load > pv during day

Estragon

A 0.5v difference between batteries while charging seems pretty high to me, implying some higher resistance batteries and/or connections (see Bill's comment about higher resistance showing up with increasing current).

OldMan

Shazam! Y'll have convinced me lithium iron phosphate batteries are the way to go. Thanks!!! :-D

mevenable

I finally got the problem sorted. Although I was 100% sure I had checked all the terminals to make sure they were tightened properly, it turns out that one of the connections in the back on the bottom row was very loose - the installers had not tightened it.  It could be tightened about two entire turns with ease, then a bit more with effort.  Since I tightened it, there have been no more issues with the bottom battery bank checking itself in and out, and everything is functioning better, although the two batteries to the left of what was the loose connection are consistently a bit lower than the 2 on the right, at rest or during charge or whatever, maybe with time they will even back out

BB.

Yep. You (and your installers) have to be very thorough during installation and verification of the system. All it takes is one little mistake, and everyone is plowing through the wiring and hardware trying to figure out what is wrong.

When something does not look right (battery voltages do not add up to battery bus voltage)--There is either a bad connection or a violation of the laws of physics. Unlike human laws which people twist to their own ends--Mother Natures laws cannot be...

-Bill

mevenable

Another lesson learned.  It seems that when battery strings are connected in parallel, the string reading will always be the average of the two strings, and if you disconnect one string (which is what was happening with my loose connection) then you would end up with a reading equal to the other string.  So, you could have the individual batteries in the detached string adding up to something quite different than what you get for the string.  When you put together two strings, one that was used all night and was at 46.8 V and one that was detached and has 49.2 V, they both will instantly measure 48.4 V

mevenable

I supposed I should have realized this, but since I was always being told to test the voltage of both strings, I assumed they could be different

mevenable

I think most people I was working with presumed I already understood some things.  Again, I really sort of do, but it escaped me at that moment. It actually dawned on me while dealing with something else, and then I went back to double check connections.  Seems like testing the terminals for resistance was not an effective thing either, I had been told to do that, but it must depend on very precise placement of the electrodes.  Figuring out these systems and getting the kinks out of them on my own in a remote location, with no end-user tech support, and no effective back-up from the installer, is challenging.  I am very grateful to the people at NAWS that have been corresponding with me, and would recommend this approach to others.  Even when people don't understand the question, or their answer is not quite what I was looking for, something they say is often very helpful, sparking another train of thought.  I will keep up with you guys and am enjoying looking at other people's submittals too.  Maybe I will find something to contribute to someone else someday.

Estragon

We all started in exactly the same situation. None of us is born with the knowledge on offer here. We all learned through a combination of reading and actual experience.

BB.

A very quick way to compare the "health" of your battery banks is to use a DC Current Clamp DMM (digital multi meter). DC Current Clamp meters are very safe and easy to use. You set the meter to DC range, zero the meter (DC clamp meter readings tend to drift over time/temperature) and then clip on one of the battery bank string's cables. (just one cable at a time, if you put the + and - cables both in the clamp, the readings add up to zero current --- I.e., -40 amps + 40 amps = zero amps).

If you see an issue (battery strings should be sort of close, but they will probably never exactly match charging/discharging currents--Something like 1/3 vs 2/3 current "balance" or greater, something is probably wrong)... Then you can use the DMM to measure voltage drop.

Again, with relatively high charging or discharging current, in a series string, you can measure the cable+connection voltage drops by placing one meter lead on the negative terminal of battery A and on the positive lead of battery B. You are now measuing the entire cable+connection voltage drop (again, with current flowing through the battery string). And typically, you are looking for something that stands out. You may measure 0.02, 0.05, 0.1 volts drop and find one cable that has 1.0 volt drop--That is the guy you need to really pay attention too (dirty connections, loose connections, etc.).

Here are a couple meters that have been used around here (you can pick one, or find another that meets your needs... Decent meters usually run from $60-$100, and a Fluke high end meter can run you $300 or more--For our needs, typically the $60-$100 meters are "good enough").

http://www.sears.com/craftsman-digital-clamp-on-ammeter/p-03482369000P
https://www.amazon.com/Uni-T-B4Q094-UT210E-Current-Capacitance/dp/B00O1Q2HOQ
https://www.amazon.com/Uni-T-B4Q094-UT210E-Current-Capacitance

I have the first one--Works well for $60 (I did have one meter fail and cause a AAA battery leak--Not sure if failure or leak happened first). The last one, is $41 and seems to have good ratings. The middle one was recommended by a poster here.

You can use your car's electrical system to experiment on how to measure DC current and voltage drops. Car not running, lights off. Lights on, start car, etc... and get some experience with the meters (the DC Current Clamp Zero function is not always obvious which button to press and how to use it. For the first meter, with no current flow (or no wire in clamp), you push the button on and the ZERO displays on the LCD, and then you make your current measurements--To rezero the meter, disconnect, press button to ZERO off, then press again for ZERO on, then measure). You can also ZERO with current flowing (i.e., ZERO with 10 amps, and then if the current rises or lowers, you will see the change i.e., +1 amp, -2 amps, etc.).

Note that DC current clamp meters can measure the direction of current flow (charging or discharging)--Take a bit of getting used to which direction to put the clamp on (i.e., flipping the wire direction changes the +/- polarity on the display).

-Bill

mevenable

we have two that style, one from home depot, one from lowes. I have been attaching the wires to the terminals on each battery, and on each end of the string.  The two give me readings that differ by more than 0.1 V on a 12 V battery, but supposedly that is well within the range of accuracy they guarantee.  Anyhow, I will try the stuff you describe

mevenable

sorry - the difference was 0.2 V for the four batteries - the lowes version (which we believe more) gives 51.3 at rest after 24 hours and the home depot version gives more like 51.1 V

BB.

I am specifically (last post} talking about voltage drop on each battery to battery cable under load/charging. Typically you set the meter to 2.00 or 0.200 (200mV) full scale.

Makes it very easy to verify the quality of you battery interconnects.

Bill

Photowhit

Hi @BB. your Uni-T meter link doesn't work, not sure how you find these short link names, but this is the meter I believe.

https://www.amazon.com/Uni-T-B4Q094-UT210E-Current-Capacitance/dp/B00O1Q2HOQ/ref=sr_1_1?ie=UTF8&qid=1522035201&sr=8-1&keywords=B4Q094-UT210E&dpID=41sD0XebR4L&preST=_SX342_QL70_&dpSrc=srch

BB.

Thank you Photowhit--I have fixed the link. I messed up editing the "shorter" URL...

By the way, after the ...HOQ item number, the rest of the link you really do not need:

/ref=sr_1_1?ie=UTF8&qid=1522038964&sr=8-1&keywords=Uni-T-B4Q094-UT210E

In earlier Amazon days, the "long link" would even take you the the personal cart information... Above, the link includes a GID (global ID) and other things like keyword search terms... I (usually) cut that stuff off for a bit of privacy/less tracking.

-Bill

mevenable

I have not been able to figure out how to use the clamp meter as you indicated

BB.

Which brand and model?

What are you trying to do?

Bill

mevenable

commercial electric MS2033C Clamp Meter, trying to follow instructions:  A very quick way to compare the "health" of your battery banks is to use a DC Current Clamp DMM (digital multi meter). DC Current Clamp meters are very safe and easy to use. You set the meter to DC range, zero the meter (DC clamp meter readings tend to drift over time/temperature) and then clip on one of the battery bank string's cables. (just one cable at a time, if you put the + and - cables both in the clamp, the readings add up to zero current --- I.e., -40 amps + 40 amps = zero amps).

If you see an issue (battery strings should be sort of close, but they will probably never exactly match charging/discharging currents--Something like 1/3 vs 2/3 current "balance" or greater, something is probably wrong)... Then you can use the DMM to measure voltage drop.

Again, with relatively high charging or discharging current, in a series string, you can measure the cable+connection voltage drops by placing one meter lead on the negative terminal of battery A and on the positive lead of battery B. You are now measuing the entire cable+connection voltage drop (again, with current flowing through the battery string). And typically, you are looking for something that stands out. You may measure 0.02, 0.05, 0.1 volts drop and find one cable that has 1.0 volt drop--That is the guy you need to really pay attention too (dirty connections, loose connections, etc.).

Photowhit

mevenable said:

commercial electric MS2033C Clamp Meter, trying to follow instructions:  A very quick way to compare the "health" of your battery banks is to use a DC Current Clamp DMM (digital multi meter). 

Unfortunately the MS2033C doesn't measure DC current. Often people sell clamp meters as AC/DC that just measure DC voltage.
You are looking for a clamp meter which also measures a range of Amperage for DC;

jonr

If you have a really good multimeter (like a Sanwa PC7000), you can also find current by measuring voltage drop  along say a 12" section of wire.

mevenable

great - I am going to the states for a couple weeks, I will pick up one of the good ones, seems like I need it