Battery Charger Testing Results

[JHZR2]

Like a number of us on here, I have a number of battery chargers. I keep a bunch because I have a number of cars that dont get year-round use, and I also have some for maintenance.

Because there is now an Automotive Electrical subforum, and because there are a lot of good questions, discussions, and general interest on battery chargers, I thought it would be interesting to test a few, to better observe and discuss how they operate.

I intend to test a number of chargers that I own, as time allows, and invite others to test them to if they are able and so inclined. I intend to edit this thread heavily, questions and good discussion welcome, but unrelated questions should be asked separately.

I wont get the testing perfect. I dont have a battery cycler to extract exact amounts of energy from a specific test battery. Who knows how this will go in the end. But its a shot to see how these thing work, one after another, as time and energy allow me to do it.

The testing will be done using two Owon B35T+ Bluetooth Datalogging multimeters. These are cheap and available on Amazon. I am not looking for low mA accuracy, so will take current readings through the 20A sensing port. Data is taken at a pre-defined interval, and the CSVs that are pulled from the meters will be processed in Excel. Ill try to keep the charts consistent. In the end each chart can be tens of thousands of data points, but its nothing that a computer cant handle. My time and energy to manipulate charts may be a different matter though...

Example of setup in vehicle:

Meter Bluetooth GUI example:

1/10/21 Edit: Going to use a spare Group 29 DC battery that I have in my garage. It read 12.6V initially, but to load it, I am using a 150W inverter an a 4.5W and 25W LED (just whats convenient in my garage). Voltage quickly dropped with load, but this plot shows a very consistent voltage when the 4.5W load was applied. At the end, far right, is when I used a 25W LED.

Efficiency of the load appears to be ~ 60% for the inverter and light. Note the image below, that shows 3.27A @ 11.72V, which is 43.6W. I don’t know that this battery is the best as a result, but it’s what I have to easily and consistently test in my garage.

Im going to reserve a few response windows right below this for future tests. When I get as far as Im going to go, I can always delete the rest. Each reply window only allows a certain number of embedded images, so I want to reserve enough for my anticipated use.

Please keep unrelated questions/comments to a minimum in this thread. Non-specific discussion, not associated with the details of the chargers and their voltage/current profiles will be removed at my discretion to ensure consistent discussion and flow of relevant information. No hard feelings.

Enjoy, and I look forward to good discussion!

 

[JHZR2]

Test #1: NOCO Genius 2A charger on Group 51 Honda Accord Hybrid Battery. 6-7 Jan 2021; ambient conditions high 30s-low 40s.

For whatever reason, my Accord Hybrid doesnt charge the 12V battery up much, so I routinely come upon the battery sitting in the low 12s, or even high 11s. This is fully discharged, but it doesnt matter much to the car, because the hybrid battery does all the engine start and provision of hotel power. The 12V battery is really just there to unlock and keep the memory mostly, I guess. So the battery was at very low voltage from sitting for a while over the Holidays. I decided to use the 2A NOCO because I didnt want the test to take forever (I have a 1A and 2A model), and there have been a number of recent threads about the NOCO chargers.

Near-dead battery, starting to charge:

I connected the two datalogging meters up to the battery, with the one measuring current connected in series with the charger. I initially set the data acquisition to 2s per sample, but the meters only have a 10000 sample data bank each, so that became an overwhelming amount of data. Later on I changed it to 4s samples so I could go 11 hours between restarts. Lessons for the future I guess...

Here is the result:

It is interesting to observe all the voltage drops. These arent singular points, they are points over a few minutes time, indicating that the charger must allow some sort of a test to look at the relaxation of voltage. It doesnt seem to be a regular interval, so it must be based upon some prescribed event or logic. One can immediately note that no time is spent at a constant voltage level. Towards the end (right side), it holds just north of 14.5V for a while, but we do not see the logarithmic decrease in current that would be expected over a long-term CV charge. It is also interesting to see how as the low (0.2A) charge progresses along the right side of the chart, the voltage is manipulated down by the charger, while constant current is retained.

I suspect that this is all in an abundance of caution so as to not damage the battery or a vehicle's electronics, and to avoid high rates of corrosion due to charging at elevated voltages for long times. I am surprised that they didnt allow for at least one logarithmic type current decrease at constant voltage though. Overall this is "good enough" for an SLI battery, and will keep it from being overcharged or from significant sulfation occurring. It is a conservative charge, with limited opportunity to spend time at elevated voltages.

NOCO customer service indicated that after it reaches full charge (the colored lights will all sequentially turn on as it charges, then it will progress to flashing green only, then solid green only, then a very slow blinking green, which I didnt achieve), the charger turns off the current, and watches until it observes 12.8V +/-0.2V. 1/8/2020 EDIT: I put the meter on measuring voltage only, to see how it would operate in maintenance mode. Twice I opened the car door, which causes the battery voltage to drop, spurring the charger to top off. The other times it did it on its own.

Here are the results:

The logic consistently turned on the charge at 12.77V, and turned it off at a surprisingly high 14.8-14.85V. Granted it was a high of 36F today.

Feb 12, 2021 update. I had re-run the Noco unit on the standalone battery, and now have much smoother/better data.

Test #2: Battery Minder 12248 @ 8A. Cold conditions, 25-45F, in unheated garage.

This test used my Group 29DC battery, described in the original post. I ran the test at 8A to see how it would perform. My observation is that the higher the current, the faster the first cutoff voltage is achieved, and this sort of "tricks" the charger to rapidly reduce to a more conservative ramp-up.

I ran the test to see what the charging profile would look like, and then at the end, tried to play with the thermistor a bit. I put the thermistor in icewater and also in my hand. Interestingly, I observed that if the charger saw a rapid change in temperature, it took the current to zero. I observed as the low current charge progressed, that the current swung rapidly. A "zoomed in" profile of this (during the timeframe I was doing some of the thermistor tests) is provided. It was interesting to observe that when the thermistor came back "into spec", the charger turned on for 0.8-1A, and then rapidly settled down to the voltage and current it was at before.

 

[JHZR2]

Test #3: CTEK MUS 4.3, Ambient 28-43F, inside of unheated garage, Group 29DC battery.

Used my CTEK MUS 4.3 this time to see how it charges. Got a very nice, smooth charging profile, but when it went to float, it went on and off like others have observed with other chargers. Seems like this proifile is simpler and more traditional; not sure if there is other logic built in. But seemed to do a good job.

Next up is the Battery Tender Classic 1.25A Charger (metal box). This one seems to have a different charge profile, it I think it may just be an artifact of the fact that it is so current limited, it doesn’t need/can’t generate enough overpotential at low/early SOCs. It also has an interesting float. When the charger reached higher voltage levels, there is some funny behavior in there, not sure why. I assume it has something to do with the desire to avoid hitting max voltage and current, perhaps a thermal limitation on power? Interesting that it starts dropping current in an expected logarithmic fashion, but allows the voltage to keep creeping up even as the current that the battery is tending down. I would think this would give a more complete charge than some others that we have seen, since it is capturing the end part of that logarithmic decay, at elevated voltage. The voltage is pretty high though, even for it being in the high 20s to low 40s. Thus there is chance of higher levels of gassing and corrosion. Granted, to generate gas, you need more coulombs, and a 1.25A charger cant charge and generate a lot of gas because there's just not enough amps to drive it.

Zooming in on float:

 

[JHZR2]

12/26/21 - it has taken me a long time to focus on this. Too much other stuff going on. I had some archived data from the Tankpow 2/4/8A charger, that I tested at 2A last February.

This seems like a pretty simplistic charger. The charger has a thermistor lead on it, and seems to have a temperature compensation capability.

My interpretation on this one is that it displays the setpoint voltage, and then works to achieve it through a CC or a tapered CC charging scheme. Not much else. But it is probably "good enough". What Ive seen is that it "times out" and gives an error when Ive left it attached to my Ram for weeks/months at a time. I dont know what that means. The truck is always at 12.7/12.8V anytime I check on it. I suspect that at some point when it senses a voltage below 12.8V, it will push some more current into the battery to get the voltage right. That's something that takes long enough that I can't pull data for that long to find it. But I doubt that the profile is different than what we see above.

12/28/21 - Optimate 6

I know this is one that @UncleDave has high regards for. To me it is less intuitive to use, in that there really is no control to current limit, display voltage, charge status, etc. I put this one onto one of my Ram Diesels the other day, and noted that the voltmeter/USB charger never showed higher than 13.8V, even though I knew that the battery had been low due to spending an extended amount of time dealing with a stereo install and the interior lights on. So I decided to test it, and its the first new one Ive done...

First the truncated data. First, I was surprised at how little current this unit provided into the battery. I dont know why it limited itself to ~2.5A when it is a 6A charger. After this test was done, I was messing around with things, including using the test battery as a power source, and turned on the charger through my ammeter. I saw >4A at that point, so there is the ability to provide more current, thus Im curious why it didnt. This would also explain why when Ive used this in the past, I dont see as fast a progression towards a real charging voltage, as Id like.

It has selectable output voltages, I had 14.4V selected, and it looks like it achieved that fine. What I dont understand is why it abruptly cut off the charge current at the very start, and then ramped it up so slowly with current. Since there was more than enough current capability in the charger, it could have accelerated into a CC charge at a higher rate, instead it gave an initial impulse, then crept the current up with voltage, and then back down. The logic in this must be to titrate the battery with increasing voltage and current, as part of the "smarts".

I dont really understand what its doing on the subsequent long-pulse charging operations. As you can see, they are changing in time, my guess is because of temperature compensation. My one logging meter had a dead battery after a while, so there is a gap in data, but the trend can be observed.

As of now Im attributing it to temperature compensation and the battery indeed getting to a slightly higher SOC. Overall my feelings on this one are still mixed. It didnt appear to go through all charging modes, and severely limited current going into the battery, with no real way to control it other than to disconnect/uplug and plug it back. I think this is a trustworthy charger to leave alone, but its hard to keep tabs on what the status is compared to some others...


12/30/2021 Update, tested a ProLogix2320. This test was slightly different. Instead of the group 29 battery I had been testing on, I needed to charge my group 49 in my 1982 Mercedes 300CD. It has a recent East Penn battery, in good condition. It had sat without a charger for about two months, I didnt realize I hadnt connected a charger.

I did a quick start at 2A because I wanted to test the leads and setup. The current went right to 2A, and voltage crept up slowly (from ~12.22v to 12.4V); I left it on for just a few seconds. When I got the data logger ready and started at 10A, the voltage very rapidly jumped to >14V. You’ll notice in the image above, when doing the early stage charging, the charger shows/claims 14.7V in the display, while the actual meter measurement was lower. In time they converged. So not sure if this is due to location of placement of the voltage sense, display of a control value vs measured value, or what.

The profile is like others, where in my interpretation as the dV/dt and voltage hits too close to to the setpoint too rapidly, it pulls current back as part of its "smarts". You can see that it wanted to give 10A, but with the voltage rising rapidly, it pulled the current to watch how the titration of current affected voltage. Then it started to creep it up a bit more before allowing the natural logartihmic decay.

So its not a CC charge really; I thought this one would be. I think that these chargers all pull back too quickly, Id have preferred to see it let the voltage quickly get to 14.7 and then sit at CV and decay down. But you can see that most all of these chargers do this - they operate to slowly ramp voltage and then go to a lower current setting and track the voltage rise. That said, I cant fault this charger as a one stop solution with a lot of good features such as all the control settings you might want, a voltage readout, settings tuned to AGM/gel/standard batteries, etc. It doesnt use SAE plugs, so its more of a once in a while topping charge unit, not a float/maintenance charger, but it works well.

I ran it all the way to the Exercising Phase (EP) that @Quattro Pete asked for. I had never ran this unit long enough to see it, usually I used this unit as a 10-20A charger to revive a low battery quickly, or just give a quick boost.

Here are the data all the way through the EP phase:

Looks like it does a minimal discharge just to get the voltage down enough to accept a suitable charge. Then a 2A charge and a similar cutoff and re-test to before. I believe we are seeing some temperature compensation in this.

Here are data "zoomed in" on the EP phase:

The discharge was only -0.3A. You can see that since the voltage doesnt peak quite so fast at 2A vs 10, that it does stay in CC for a while. It doesnt have to pull back to titrate dV/dI and dV/dt to prevent an overvoltage condition.
Im going to run the 2A setting on the group 29 battery at some point.

 

[JHZR2]

During the cold spell a few weeks ago I had interest in a cheaper charger I could leave outside and maybe mount to the house to just have a convenient unit in one spot. As I don’t know the longevity of it, and this would be used to just boost the battery on cold nights or when sitting a few days, I wanted to start cheap. I’m not going to leave it connected for months like I do with others on some of my other vehicles.

The Mroinge unit looked to be reasonably small, had a voltage readout, and enough current to charge a battery in the 75-100% range, etc. This unit was around $35.

I did the data acquisition on a group 49 battery in my 1993 MB 300SD. Not my usual test battery. It had been somewhat depleted because there is a parasitic load of around 600mA associated with one of the grouped relays that coordinate windows, steering tilt/telescope, seats, etc.

The results are as one would expect. It does keep the battery on a “float” voltage which doesn’t much matter for my intended use but may for others. Again, I’m not saying this should be a long term storage charger, for a few reasons. But it works for its intended purpose.

The charger works like many others in that it keeps the CC period fairly conservative in terms of cutoff (looks like a bit under 13.5V) and then starts to taper the CC as it titrates di/dV. The cutoff point on this charger is pretty conservative, stopping the charge around 14.2V.

The charger indicated full; as you can see, it continued to source around 0.6A which is consistent with the parasitic I see when I dont have a specific fuse in the rear fusebox pulled. I need to diagnose that, currently its easier to pull it (car runs and operates fine as long as Im not trying to move the seat or the steering wheel), because I justhad no time, and it was too dark, too cold... Now that spring is here Ill fix it...

As a temporary use charger to boost batteries in cold conditions, and as a cheap option that has some OK features, this unit is OK. It has spent the last month or so outside, in the rain and snow, sometimes plugged in, sometimes not, and is no worse for wear... So there's that...

 

[JHZR2]

Reserved

 

[spackard]

Not to distract from your NOCO 2A, but I recently bought a NOCO 15A, hooked it up to a mostly-charged 151R (very small car battery) and I didn't like the charge curve at all. I didn't use a logging multimeter; I watched with a current clamp meter and saw "many dips and restarts". So much so I felt I got a bad unit. I unhooked the NOCO and went to the Battery Minder 8A I have to see what the current curve on that one looked like. It was a much more conventional charge curve. So, it's not my clamp meter, or a wonky battery.

Also, the cable is a different size than the other NOCOs, and, one cannot buy any lug adapter cables anymore for it. It really feels like a bastard model in their line. But, it has a useful 13.6V power supply mode, so I'll keep it just for that feature.

JHZR2 - PM me with your address if you want qty 2 of GC002. I have no use for them and will ship them to you free for your work on this thread, among other things you do for the forum.

 

[Quattro Pete]

Not to distract from your NOCO 2A, but I recently bought a NOCO 15A, hooked it up to a mostly-charged 151R (very small car battery) and I didn't like the charge curve at all. I didn't use a logging multimeter; I watched with a current clamp meter and saw "many dips and restarts". So much so I felt I got a bad unit. I unhooked the NOCO and went to the Battery Minder 8A I have to see what the current curve on that one looked like. It was a much more conventional charge curve. So, it's not my clamp meter, or a wonky battery.

My Noco g7200 behaves weirdly like that, too. I no longer use it.

 

[JHZR2]

Not to distract from your NOCO 2A, but I recently bought a NOCO 15A, hooked it up to a mostly-charged 151R (very small car battery) and I didn't like the charge curve at all. I didn't use a logging multimeter; I watched with a current clamp meter and saw "many dips and restarts". So much so I felt I got a bad unit. I unhooked the NOCO and went to the Battery Minder 8A I have to see what the current curve on that one looked like. It was a much more conventional charge curve. So, it's not my clamp meter, or a wonky battery.

Also, the cable is a different size than the other NOCOs, and, one cannot buy any lug adapter cables anymore for it. It really feels like a bastard model in their line. But, it has a useful 13.6V power supply mode, so I'll keep it just for that feature.

JHZR2 - PM me with your address if you want qty 2 of GC002. I have no use for them and will ship them to you free for your work on this thread, among other things you do for the forum.

Thats interesting - without logging, how did you see dips and restarts? How did you determine they were an issue?

My suspicion is that they are essentially like titrations in chemistry. These electronics can push voltage and current independently up to their ratings. So they can drop the voltage, and as long as current doesn’t drop, then they can extrapolate state of charge and other characteristics such as impedance. I suspect that if they watch something like dI/dV, and it’s small or zero, then they know something. And if they take the voltage down a certain amount and see no change, it means something.

I’m inclined to test my Battery Minder 12248 next. I put it on our van which had sat a while over the holidays, and was surprised how fast it went to blinking green float charge. I didn’t have the meters on at the time...

So there’s something to be said about not trusting a charger when it says it is done, when charging fast. I think that if your algorithm doesn’t allow long times at high voltage, and you have a high charge rate (which requires more overpotential to drive it), it’s hard to keep at the highest voltage levels, and they inadvertently trip to lower states.

 

[spackard]

> Thats interesting - without logging, how did you see dips and restarts? How did you determine they were an issue?
I just watch the clamp meter for 10 minutes, mentally drawing the graph, and looking for patterns. Seems it had 4 current step increases (say 2, 5, 7, 15 amps), would trip to 0 amps during the 4th step, wait 30 or so seconds, and repeat. As time went by it spent more seconds at 0 amps.

I started thinking how many amp-seconds is really getting into the battery and maybe a lower current for more seconds would charge faster, switched to the Battery Minder, and went from there. Only when the Battery Minder gets to full and goes into float do you see on-and-off current transitions, which I expect to see there anyway.

 

[IcarusLSC]

Not to distract from your NOCO 2A, but I recently bought a NOCO 15A, hooked it up to a mostly-charged 151R (very small car battery) and I didn't like the charge curve at all. I didn't use a logging multimeter; I watched with a current clamp meter and saw "many dips and restarts". So much so I felt I got a bad unit. ...

Don't all the newer NOCOs do this as a 'Recovery desulfation process' to start with?

NOCO - Understanding Charge Steps - Support

Understand the charging steps that will occur while using a NOCO charger.

no.co

 

[Quattro Pete]

Don't all the newer NOCOs do this as a 'Recovery desulfation process' to start with?

NOCO - Understanding Charge Steps - Support

Understand the charging steps that will occur while using a NOCO charger.

no.co

Mine did it at the end.
It was not in any kind if repair mode.

 

[JHZR2]

I’m not sure that varying voltage is a problem. I’m not sure that letting the current go to zero is an issue. This may be a way to sample the state of charge, reduce gassing/support recombination, etc.

 

[y_p_w]

I’m not sure that varying voltage is a problem. I’m not sure that letting the current go to zero is an issue. This may be a way to sample the state of charge, reduce gassing/support recombination, etc.

What does it look like in real time? It was totally nonscientific, but I've used a multimeter with a microprocessor controlled charger and it was showing the voltage constantly oscillating. I believe that "pulsed charging" is pretty common these days for any number of reasons.

 

[ripcord]

I don't have a NOCO charger, but the graph on the first post looks just like what I observed with a volt meter and a DC amp clamp with a few different smart chargers I have.

I think the strategy these "smart" chargers use is to push constant current levels and use inductance and voltage levels to monitor the battery and stepping current levels down until the battery is more or less fully charged.

The charge cycle basically works like this: if the charger detects the battery is low, it will charge at it's highest current level until the battery reaches a certain voltage (~14.5v), then stops charging to test (via inductance probably), then starts charging again at a lower current level and repeat until it takes ~14.5ish volts to push a very small amount of current into the battery. One 4 amp charger I have charges at 4 amps, up to 14.6v, then repeats with 2 amps, 1 amps, and .5 amps before going to a float charge of 13.4 volts.

 

[circuitsmith]

I initially set the data acquisition to 2s per sample, but the meters only have a 10000 sample data bank each, so that became an overwhelming amount of data. Later on I changed it to 4s samples so I could go 11 hours between restarts. Lessons for the future I guess...

I think 30s samples would be adequate for this.

 

[JHZR2]

I don't have a NOCO charger, but the graph on the first post looks just like what I observed with a volt meter and a DC amp clamp with a few different smart chargers I have.

I think the strategy these "smart" chargers use is to push constant current levels and use inductance and voltage levels to monitor the battery and stepping current levels down until the battery is more or less fully charged.

The charge cycle basically works like this: if the charger detects the battery is low, it will charge at it's highest current level until the battery reaches a certain voltage (~14.5v), then stops charging to test (via inductance probably), then starts charging again at a lower current level and repeat until it takes ~14.5ish volts to push a very small amount of current into the battery. One 4 amp charger I have charges at 4 amps, up to 14.6v, then repeats with 2 amps, 1 amps, and .5 amps before going to a float charge of 13.4 volts.

Its interesting that you saw this with a meter and amp clamp. It is quite true that many chargers use a PWM scheme. So a 10A charger might push 20A, 50% of the time. It’s also interesting what you say of the logic approach. CC/CV is a common, and notionally easy approach to administer, and Id suspect faster than the scheme that you mention and I observed. I think it has to do with safety (gassing and other issues) and corrosive issues at elevated voltage.

What does it look like in real time? It was totally nonscientific, but I've used a multimeter with a microprocessor controlled charger and it was showing the voltage constantly oscillating. I believe that "pulsed charging" is pretty common these days for any number of reasons.

Im surprised if you could see PWM with a multimeter, but I do agree it’s used.

My Noco g7200 behaves weirdly like that, too. I no longer use it.

Not to distract from your NOCO 2A, but I recently bought a NOCO 15A, hooked it up to a mostly-charged 151R (very small car battery) and I didn't like the charge curve at all. I didn't use a logging multimeter; I watched with a current clamp meter and saw "many dips and restarts". So much so I felt I got a bad unit. I unhooked the NOCO and went to the Battery Minder 8A I have to see what the current curve on that one looked like. It was a much more conventional charge curve. So, it's not my clamp meter, or a wonky battery.

Flagging these two because of my edits, including more data posted above. Taking data with it in maintenance mode, you can see a much smoother operation, with none of those drops. Also can notice that the controls allow for a much more rapid rise to the upper voltage limit.

These data are substantially smoother. Because of that I believe that those dips are all purposefully done as part of the logic, to sense voltage relaxation and better characterize SOC. I don’t think the drops observed are an issue, my guess is they’re some sort of analysis and control.

 

[Jetronic]

The optimate chargers we use at work do test the batteries regularly and show this on the led readouts. Likely a very similar thing to what the Noco does as the charge current is then also 0

 

[ripcord]

Im surprised if you could see PWM with a multimeter, but I do agree it’s used.

I wasn't observing pulse width modulation... I don't have the equipment for that. What I was seeing was constant current charging, with the current stepping down each time the voltage reached 14 and a half volts and after short periods of zero current where I assume testing is happening. This seems to get the batteries close to a full SOC (~1.275 SG), but they take a long time to do it.

 

[y_p_w]

Im surprised if you could see PWM with a multimeter, but I do agree it’s used.

The voltage was just jumping around from 11V to 16V. Like maybe every 10 seconds.