Double your battery life

[banjineer]

This is a great idea and thread.  If / when the next batch happens please add me to the list for 2....

[BadNews]

Hi DickB,

I wont get my Radrunner 2 for a few more days, but, I also would like to be on your list for your next batch.

[murbot]

A switch like this with the option to disable it should already included as part of the battery. License it, get it done.

[smunsey]

Please add me for two as well.  Second RadMini due to arrive Tuesday!

Thanks,
SMunsey

[handlebar]

Probably just meter accuracy. My meter is specified as +_1.2% accuracy, so that's +- 0.5V.  Probably just a variation in meters.  I have two DVMs that do read within 0.1V on the Rad battery, but I have one that I got for free at Harbor Freight and it's different by almost 1 V.

Next time I order parts I'm going to get a Maxxim MAX6350 voltage reference IC and check it out against my DVMs.

In 2015, I paid $11 for this item:
https://amzn.to/3GQ6p9e

All I had to do was solder on leads for two 9 volt battery connectors. Since then, I have once sprayed the selector switch with contact cleaner. It was calibrated and came with a paper showing each of the four voltages to the nearest 0.01 millivolt.

It's advertised as an AD548. In fact, it's an AD584. The item I bought is out of stock, but searching for AD584 brought up quite an assortment.

I'm amazed at how accurate some cheap meters are these days.

[handlebar]

Although my smart switch works, I have a much simpler solution to partial charging with the Rad charger.

Standard silicone diodes have a forward voltage drop of about 0.8 – 1.0 volts at 1-5 amps. Simply placing two diodes in series in the circuit between the Rad charger and the Rad battery will drop the float voltage to provide less than a full charge. I've tested one cycle with two diodes and achieved an 80% charge level. Other than the reduced float voltage, the charger behaves just like before, charging at a constant 2.03 amps, then tapering off.  The Charge Indicator starts red and then turns green at charge completion as usual.

I've verified the charge level both by measuring rest voltage of the battery after charging and using the Rad Battery Charge Indicator.

The float voltage after the diodes was 52.7 volts, and the rest voltage after my dual diode charge was 52.0 volts.  This correlates to the following table from Benzo Energy, a Li-Ion battery manufacturer (I added the Rad battery column):

  Voltage
Rad    Cell    Charge
54.6   4.20   100.00%
52.8   4.06   90.00%
51.7   3.98   80.00%
51.0   3.92   70.00%
50.3   3.87   60.00%
49.7   3.82   50.00%
49.3   3.79   40.00%
49.0   3.77   30.00%
48.6   3.74   20.00%
47.8   3.68   10.00%
44.9   3.45   5.00%
39.0   3.00   0.00%

I made a similar chart last April, based on data from someone who had used a computer to record voltage each second as he discharged a Samsung 18650 at a constant current. I don't remember if it was a 35E. As I recall, each reading listed time, voltage, and culombs drained so far. There were thousands of entries. I took the culombs at the bottom and multiplied by .1, .2, .3... etc, to find the points on the list showing voltage at 90%, 80%, etc. Like you, I then multiplied by 13.

My chart varies from  yours.

54.7  100%
51.7    90%
51.0    80%
50.2    70%
49.3    60%
47.9    50%
46.3    40%
44.5    30%
41.7    20%
39.0    10%

At the time, I thought I'd validate this empirically, but that would entail routinely running the battery down until the controller kept shutting off the motor. Besides, my energy use per mile seems to vary from day to day, depending on things like hills on my route, pedal assist gear, throttle use, and speed. (I think a hub motor may use more energy per mile at lower speeds than at higher speeds.)

I've ordered a mechanical countdown timer switch, up to 12 hours in 15-minute increments. With the data I've logged in the past year, I should have a good idea how long to set the timer if I want to charge the battery from one voltage to another.

Oh yes... the display on my Bolton aftermarket controller reads 0.5 volt low, like yours.

[DickB]

I did run an empirical test on my Standard Rad battery and Rover 5. I ran full throttle on level roads, stopping every two miles to measure battery voltage after 30 seconds. I went 20 miles before the battery was basically exhausted, so for the most part every mile is 5% of capacity. I based my meter chart on these results.

I have 2300 miles on the battery after about 6 months. I plan to repeat the test at 1 year.

[handlebar]

I did run an empirical test on my Standard Rad battery and Rover 5. I ran full throttle on level roads, stopping every two miles to measure battery voltage after 30 seconds. I went 20 miles before the battery was basically exhausted, so for the most part every mile is 5% of capacity. I based my meter chart on these results.

I like your chart. Your knee at 44 volts(20%) agrees with other curves I've seen. For several months, I amassed empirical data by logging voltage and mileage readings each morning before riding. I've noticed that knee, so now I recharge at that point. From my data, I've noticed another knee: above 50 volts or so, I tend to get a lot more "miles per volt."

Internal resistance may be the problem. You were apparently able to maintain an efficient speed. Around here, I often have to slow to a walk. Perhaps below 50 volts (65% on your curve), battery resistance limits the current the controller can deliver. Acceleration would be slower and I'd spend more time climbing at inefficient speeds.

In particular, a 300-yard grade at 6% leads to my house. The whole block is 600 yards. Historically, the limit was 35, and that was quite safe. There are four small houses on that block, and apparently at least one household with political pull wanted it treated as a private road. About 1990, the town posted it at 16 mph and ordered the town cop to issue tickets. It was expensive for a town of 200  households to have its own cop, but how else could they enforce a law like that? There was a rapid turnover because it was as abusive to cops as to those ticketed.

I used to observe the hill with a stopwatch. In spite of all the tickets, about half the vehicles continued to drive at 35. That meant the cops were profiling. My mother was on the council. I told her the limit was a ridiculous burden to anyone who obeyed it. She ignored me until the day she got a warning. I think the cop was politely making her aware that he'd been told to enforce a ridiculous limit. When she brought it up at the next meeting, councilmen scoffed at her and fired the cop.

A mayor was elected who loved to do favors and inconvenience people. He got only 12% of the vote, but nobody else wanted the job. He had a speed bump put in at the bottom of the grade and another 200 yards from the bottom. Vehicles including his would come to a complete stop for each bump. He didn't mind the inconvenience as long as he was inconveniencing others.

After I installed a layback seat post, my Radrunner was the only vehicle in town that crossed those bumps without slowing. Hitting them at 25 wasn't uncomfortable because my riding position put most of my weight on the pedals. I changed my mind when I got home from the grocery store with cracked eggs. That wasn't so bad, but I was afraid I might arrive with broken spokes the next time. Now I slow to a virtual stop, like everyone else, and climbing  that hill is a drag. I should have gotten a mid-drive bike with gears.

[Webfor65]

Where can  one order a smart switch?

[Webfor65]

Where does one buy a SMART SWITCH?

[JimInPT]

Where does one buy a SMART SWITCH?

What exactly are you referring to as a "smart switch"?  I have a bunch of those both as wall switches and plug-in adapters for my Alexa network to control things, including running my Rad charger for either 7 or 12 hours with a voice command.

If you're referring to a variable-charge-percentage charger controller, just post your interest in this thread; "peanutbutterpope" is working on the first batch and considering doing more later on (see Reply #63), so he's keeping a list of interested people.  The first batch has not shipped yet, so you'll need a bit of patience.

[handlebar]

Where does one buy a SMART SWITCH?

I bought a 12-hour mechanical timer instead. (Simply to count down, the digital timers I saw looked like a hassle to use.)

As a rule of thumb, the battery charges at 2 volts per hour. Suppose I'm at 47.6 volts and I want to charge to 80%. According to DickB's chart, that's 51.6 volts, so I set the timer for 2 hours.

It's possible that not topping off could be harmful. I've been keeping track of voltages and watt-hour inputs for several months. Generally, partial recharging requires about 1.16 x average voltage x voltage change. (in the paragraph above, it would be 1.16 x 49.6 x 4.0, or 230 watt-hours.

However, charging all the way usually takes a lot extra, and that extra amount varies so much that charging from 42 volts could take more energy than charging from 40. I think that extra energy goes into equalizing, and the need may be different every time. I think letting the cells get way out of balance could ruin a battery pack much faster than charging all the way.

Rad Power recommends disconnecting the charger as soon as the green light comes on. I may not be awake to do that. I may use the timer to avoid leaving the charger on for hours extra. To start, my estimate would be an hour for every two volts I need, plus at least an hour for topping off. I'll probably change that according to observations.

An hour after seeing a green light, I have reconnected the charger and had it continue charging with two red lights. That could mean that after the battery settled down, the controller found that some cells still needed equalizing. If the timer shuts off after a reasonable charge time while I'm sleeping, the battery can settle down before I turn the charger back on in the morning to check. Allowing that rest may help the controller equalize cells.

[BILLA]

PB, Are you still producing the Battery charger controllers? If so put me on the list. Bill Asbell

[serrj]

Looks like this device can be used as battery charger controller: https://www.aliexpress.com/item/32886201775.html

[DickB]

The charger needs to reach float voltage and the current drop for battery balancing to take place. This device would shut charging down before balancing could take place. You would need to not use it periodically to keep cells in balance. The diode device allows cell balancing to take place at a lower float voltage.

It would be a convenient way to charge to a lower voltage and State Of Charge for long-term storage.