Batteries Charging Routine

[Woosh]

Stuart reported his own observations on his battery and his charger. His method is reproducible if you have a 5ah pack. I don't,
so i can't reproduce his results.
Have you looked at sturmey's post #111?
He explained Stuart's results very well.
You can see differences even bigger than 24% on discharging depending on how fast you discharge, so why the same not happen on charging? Flecc explained that fast charging his car gave a sharp drop of mileage as soon as you start using the car, independently from the voltage.

 

[flecc]

You'll see that the capacity can be different depending on the way the battery is charged. Flecc made the same point with his car.

More information on that having done more miles today from that slow charge (2.1kW rate). There's two methods available to assess the capacity gain over the normal charge rate at this part discharged point.

The first is miles covered against the percentage capacity the car tells me has been used. That gain in mileage and therefore capacity over using the normal speed (6.6kW rate) charger is currently 10.56% .

The second method is comparing the miles covered against the car's predicted miles still available. That gain is 4.35%.

I won't be fully discharging the battery, but the truth once circa 70% discharged will be between those two.

What matters more to me though is that as well as getting getting more capacity from each slow charge, the battery still meets its specified capacity and range when new at 5.5 years old. That's thanks to well over half of all the charging during the 5.5 years being unconventionally at the slowest rate and almost none on 50 kW Rapid chargers.
.

 

[StuartsProjects]

Take the example of a single lithium cell.

There are numerous chargers out there that will keep applying charge current till the voltage across the cell reaches 4.2V, then charging stops.

For sure when the battery is then allowed to rest a bit the battery voltage falls to say 4.15V. Most people realise that, including I.

Say you charge a lithium battery at a set current and stop charging when the volts across the battery is 4.1V and then measure the capacity. If you then repeat the test but charge to say only 4.0V then discharge, you are clearly measuring the difference in cell capacity between a charge to 4.1V and a charge to 4.0V. It does not matter if the instruments used are not superbly accurate, or what the battery rest voltages were since the same instruments, charge current, discharge load and test method was used for both tests, so its a valid comparision.

Does anyone know of a lithium charger that charges the battery to 4.2V, stops for a few minutes, measures the rest battery voltage (which will have dropped a little) and then uses that rest voltage to determine if the battery is fully charged ?

 

[guerney]

Does anyone know of a lithium charger that charges the battery to 4.2V, stops for a few minutes, measures the rest battery voltage (which will have dropped a little) and then uses that rest voltage to determine if the battery is fully charged ?

At this rate, you'll end up making your own smart BMS at your ORAC factory. I'd buy one. Is there a pre-order form and do you accept Paypal? I'd like lights indicating when the balancing phase starts and stops.

 

[StuartsProjects]

I have no reason to think that he did not measure voltages correctly.

I assure you I could measure voltages for several years even before I spent 3 years in college doing electronics.

 

[StuartsProjects]

At this rate, you'll end up making your own smart BMS at your Orac factory. I'd buy one. Is there a pre-order form and do you accept Paypal?

I accept Paypal.

This is where I used to work in the early days, I only discovered the existance of the video a couple of months ago. I ran the Electronics lab, fixed the difficult stuff, checked components for suitability etc. Not many eBike using Lithium batteries in those days mind.

 

[StuartsProjects]

A question.

The constant voltage charge at the end of a Lithium battery charge, starting at say 4.18V per cell, adds in my tests around 7% of the capacity.

I have a generic Solar panel type adjustable charger, but although it goes into the constant voltage mode at the end of the charge it does not turn off as such and it hovers there adding short burst of a few mA every few seconds. A limitation of the basic design presumably.

So is there a reasonably priced adjustable charger out there, that when the set voltage is reached, and the constant voltage part of the charge drops the current to a few mA the charger actually stops all charging and would need to be reset ?

 

[saneagle]

I assure you I could measure voltages for several years even before I spent 3 years in college doing electronics.

You're reading your meter OK, but you're measuring the wrong thing.

 

[Sturmey]

So is there a reasonably priced adjustable charger out there, that when the set voltage is reached, and the constant voltage part of the charge drops the current to a few mA the charger actually stops all charging and would need to be reset ?

Some charger can be adjusted. But there is a cheap hack that works without doing this (but not perfect). That's to put one (or two) schottky power diodes in series with the charger DC power lead to the battery. The diodes can be easily switched to short out so you can have either 2, 1 or none switched in series. The actual effect depends on the characteristic of the diode. The ideal situation would be to have an almost vertical forward voltage curve. Because this is not the case, each diodes will initially reduce the charger voltage by .6 volt at 2 amp but if you leave the battery on for a long time this voltage drops as charger current drops.
One side effect of this is that the green light doesn't come on for a long time but you dont have to wait for this. I use these and in practice I find it works very well for me. Silicon power diodes or any power diode ( of say 5amp +) will work but the voltage drop will probably be higher. Its easy to experiment in this case. One diode may be enough.

 

[Woosh]

Don't you think it's a lot simpler to take the bike out for 5 minutes after charging to drop the voltage to 41V or less than modding a certified charger?

 

[Sturmey]

Don't you think it's a lot simpler to take the bike out for 5 minutes after charging to drop the voltage to 41V or less than modding a certified charger?

No. You have already saturated ( and stressed) the cells at 4.2 volts before you dropped the voltage and anyhow, with a little imagination, you do not need to modify the charger.Just get a matching plug and socket along with diode and heatsink and container as below. The diode can be used or not used as required by extending the charger lead with as shown below.

 

[StuartsProjects]

But there is a cheap hack that works without doing this (but not perfect). That's to put one (or two) schottky power diodes in series with the charger DC power lead to the battery.

I had thought of that, and tried it, but on the charger I was using, my original Swytch Brompton one, the charger did not detect the battery, so I did not try any further. I suspect this charger does not turn on until it sees the battery, and with the diode in place there is no battery voltage to detect.

I just tried it on one of my other charger and it does work with the diode in place, I will see what happens on a full charge.

I will also try the original Brompton charger with two schottky power diodes, some of these types do tend to have a significant reverse leakage, and may allow the charger to see the battery.

 

[StuartsProjects]

Don't you think it's a lot simpler to take the bike out for 5 minutes after charging to drop the voltage to 41V or less than modding a certified charger?

Dont see the logic of that, charging to 42V and then discharging to 41V, is just not the same as charging to only 41V.

No hack to a charger needed, just make up an Anderson lead with a diode in-line to go between charge monitor and battery.

 

[Woosh]

Dont see the logic of that, charging to 42V and then discharging to 41V, is just not the same as charging to only 41V.

No hack to a charger needed, just make up an Anderson lead with a diode in-line to go between charge monitor and battery.

The logic is you follow manufacturer's instructions with guaranteed results while reconcile your wish to lower the storage voltage at lower than maximum. If you skip the constant voltage charging you will damage your battery prematurely for certain.

 

[Sturmey]

Re using 2 Schottky diodes, one thing I only briefly mentioned is the delay in the green light coming on. In the example given in graph above, it will be noticed that although the forward voltage is .6v at 2 amp, this voltage is lower and is .4 volt at .2 amp which is about 10% of the chargers output and is often the point when the green light is set to come on (I think this may depend on the charger). This would indicate that the overall effect of using 2 schottky diodes would be to reduce the charge voltage by .8 volts when you wait for the green light.
Adding the two Schottky diodes somewhere in series with the battery charging has little or no effect initially to the charging. Its effect kicks in later on at under 40.8 v (for 36v battery) when the cells go into the constant voltage floating stage. The voltage rise slows down then but continues to rise very slowly in this stage (when ideally it should not) with the current reducing. Eventually the voltage reaches about 41.2v when the green light comes on. I often disconnect the batteries at about 41 volt or when the charger becomes noticeably much cooler as the final stages are very slow due to the effect of the diodes forward voltage changing with current .

 

[StuartsProjects]

The charge with a Si diode in series does work and the battery charges to 41.2V or so but the constant voltage stage takes around 2 hours to complete versus about 1 hour when there is no diode and the battery charges to 42V.

This is a 5Ahr battery pack and charging at 2A is a bit much for long life, so I have a 1A charger on the way to try.

 

[Woosh]

The charge with a Si diode in series does work and the battery charges to 41.2V or so but the constant voltage stage takes around 2 hours to complete versus about 1 hour when there is no diode and the battery charges to 42V.

my guess is the CV current is much less with 41.4V feed compared to 42V feed.

 

[StuartsProjects]

my guess is the CV current is much less with 41.4V feed compared to 42V feed.

Here is the plot. The current drop under CV is about the same until 1Amp, but then diverges considerably.

 

[Woosh]

you should get longer discharge time with the diode charging method than 106 minutes last time. I expect 135 minutes - 8 minutes = 127 minutes.
8 minutes = 0.6V for your Si diode.

 

[StuartsProjects]

I will wait for the 1A charger to arrive before checking again, its good to be kind to batteries.

I did see a 42V 1A charger on Alixpress for around £3 delivered, but thats too cheap to be true I think.