hwy17's Orion battery build

[hwy17]

Odd though, I don't have an 80% drift point configured, so I don't know exactly what logic did that.

 

[hwy17]

I have turned the corner from CC to CV without any problems so far.

 

[hwy17]

Oh but I see the most precarious part was still to come after that. Oh well, 3.55v never hurt anyone right?

My budget model Fluke 101 says 55.22v at the charger terminals, but my BMS says 55.3v. Who do I believe here? The BMS, I think.

 

[42OhmsPA]

You'll be fine at 3.55V as long as you don't leave them to long.
I'd hold them at 55.2V for a while (overnight) to let them balance out a bit.
Do you have another meter to verify / check?

 

[hwy17]

3.555v now I'm cutting the charger off and they will continue balancing as long as they stay over 3.42. Time for a better meter I think.

 

[hwy17]

Restarted at 55v and the highest cell voltage is now headed in the right direction.

 

[hwy17]

Yay when I was watching highest and lowest cell last night, they didn't seem to be moving at all for a few hours, but they did overnight.

I believe I have found the answer to one of my questions above. At least in the case of the Orion, it being described as having 200mA balancing current, it looks like 200mA of charging current is what holds steady state balancing.


Edit: <100mV delta milestone passed. BMS is configured to balance until <10mV. If I were to do this again, the excursion to 3.55 was unnecessary. I would start at more like 54v, get it steady, and then move to 55v. It didn't really click for me until I'm actually doing this now that you could have like charge to 3.375 avg and balance at 3.45 and your high cells will still poke through the balancing threshold.

 

[hwy17]

 

[hwy17]

This is the load and charge controller wiring, more or less done. Behind this battery controls box is a Lifepower4, and it's wired directly to the inverter terminals doubled up on the main battery lugs. My idea is that the Lifepower4 provides a little bit of redundancy here, particularly to keep the load online while software maintenance is being performed on the main battery. It violates the seperate charge and load design topology a bit, but I figure at the location it's in, as long as both batteries are online, the Lifepower4 should always primarily act as a load unless the batteries have been disconnected. There is also a disconnect switch so that the Lifepower4 can be maintained and individually charged if necessary.

I restarted the charge last night at 55v charging and balancing at 3.44. Glad to see the high cells are in line now. I will drop the balancing to 3.437 and see if that is enough pressure to drag the low cells all the way.

 

[hwy17]

Oh, wait, that's already a .010 delta! Balancing is nearly done. The Orion only allows settings down to .005. I am setting it to .007 now.

 

[hwy17]

Ok, the house is up and running on double conversion now. An SOC trigger will turn the charger on at 50% and off at 80%.

The plan for now is periodic balancing will be manually initiated with the charger on override of the SOC trigger. This spring, hopefully solar will get online and then solar charging should be regularly providing full charges.

 

[hwy17]

Should I be concerned with loaded resistance discrepancies? I'm not going to be, unless I find out there's a reason to.

 

[sunshine_eggo]

I would double check that all connections are torqued to the same values.

Is this also potentially the sense harness resistance? If it's the sense harness resistance, I wouldn't worry that much.

 

[hwy17]

I would double check that all connections are torqued to the same values.

Is this also potentially the sense harness resistance? If it's the sense harness resistance, I wouldn't worry that much.

I hand checked them for warmth and they're all cold. I torqued them by feel and I'm hesitant to mess with them unless there's a problem. Could be sense harness but resistances are all .50 at rest.

 

[sunshine_eggo]

IMHO, get a torque wrench. You have too much invested to leave out this comparatively inexpensive tool.

 

[hwy17]

IMHO, get a torque wrench. You have too much invested to leave out this comparatively inexpensive tool.

I have an inch pound torque wrench but I also have an ingrained superstition from working on dirtbikes not to put a torque wrench on a 10mm bolt head. I have snapped bolts that way and never snapped one by feel. I will consider giving this bus bar a check and snug though if it's consistently higher resistance.

 

[hwy17]

A check of the 1-2-3-4 cell connections and slight snug on 2 of them that would take it resulted in no IR changes as far as I can tell.

This may just be the reality of buying unmatched DIY market cells.

 

[hwy17]

Charger kicked on. They say a battery is only either charging or discharging, but they switch back and forth between the two very fast.

 

[MisterSandals]

Charger kicked on. They say a battery is only either charging or discharging, but they switch back and forth between the two very fast.

Yes, i saw mine was doing something similar just this morning. I was alarmed until i saw it was +10A to - 5A

I think its like making sausage (i grew up on a dairy farm in wisconsin), you really don't want to know what is really done. It would be GREAT and also EASY to do this better but there is no effort in making this better (as easy as it would be).

On my pic, the spike graphic is horrific while we slept and only the fridge cycled. At 6:30 the hot tub (1hp 120V pool pump) was turned on. Gas heat.

I dunno. Hopefully helpful/comforting. Seems like it should be smoother and more reactive.

I bumped my charge % up to 50% at 2:30, thus the spike. Was looking to get a little more charge and watch my suspect cell.

 

[hwy17]

My utility room is being gently warmed by the double conversion inefficiency. Maybe about 300w. Nice for me in the winter because my toolbox and sort of work bench is in there too. Normally unheated.

The 2300W is very closely matched to our average consumption. The charger started at 50% SOC at around noon, and 8pm it's at 59% now.

 

[hwy17]

1hp 120V pool pump

Above ground? I have an 18' round.

It's my one significant load that happens to be wired to the parent panel that's not going off grid here, so if all goes well that will be the summer project after solar in the spring, to run a new wire for that.

 

[MisterSandals]

Above ground? I have an 18' round.

Hot tub. I added an external pump, filter and natural gas heater in a little (3ft cube) sound-proofed out building. As much as i love a gas fired hot tub, i miss the days of cheap NG.

 

[hwy17]

Hot tub. I added an external pump, filter and natural gas heater in a little (3ft cube) sound-proofed out building. As much as i love a gas fired hot tub, i miss the days of cheap NG.

Hey it's still cheap compared to electricity, but maybe not solar.

I have a 160sq ft solar pool heater on my 18ft agp. It works well June-Aug, but not outside of that.

 

[hwy17]

I ran the battery down to 0 SOC today and I think it was successful. Under 3kW load, cells first started dipping below 3V at 2.5% SOC and we're all still in the 2.9's at 0% SOC. I don't know if this really constitutes a reliable capacity test, but I'm just glad that it's not overestimating capacity and crapping out at 20%.

I did notice that during particularly hard pull of about 100-120 amps at 20ish SOC live voltages did get erratic with spikes down to 2.5V. I need to read the voltage control logic in Orion better to understand exactly when it's going to react to an overdischarge.

 

[hwy17]

My double conversion trial ended yesterday after I went to investigate auto charger output not applying and when I poked at the wiring harness in the orion, the charge contactor would open. The controls harness is a much smaller connector than the cell tap harness, and I wish they would use something more sturdy. I can't tell how it was lose, but I reseated it and I can poke it now without issue, and auto charger works fine.

This is the first rebalance. It's not under load so balancing will probably not stay this tidy once it is. But it seems fast enough to me, idk.