion BMS2. Not the junior. There must be a HW difference as the junior appears to measure in groups of 8 and not 12 like most BMS chipsets. Sofware looks the same.

The drift points are just to force the SOC reading to show whatever you want. I would keep the default for your cells.
True DeltaV is better checked at high SOC. less than 90% can make DeltaV look better than it is.

DeltaR is not good. Dielectric grease between busbars is a good idea. If DeltaR stays high I would swap out the cell. The sense wire resistance does not affect this.

The resistance chart needs to be populated with the correct cell resistances. Will prevent CCL & DCL from being reduced.

The Orion only balances the individual cells at a fraction of an Amp. <0.5 I think. fine with 20Ahr, will take days with 200Ahr.
I would only balance if a cell is above 3.5 with a max of 3.65. The fastest way to balance with Orion is to set the charging current at an Amp or less.

Cheers,
Iain.

PS, Cell #10 is the weak link.

 

[hwy17]

Greetings,
I am very familiar with Orion BMS2. Not the junior. There must be a HW difference as the junior appears to measure in groups of 8 and not 12 like most BMS chipsets. Sofware looks the same.

The drift points are just to force the SOC reading to show whatever you want. I would keep the default for your cells.
True DeltaV is better checked at high SOC. less than 90% can make DeltaV look better than it is.

DeltaR is not good. Dielectric grease between busbars is a good idea. If DeltaR stays high I would swap out the cell. The sense wire resistance does not affect this.

The resistance chart needs to be populated with the correct cell resistances. Will prevent CCL & DCL from being reduced.

The Orion only balances the individual cells at a fraction of an Amp. <0.5 I think. fine with 20Ahr, will take days with 200Ahr.
I would only balance if a cell is above 3.5 with a max of 3.65. The fastest way to balance with Orion is to set the charging current at an Amp or less.

Cheers,
Iain.

PS, Cell #10 is the weak link.

Hey! Thank you for reading.

As I understand it the Jr uses 2 of the the same 12 point IC's, with the last 6 disabled or closed. And I believe the Jr1 did so in a 12/4 split, so the even 8/8 split was an improvement for the Jr2 to allow pack splitting fuses like mine.

CCL/DCL are not a feature I consider myself to be utilizing yet, because I have no loads or chargers that can receive a DCL/CCL comm and react to it. I have the pack fused at 150 and I have my DCL/CCL set to allow 150+ with spikes to 300 under almost all circumstances. I'm not sure if I will ever tighten that down.

For now I plan to only charge to 3.437 at all stages from all charge sources, and only balance at 3.438+. I've considered charging to 3.437 and balancing at 3.44 to really ease back on the balancing activity and I might try that later.

I'm gonna try tightening the bus bars on 10 next to see if that resistance goes away.

 

[spitfire6]

Hi,
I was able to use CCL/DCL over Can. Shame you cannot.
it's a good idea to set DCL/CCL so you can set conditions for the contractors to open under fault conditions. I see you have set this.
I would initially charge to 3.65v so that DeltaV was more informative.
I would relocate the weakest of the cells to the end of the string & have the end of the string easiest to get to for cell replacement.
I do like the Orion BMS SW.

I wonder if I can paste a profile? Nope. LOL.

Cheers,
Iain.

 

[hwy17]

Hi,
I was able to use CCL/DCL over Can. Shame you cannot.
it's a good idea to set DCL/CCL so you can set conditions for the contractors to open under fault conditions. I see you have set this.
I would initially charge to 3.65v so that DeltaV was more informative.
I would relocate the weakest of the cells to the end of the string & have the end of the string easiest to get to for cell replacement.
I do like the Orion BMS SW.

I wonder if I can paste a profile? Nope. LOL.

Cheers,
Iain.

Yeah I have some canbus messages that I might be able to get to shutdown my inverter, but otherwise the only option and the only currently enabled option is for contactors to open. I'm relying on my high DCL and CCL only to open the contactors in an extreme fault scenario, if the 150 amp fuse has somehow not already blown.

Contactors also react to high and low cell voltages of course. Also expecting those to never have to open in normal operating conditions.

I'm gonna keep working with 3.437 to try it out, it's kind of my pet project to trial single voltage charging in home storage applications - absorb and float and balance at 55v even.

I probably won't be re arranging since my case is so hard to work in. Luckily 10 is already in a more accessible position.

Do you have any advice on finding resistance factors for the EVE 304's? I'm a bit lost on that, and just relying on my uncalibrated figures as a delta indicator to point out the odd cells, like 10.

Have you ever looked at Nuvation or FoxBMS? I think I will try Nuvation on my next build. They are a north american commercialized version of FoxBMS.

 

[spitfire6]

Drift I'm gonna have to learn as I go. I think it's gonna be a while of not trusting my SOC calculation since you're basically designing the calculation yourself. Maybe @Ampster has drift setting advice.

Hi,
I suspect you have a current transformer fitted to the Orion.
The Orion Coulomb counts to calculate SOC. A very good method but errors can add up so drift points & times are available so you can use the non-accurate voltage vs SOC to drift the SOC up/down AKA reset the SOC. it is a forced override of the accurate coulomb counting. I would think carefully before adjusting the curve.
Cheers,
Iain.

 

[spitfire6]

Yeah I have some canbus messages that I might be able to get to shutdown my inverter, but otherwise the only option and the only currently enabled option is for contactors to open. I'm relying on my high DCL and CCL only to open the contactors in an extreme fault scenario, if the 150 amp fuse has somehow not already blown.

Contactors also react to high and low cell voltages of course. Also expecting those to never have to open in normal operating conditions.

I'm gonna keep working with 3.437 to try it out, it's kind of my pet project to trial single voltage charging in home storage applications - absorb and float and balance at 55v even.

I probably won't be re arranging since my case is so hard to work in. Luckily 10 is already in a more accessible position.

Do you have any advice on finding resistance factors for the EVE 304's? I'm a bit lost on that, and just relying on my uncalibrated figures as a delta indicator to point out the odd cells, like 10.

Have you ever looked at Nuvation or FoxBMS? I think I will try Nuvation on my next build. They are a north american commercialized version of FoxBMS.

Hi,
For cell resistances, if Orion is not aware of your cell type, I would look at similar Ahr ratings in their DataBase. I would err on the low side. It's not critical but too high and you will get weak cell errors when there are none,
Cheers,
Iain.

 

[hwy17]

A note here for any future Orion builders reading. I missed a key potential signaling mechanism which is that my XW inverter accepts a basic open closed switch remote power enable signal. I.e. you could wire the two control pins to a light switch and the light switch would enable or disable the inverter.

This would be a far preferable shutdown mechanism compared to opening the load contactor. Eventually I will go back now and wire a relay to one of the Orion MPO's connected to the inverter on off, and have that relay open when low cell hits a threshold slightly above the threshold that opens the discharge contactor. E.g. turn off inverter if a cell hits 2.6 and open contactor at 2.5 to hopefully avoid the contactor opening under load.

Graceful recovery and automatic restart from a low battery condition might even be possible this way. The Rosie accepts remote power off relay as well. I don't know if it is common in AIO's or other inverters.

 

[hwy17]

System's getting a first good workout in a light heatwave.

Ambient temp in equipment room: 37C. 1"x1/4" 6011 Aluminum Bus bars are running 42C in a 37C ambient environment.

On one hand, that's only a 5C rise over ambient. On the other hand, becoming uncomfortable to touch, I'm heading towards the conclusion that V2 of system bus bars and contactors is going to be overengineered. Maybe 1"x1/2" or 2"x1/4" 6011 or possibly even larger, and maybe 500A contactors just for thermals as well.

And now that I think about it, active ventilation for the bus bar box. Still design for it to be adequate without, but why not have some air flow too.

 

[hogback]

Curious as to the thermal dissipation of the bus bars. Why wouldn't they present the same or less resistance as any other cabling? I'm trying to borrow a smartphone thermal cam from a coworker, for solar thermal water panel check, but can hit my bus bars if I ever get it from him. Should cool off tomorrow. Watch out for rattlers. Seen more in the last week than I have in the last ten years up here. Very close call last Sunday.... ordered some chaps for weekending the hillside.

 

[hwy17]

Curious as to the thermal dissipation of the bus bars. Why wouldn't they present the same or less resistance as any other cabling?

They do, I just never did any proper calcs on 1/4" x 1" 6011 bar to compare it to wire. I gave it the seat of my pants certification for 150 amps based on the look of it.

 

[hogback]

Did 1/4" x 2" for mine. Seat of pants also. Looked at what some others did and upped it a notch or two. One thing I need to do is enclose it, for mice mostly. Really don't want to have to do a full box with tons of grommets, but nothing is jumping out at me. I have eight batteries all 1agw wires to the bars. Bit of a Medusa. What did you do again?

 

[hwy17]

What did you do again?

I used two of these pricey things: https://www.digikey.com/en/products/detail/panduit-corp/CESUM3-D12/16995694

This one is all same size holes and might fit 1 awg not positive about that. It says .47" max.

CESUM3-E14 | DigiKey Electronics

Order today, ships today. CESUM3-E14 – Rectangular - 5.870" L x 2.400" W (149.00mm x 60.96mm) Grommet, Entry Plate Black from Panduit Corp. Pricing and Availability on millions of electronic components from Digi-Key Electronics.

www.digikey.com

 

[SeaGal]

System's getting a first good workout in a light heatwave.

... is that like a solar flare

 

[hwy17]

... is that like a solar flare

Yeah it's like that but with infrared, causing an intense heating of the skin and any other surface it lands on. I believe they call it "strong sun" in the UK, and they get on planes to go look for it

 

[hwy17]

Watch out for rattlers. Seen more in the last week than I have in the last ten years up here.

Thanks for the reminder, cause I honestly usually forget they even exist here, in between the times I'm reminded of it. Haven't seen one yet ever before.

 

[hwy17]

I'm thinking about expanding the pack to 2p, split between two metal cases. Does this look right? 48 total bus bars required if using individual bus bars stacked.

 

[sunshine_eggo]

Note how horizontally, you only have one bus bar connecting parallel cells, but when you transition vertically to reverse direction, you have two.

I don't recommend this. It creates imbalance in those 4 cells.



For the single interconnect between the two halves, I would try to make that resistance the same as the other single interconnects.

 

[hwy17]

Note how horizontally, you only have one bus bar connecting parallel cells, but when you transition vertically to reverse direction, you have two.

I don't recommend this. It creates imbalance in those 4 cells.

For the single interconnect between the two halves, I would try to make that resistance the same as the other single interconnects.

Thanks. I forgot about how these create varying resistances.

My understanding is that varying bus bar resistances doesn't inherently create cell imbalancing forces. If it does, that's mysterious to me.

It does affect BMS measurements though (and therefore would affect it's ability correctly balance, maybe that's all you meant anyway), and my BMS measures in two series strings of 8 cells. So the BMS will not be affected by the highest resistance connection between the packs, which will also have the Class T in it, but it will be affected by this 4 square bus bar vs 3 line bus bars.

Aha, I see it now! I was going to start trying to justify it as a minor difference but what you're showing me is we literally just delete one of them and then it's basically identical to the rest of them, 3 bus bars in a line. Thanks again.

 

[hwy17]