I'm trying to find a digital voltmeter/ammeter that will read both the charge and the discharge of a battery. The vast majority of the products available will only read current 'in one direction'. I need something that will indicate discharge by showing a minus sign, and will display the value thereof.
I'm wondering about getting one of these. I've never used anything that works by the hall effect though. Anyone know (a) if it will do what I want and (b) what snags I might encounter?
formatting link
The problem with using an analogue meter is that if it's scaled for 25A (which I need) small currents such as 2A hardly move the needle.
The only slight snag I have found using a clamp meter with a hall efect sensor is that they are slightly affected by the position of the wire inside the loop.
There will also be a constant small drain on the battery to run the electronics.
As a bonus, the percentage reading is usually wildly inaccurate, so you need to go by the voltage and accumualted amp hours readings..
and is the +ve -ve indicator just the single LED labelled "charging"?
4 digit display for Volts, and it looks like a 3 digit display for Amps The headline description says 0 to 30A, the spec says 0 to 9.99A, the label on the back of the unit says 0 to 20A
This article goes over some methods for estimating SOC (State of Charge) in a battery. Once it drifts into Kalman filters and Fuzzy logic, you know something is wrong :-) Finding such references tells you it's a hard problem to solve.
formatting link
Your laptop has a "fuel gauge", and the chip tracks how many times the pack has been charged. As the pack goes through charge cycles, its amp-hour capacity changes. Fuel gauge chips attempt to track this for you. Occasional "calibration cycles" help the unit in your laptop pack, from drifting too far off using its estimation methods. The "wide range discharge cycle" of Li packs, helps this process too.
This is an example of a fuel gauge chip. It uses the impedance sensing method, where the "stiffness" of the battery output is a measure of the state of charge. This device claims up to 1% accuracy, but that's not going to happen without a manufacturer of energy storage solutions, programming the device. It's not magical. It's not consumer ready. Battery gauges are meant to stay with the pack, and that's why this one has Flash Memory inside the chip, and it is good for 20,000 write cycles. Perhaps it would have some parameters written after each charge cycle. If the charge process was discontinuous ("cloudy periods" day), this thing might pooch on you after a while, because this one is meant more for laptop cycling (fast charge for two hours, use for N hours, repeat tomorrow).
formatting link
That fuel gauge chip has up to 16 LEDs for a readout and uses a shift register to drive more LEDs than there are legs on the chip for drive. Display driving then, is augmented externally. You might have noticed on some old battery packs for laptops, they have four LEDs, and if all four light, it's 100% full, three light, it's 75% full. And there used to be a button, to cause the LED display to light up (as you want the fuel gauge sleeping when not in usage, and using microamps of current).
In your lead acid application, you'd probably need all 16 LEDs and you would be walking the region from, say, 11 LEDs on to 16 LEDs on. Taking the automotive battery below 11 LEDs on state, would endanger the capacity measured as a couple hundred charge cycles. A deep discharge lead acid, could use more of the "range of LEDs on", but with some impact on battery life. When the above fuel gauge is used for LiLon, practically all the LEDs could be used. You'd have almost the whole display range at your disposal.
*******
Now, counting LEDs seems pretty hammy, as an interface. But, these devices either integrate coulombs, or use other methods to keep track of charge.
Reading the open cell voltage of the battery (neither charging nor loading), the first article says the battery has to settle for two hours, before a valid reading can be taken. And some battery chemistries have a rather non-linear relationship between voltage and SOC.
In the case of lead acid, it probably is a bit more predictable, but you need the application to be resting once a day, if you, say, expected to "validate" your SOC once a day. It would be pretty hard to rest the battery and make multiple determinations using that method all during the day.
*******
It's possible to know *precisely*, what state the battery is in. Using an electrolyte density meter (hydrometer), and temperature correcting the measured density, tells you the state of charge. The battery cannot be stratified, for this to work. Some batteries, the acid "settles out", and there is a layer of acid, with a layer of "distilled water" on top, in a sense. At an automotive shop, they sometimes "boil" a battery with their charger, in an attempt to reduce the stratification in the battery. (The gases released should be vented to open air - don't keep a maintenance-type battery in a shed, fill the shed with hydrogen, and then be surprised at the results.)
So we could measure density, assuming the battery is reasonably healthy. The local conditions in the cell, would be just as dynamic as the "SOC by open circuit voltage method", in the sense that if you were pumping 20 amps into the battery with your solar cells, the gas bubbles rising off the plates, screw up the density reading. You might consider then, that the battery still has to "settle for two hours to take a reading".
And I'm not aware of anyone making a continuous density readout as an electromechanical device of some sort.
12M H2SO4 is a pretty intense environment for a sensor to work in.
*******
OK, so you've intuitively settled on your own coulombometric method. Looking at the amps, counting them, working out how many hours it will take to fill. In terms of fuel, the battery gives back energy at a variable rate. When using a heavy discharge, the capacity of the battery is reduced. If running a 50mA LED off the 12V battery, well, it probably runs for days on end, and you also "milk the max amp-hours" from the thing that way. Maybe you get half the amp hours, by running a 20A motor load instead. Think of it as "efficiency", where the battery gets a bit warm and wastes some of the energy as heat, when the motor load runs.
So if you take readings in that sense, you need to take the readings with sufficient frequency (for integration purposes). And you need to store the results somewhere while working this out. Some of the Amazon products with an "amp-hour" readout, are doing this sort of simple-minded, bipolar "amps-in" versus "amps-out". And this is a hopeless method, because when a motor loads the battery, the "amps-out" are a penalty on the process. Whereas proper fuel gauge chips, they take more specifics of the battery chemistry into account (the academic paper up top, had some quadratic function to estimate the relationship). But they also require someone qualified in this sort of mumbo-jumbo, to set the parameters for the battery type. The "bank balance method" used by the Amazon hall probe meter, doesn't take the battery efficiency into account when doing coulombimetry.
*******
As John Williamson points out,
"the percentage reading is usually wildly inaccurate"
he is bang on, and I hope that doing a few Googles, you'll find other papers on these black arts. You can pretend to build a fuel gauge, and some chemistries make this easier than others. Perhaps even a purchased "energy bank" would come with a fuel gauge fitted to the unit.
But if you expect to bodge this with your favorite Amazon panel meter, it's a solar charger (all over the place on a minute by minute basis), you would be hard pressed to reach any precise conclusions. And the open circuit voltage can't be used to sense SOC, because the battery might only be sleeping at night. And you'd only get the one SOC calc per day that way. At 3PM, you might not have a clue precisely how much fuel is left. If the motor (pump) runs all night, on a cyclic basis, now the open cell voltage is modified in the negative direction. I tried this on my car battery, and got all sorts of inappropriate info by doing that (I would have concluded the battery had expended the 25% energy it was safe to extract because it was reading 11 volts or so).
As a consequence of all these observations (this isn't the first time I've lightly researched this), my conclusion would be:
1) You're getting a plus/minus amps meter here. Time to celebrate! This is the absolute minimum equipment for a solar charger dude.
formatting link
2) Don't bother staring at the SOC display at the top. Rubbish.
3) If the current measures zero for 2 hours, measure the OC voltage on the right, measure the air temperature around the setup, consult temperature correction tables or use the equation, compute the normalized to 25C OC voltage value, and determine from that, how much fuel is left. Maybe don't run the battery below 11V or so. The nominal might be 12.6 or so. Right after charging, it could be 13.8V for a while, but then we allow hours and hours to pass for the battery to settle properly, to discover it is 12.6 at 25C and it's um, "full".
4) With Lead Acid systems and these sorts of half-assed estimation methods, you could be off by as much as 50%. That's my result from working with automotive batteries casually. Temperature compensation is important. Settling time at zero load is important (if using open circuit SOC estimation methods).
Purchasing the above product only gives you an approximate handle on what is going on. It's better than nothing, but it's much worse than a properly engineered fuel gauge for the particular battery you're using. That amazon product has the little battery symbol in the upper left corner, but it's not subject to the careful maths that the ti.com device is doing. And that TI device is impedance based (or so they claim), while other ones are colombimetric but with corrections based on the chemistry. The corrections are different for Li than for Pb.
*******
Take a look around too, for more recommendations on end-voltage.
formatting link
Actually, 12.7v is considered fully charged, and 11.4v is considered fully discharged, for lead-acid batteries.
[At 25C. Don't forget temperature compensation...]
And the 11.4v value, would be measured when the battery has rested for two hours. Makes it kinda hard, while the motor is running, to figure out what voltage value to use to open the relay. Maybe if the battery was still stiff at that point, it might be reading 11.0V while running the motor, then if stopped for two hours to rest, it might rise to 11.4V.
This site has good articles. And it has yet different values for percentage charge.
formatting link
100% 12.65 \ 75% 12.45 \ 50% 12.24 \___ Notice we need a 3 1/2 digit DMM for voltage. 25% 12.06 / The Amazon meter does not have 3 1/2 digits for voltage. 0% 11.89 / The current measurement requirements are different
But they would need approx 2V to 3V across them to work. So if you are considering just wiring them in parallel with the shunt the shunt would have to drop 2V across it when the equipment takes 2A. When the equipment takes 20A the volts drop across the same shunt would be 20V, but long after the equipment stopped working due to a lack of a power supply from the 12V battery :)
Huh!? Just about every digital meter I have ever come across works 'both ways', you just get a number (for +ve) or a number with a - sign in front (for -ve).
My little 4 digit displays that I use with my BeagleBone black are (if I remember right) 100mV full scale, but that means +- 100mV and they have a - sign.
What I've found too. Although don't have a brand new DC clamp meter. They're fine for large currents, but near useless for measuring small (less than say 10A) accurately. Which is mostly what I need.
I think Bill is conflating battery current charge / discharge with capacity (or has two different projects).
I have a similar issue in that I have an 'in-line' power meter that logs watts and monitors volts and amps that I want to use on a mobility scooter but it seems to be unidirectional, in that it has a 'Source' and 'Load'.
So, whilst it will log the power drawn from the battery, I think it's a cumulative thing till reset and wouldn't monitor any power being put back into the battery.
I have (somewhere) something I bought a while ago that is in two parts, with the base part being a shunt and a remote (wired or wireless) that shows all the data. I think they were designed to be used on the likes of narrow boats with solar panels so you could both monitor / log the energy in / out the battery and do so remotely.
I believe that also 'learns' the battery capacity (an issue I believe I've been seeing when charging the scooter batteries (two scooters) on the bench rather than though the scooters themselves.
Just checked with mine - and you do get a minus sign when you reverse the leads. Not something I was sure about, as apart from with a charging system like on a car, not usually important.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.