diy solar

diy solar

Help diagnose a mysterious but substantial voltage drop!

Dzl

Unofficial Forum Librarian & Perpetual Newbie
Staff member
Moderator
Joined
Dec 10, 2019
Messages
5,110
Location
West of the rockies
I am posting this here on behalf of an acquaintance on another forum, since the good folks here have much more expertise in this area than the people in the forum it was originally posted to. Can anyone help diagnose the problem here, or point out any obvious things to test?

Now, the battery bank is giving me some problems, so I'm hoping someone here has an idea of how to proceed. I have a 24v LiFePo4 bank with 120 amps capacity. The positive terminal is wired to a 100 amp inline fuse, which then goes to the master battery switch on the electrical panel, but it also goes directly to the positive terminal on my AIMS 2kw inverter/charger. The negative terminal of the batteries is attached to the BMS from electrical car parts company, which is then hooked up to a Victron Energy BMV-700 shunt for battery bank monitoring. That shunt, like the inline fuse, is attached to both the negative bus bar at the electrical panel, and also the negative terminal on the inverter/charger.
Everything was working great--the inverter/charger had charged the batteries to max capacity, and when I unplugged the shore power, found that the batteries could power my shop vac no problem through the inverter. Great. Well, after I topped up the batteries overnight, I unplugged the shore power and turned the inverter off. The battery voltage then rapidly dropped from 26 to 24 to 22 and then the BMV-700 display went blank.
I tested the voltage of the batteries themselves, and they were good (like 26-28v, I don't remember). Then I tested the voltage between the shunt and the fuse, and it was like 4v. I disconnected everything, and reconnected it again, and it went back to normal. Okay.
A few days later, the same thing happened. I suspected the inverter was the culprit (because it happened only after I turned it off), so I disconnected the inverter--but it didn't work. I disconnected everything again, and reconnected, and it went back to normal. But this time, I left the inverter disconnected. A few hours later, I noticed it had happened again. So obviously it's not the inverter, it's either the shunt or the BMS.
I can't trouble shoot for a while, but has anyone heard of this happening? Anyone have any ideas? Is the shunt confused and blocking the battery power? Or is this normal?
 
Did by any chance did the display 'Wake' when the inverter was turned back on?
Sounds like a 'Sleep' mode to me since the batteries aren't being drained.

The Shunt only senses energy MOVING through it, when energy stops moving, it will often 'Blank' out when internal capacitors discharge, and they often have a 'Sleep' mode.

Check to make sure the Shunt wiring is connected correctly, the display shows 'XX.xx' Amps, watts, volts or whatever, and NOT '-XX.xx' (-) meaning the connections are backwards or the shunt itself is installed backwards.

Where is the display power coming from?
Again, if it's not connected to full time power that stays on when the inverter/battery switch is shut down, it will show charge left in capacitors until that's exhausted and shut down.
 
Last edited:
  • Like
Reactions: Dzl
I'd check how he has the BMV wired up. That it sees a drop off in voltage while voltage on the battery is measured as correct might be a clue. If he is turning the inverter off with an external isolation switch it could be that the BMV is monitoring the voltage on the inverter side of the switch and is seeing the voltage on the inverter's capacitors instead of the battery.

Just a thought, may not have any relationship with reality.
 
I'd check how he has the BMV wired up. That it sees a drop off in voltage while voltage on the battery is measured as correct might be a clue. If he is turning the inverter off with an external isolation switch it could be that the BMV is monitoring the voltage on the inverter side of the switch and is seeing the voltage on the inverter's capacitors instead of the battery.

Just a thought, may not have any relationship with reality.

Having screwed up in every way I could find, I've done things like get shunts in backwards, connections on backwards, display wired on the wrong side of switches, etc.

Does No One diagram their systems?
 
Okay, I've asked him to provide a diagram if he has one, and asked if it could be a sleep mode issue, and passed on your suggestions.
 
Last edited:
This is my understanding of the description.
1578089363179.png
@Dzl, Could you confirm? Better yet, invite him to this forum.
 
Assuming the above diagram is correct....

If the voltage at the battery is good and the voltage between the shunt and fuse is bad, I suspect the BMS has tripped. When you take it all apart and put it back together, it resets the BMS and you are good till it trips again.

> The battery voltage then rapidly dropped from 26 to 24 to 22 and then the BMV-700 display went blank.
This was most likely the inverter capacitors discharging after the BMS tripped.

When it happens again, measure the voltage from the battery positive to the output of the BMS. If it is low, it is a strong indication the BMS has tripped.

The next question is "Why did the BMS trip". I don't have enough info to even speculate on that.
 
This is my understanding of the description.
View attachment 4437
@Dzl, Could you confirm? Better yet, invite him to this forum.

Can't confirm, sorry, I have no more information than you I posted his post in its entirety. But I have asked him to provide a diagram and to followup on some of the questions you all have brought up, and I have given him the link to this post. I will forward the diagram along as well
 
Got this reply:

I appreciate the extra help, thanks. I’ve actually isolated the problem. It’s not the inverter or the shunt... it’s the BMS. Either the BMS is malfunctioning and turning off the current, or its working properly and turning off the current because my batteries are damaged. The BMS allows power to flow through again once I’ve completely removed any load, but then once I put a load back on—the shiny battery monitor, or even the battery panel indicator lights—it’ll turn everything off in a minute or two.

I suspect my batteries are bad, even though their voltage is ~26 volts throughout this whole thing, because they sat for 18 months in my closet in its original packaging. But then again, the voltage says it’s fully charged—maybe the BMS is malfunctioning. I have to hit up Carl from the Car Parts company to troubleshoot. Unfortunately for me I’m about to deploy for two months so this will have to wait. Thanks again for the extra help.

Here is the post from the other forum if anyone is interested, its a short bus build.
 
A few days later, the same thing happened. I suspected the inverter was the culprit (because it happened only after I turned it off), so I disconnected the inverter--but it didn't work. I disconnected everything again, and reconnected, and it went back to normal. But this time, I left the inverter disconnected. A few hours later, I noticed it had happened again. So obviously it's not the inverter, it's either the shunt or the BMS.

OK.... try this. Leave the Inverter out and turn off the battery disconnect. If the problem happens again then you have really narrowed down the potential problem areas.

Based on the limited data provided, it sounds like either the BMS is bad or something is overloading the DC. (There are a LOT of assumptions built into that statement so take it with a grain of salt)
 
  • Like
Reactions: Dzl
His next step should be to check the voltages on the individual cells. If they all look good, then I would suspect the BMS (or how it is hooked up). If a cell is bad...then that is the problem.
 
His next step should be to check the voltages on the individual cells. If they all look good, then I would suspect the BMS (or how it is hooked up). If a cell is bad...then that is the problem.

I forwarded your advice along, thanks!
 
This is my understanding of the description.
View attachment 4437
@Dzl, Could you confirm? Better yet, invite him to this forum.

His response:
I’ll be posting a write up on how I built the battery bank with pics and schematics soon, but yes that diagram is accurate. I dread having to unwire the individual cells because it was a huge pain to wire them up... all 32 of them... so I’m just going to wait until I get back to troubleshoot that much further. Fortunately it charges just fine when plugged in to shore power so I’m comfortable leaving the whole thing as is for the next two months.
 
I dread having to unwire the individual cells because it was a huge pain to wire them up... all 32 of them...
Interesting, if he has 24 volts and 32 cells, he either has 8S4P or 4P8S. If it is 8S4P he can measure each cell without taking it apart. If he has 4P8S, measuring the voltage on each 4P grouping if one of them is way low, that pack has one or more bad cells.
 
Thank you for offering some advice... this is my first time building a battery bank and so I'm pretty much learning as I go. Dzl is a huge bro for disseminating my issue here.

Here's a quick run down of my setup:

One of my colleagues ordered hundreds of these pouch cells (3.2V, 30 amps) from China for his toy hauler, and sold me forty of them for about $17 a piece. He was moving across the country, so even though I wasn't even close to doing the electrical system for my skoolie yet, I had to purchase them then or else they'd be gone. This was spring 2018.

Well, a few deployments and bone tumors later, I finally got to the electrical portion of my buildout 18 months later this past November. The pouches were sitting in their original packaging in my closet the entire time (as in, the ambient temp never deviated from 60-80 degrees, and they weren't wired to anything). I checked the voltage on each cell; one box was 3.30, the other 3.34, with maybe five that were slightly different (like 3.29, 3.35, etc...). I only needed 32 cells, so I shelved the odd ones and wired the rest up using 6 AWG welding wire (see the pics below) in an 8S4P configuration for 24V and 120 amps. I also purchased the 24v 8s BMS from Electric Car Parts Company (what my colleague used in his set up) and their balancers as well. I also use the BMV-700 shunt/battery monitor (the black rectangle in the schematic) and a 100 amp inline fuse (the red rectangle in the schematic). Both of these are then connected to an AIMS 2kw 24v inverter charger and the electrical panel.

how-to-install-lifepo4-battery-skoolie-0.PNG


how-to-install-lifepo4-battery-skoolie-00.jpg



how-to-install-lifepo4-battery-skoolie-1.JPG


how-to-install-lifepo4-battery-skoolie-2.JPG


how-to-install-lifepo4-battery-skoolie-4.JPG



how-to-install-lifepo4-battery-skoolie-8.JPG


how-to-install-lifepo4-battery-skoolie-13.JPG


When it's connected to shore power, everything works great. When it's disconnected from shore power, it worked great as well but only for a few weeks. Now it appears that any time there is a load applied to the system, no matter how great or small, the BMS will cutoff the power after a minute or two. That load could be something as small as the inverter monitoring itself, or the BMV-700 monitoring the battery voltage, or even simply just the indicator LED at the battery control switch all by itself. Apparently that's enough to trip the BMS. Then, once I disconnect the load (for instance, turn the master battery switch off (thereby disconnecting the indicator light), and then reconnect it--boom, it works great.

So either the BMS is malfunctioning and erroneously cutting off power when it's actually fine, or my batteries are bad and the BMS is simply doing it's job. The voltage at the battery tabs is about ~26V. Could it be that the batteries completely lost their capacity while sitting in my closet for 18 months? And instead of each cell having 30 amp hours, they only have a fraction of that? I don't think that makes sense, because surely if they had no capacity, the voltage would be far less than 26.

Is it possible the BMS has a malfunction? Could be, but Electric Car Parts Company seems pretty shit-hot with their equipment.

In any case, I'm about to deploy for 2 months here so I don't have time to troubleshoot now, it'll have to wait until when I get back. My next course of action is to start emailing customer support, but there's no rush there. Anyway, thanks for what everyone has said here, it's definitely helping an amateur new guy like me.
 
  • Like
Reactions: Dzl
First: Thank you for your service!


Wow. That is quite the set-up. I have never worked with pouch cells but they should not be all that different electrically. Do you have a data sheet for the cells?

Here are a few first impressions/questions

* The BMS you have is a Daly. That is used a lot by the folks on this forum.
* You have the 1S balancers. The Daly also has a balancing function (All be it weaker than the 1S balancers)
* The connections on the cell tabs look susceptible to shorts. Be sure to at least tape them up to prevent shorts. (A short could explain what you are seeing)

So either the BMS is malfunctioning and erroneously cutting off power when it's actually fine, or my batteries are bad and the BMS is simply doing it's job. The voltage at the battery tabs is about ~26V.

I mostly concur with the analysis. Some other possibilities are:
* There could also be a problem with the 1 S balancers
* There could be a problem with the BMS wiring
* There could be a loose/bad connection.
* A (Intermittent?) short across a couple of the tabs could cause the BMS to shut down.

The first things I would do:
1) Remove the 1S balancers. You don't need them for now. (Maybe not at all)
2) Re-verify the wiring on the BMS.
3) Check all the connections looking for a loose or corroded connection.
4) Verify there are no shorts between the tabs.
5) when it fails, check the voltages across each 4P set off cells. (You don't need to disconnect them to do this) If one of the sets is way off from the others it is a big clue.

Could it be that the batteries completely lost their capacity while sitting in my closet for 18 months? And instead of each cell having 30 amp hours, they only have a fraction of that? I don't think that makes sense, because surely if they had no capacity, the voltage would be far less than 26.

Is it possible the BMS has a malfunction? Could be, but Electric Car Parts Company seems pretty shit-hot with their equipment.

Anything is possible, but I wouldn't think the cells would degrade that much. However, voltage is not a good indication of capacity, so even if the voltage is reasonable, the capacity could be bad.

Be safe on your deployment!!!
 
Last edited:
Can you measure the voltage across the BMS? That should be essentially a short circuit in normal operation, so zero volts across it. If it trips, it will open up and the voltage across it will increase. If you see voltage across it, it has tripped...next step is to figure out why. :)
 
I'd like to offer an update to this mysterious problem. I contacted the company I ordered from, who then contacted the manufacturer. After some further testing, they determined it was a defective unit and sent me a new one. Well, the new one worked even worse than the defective old unit. The old one would charge no problem, and only cut out after shore power was removed. The new one wouldn't even charge, it would just keep the voltage cut off to 3 volts. I decided to pretty much give up on BMS's at this point. This was also when SIP began, so since I had a ton of time and the balancers worked great, I setup an 8S1P back and charged/balanced all 40 pouches to within .05V. Just for shits and giggles, I hooked up the old BMS again, and to my surprise, it works as advertised. I tried the new one (it's a bigger better model) and it still malfunctions... so I have no idea what's going on... but the original BMS works great and I'm proceeding with that one. Despite these complications, I still recommend electriccarpartscompany.com but in the future I will probably just get the Renogy LiFePo4 batteries with a builtin BMS.
 
Back
Top