Electronics > Beginners

19V PSU charging 12V lead acid battery

<< < (6/7) > >>

nuclearcat:

--- Quote from: ThickPhilM on September 17, 2019, 05:26:43 pm ---
--- Quote from: nuclearcat on September 16, 2019, 06:26:30 pm ---
--- Quote from: ThickPhilM on September 16, 2019, 06:13:02 pm ---Perhaps a small automotive lamp in series with the PSU output?

More information would be helpful, like the output current of the PSU, the type of lead acid battery (wet, gel etc) and its capacity

--- End quote ---
Lamp is worse than adjustable linear regulator :)

Let's assume PSU current is sufficient in any case, i dropped idea relying on PSU current limit.
Most likely not smaller than 12V 2.6Ah gel, this means with max 0.1C it should be 0.26A. As i charge by PWM, probably in pulse i can use 0.2C (if i wont exceed 50% duty cycle). I plan to implement it by using typical 2 NPN transistor circuit.
But thats the question, if battery chemistry is ok with fact, that on 10-20khz PWM pulse current is 0.2C, but average current is 0.1C.

--- End quote ---

Well, I was making a couple of guesses based on your previous posts, ie that you were looking for lowest complexity due to 'third world issues', and that you were using small car or motorbike wet lead acid batteries.
Now that I know you're using very small gel types, that changes things significantly.

Gel aka AGM type lead acid batteries really don't like over-voltage situations, especially very small ones, and doubly especially the likely low quality ones you'll be using, which will have a lot of very impure recycled lead in the plates. Do not use a pwm regulated 19v supply, you will kill the battery in a year or so, depending on a couple of other things (ambient temperature, depth of cycling).
Cyclic charging of these batteries requires a constant current for the bulk charge, and a constant voltage for equalisation. PWM charging from a 19v source is not a good idea, a simple linear regulator and current limiting resistor are the best bet for such small batteries. A laptop charger is massive overkill.
C/2 is more than adequate for the bulk charge in this case, exceed C and you will cook it. Float current shouldn't be more than a few tens of milliamps at most, if it draws more, it's on its way out.

Do not believe the manufacturers datasheets on charge voltages; they want to sell you more batteries.

If you want a good lifespan, a float charge voltage 13.6v really is the highest you should go, cyclic use, 13.8v, but don't let it sit at that voltage for more than a couple of hours.

The quality of these batteries varies considerably. Cheap Chinese ones are not very different from cheap Chinese NiMH in their capacity claims in my experience; adjust your bulk charge rates accordingly.

With reasonable ambient temperatures, and sympathetic charge/discharge profiles, I've had good quality batteries last 20 years.

I've also seen good quality batteries cooked in months, by crappy quality/badly adjusted chargers. Some of the bigger ones can be quite exciting when they fail  :-DD


--- End quote ---
Thank you very much for the detailed advice!!!
Most likely I will go this way, bulk charge stage C/2 (max) or even less current, depending on the remaining battery capacity and temperature, and static CV, 13.6V for equalizing charge.
Probably i will not do it simplest way, as little bit more components will increase a lot efficiency and reliability of solution.
I will use ADC from ESP8266 to measure battery voltage, one pin to enable topping charge, and
1) I2S/PWM for current offset over RC, to LM317.
or
2) I2S/PWM driving crude step-down buck converter (which is more preferable, as more efficient, but probably more tricky to implement)

I am considering also to shift measurement range from 0-14V to 12-14V by using inverting opamp with virtual ground at 12V, but probably better to open separate topic for that.

iMo:
I would go with RPi 4 (4GB ram), LTZ1000 as the Vref, double ovenized OCXO for precise PWM, LTC2500-32 ADC, 0.001% 0.1ppm/C shunt. Your lead acid battery will be pretty happy then :)

floobydust:
But the lead acid battery tempco is -3.3mV to -6.7mV/°C/cell ... it goes in the oven too? ;)

I think a car battery charger with Wi-Fi is useful, you can remotely check up on it instead of walking back and forth to the garage. Lowest cost would be to mod the 19V laptop adapter to put out less, say 14V or whatever an MCU commands with PWM, but that is beyond a beginner's project. A PWM controller buck-converter or linear regulator might be something basic to add after the 19V PSU.

ESP8266 has a crappy A/D, not good enough here. Noisy, non-linear 10-bit 0-1V input range, poor accuracy you need to calibrate it and have to shut off the RF TX circuit to get a reasonable reading. I thought the internal 1V reference does poor over temperature too but can't find that correction graph.
https://www.espressif.com/en/support/download/documents

iMo:
Here is the simplest CC CV charger I saw somewhere in past.
And it works, I've been using it from time to time (nimh, lion).
Use it at your own risk :)

PS:
I_charging = 1.25V / R1      [A, ohm]        max 1A, LM317 likes a small heatsink, R1 2-3W
Vbat = (R2+R3)/R3 * 2.5V                       TL431 trips at aprox 2.500V

nuclearcat:

--- Quote from: floobydust on September 17, 2019, 08:22:41 pm ---I think a car battery charger with Wi-Fi is useful, you can remotely check up on it instead of walking back and forth to the garage. Lowest cost would be to mod the 19V laptop adapter to put out less, say 14V or whatever an MCU commands with PWM, but that is beyond a beginner's project. A PWM controller buck-converter or linear regulator might be something basic to add after the 19V PSU.

--- End quote ---
Purpose of device is a bit different. I am living in country where government supplied electricity lasts max 12-18hours per day, when it go off, in some households private generator supply kicks in (still, it takes 30-60 seconds, enough to reboot router or DSL modem), or if it is less fortunate people - they stay in dark. Sometimes electricity supply is very unstable and might go on/off tens of times per day. Regular offline UPS doesn't last long, too much conversion losses, and need some hacking (hacking away buzzer, hehe).
This device will be connected in the middle, between AC/DC SMPS and wifi router, to provide continuous supply, so at least people can stay online from their mobiles. Most of routers is OK with 14V (step-down buck inside, only problem sometimes caps max voltage on DC IN), but on some old routers i seen LDO, there i i can use cheap LM2596 step-down brick.
ESP8266 will keep an eye on battery voltage, and if it is too low, send data to cloud/owner mobile, cut off router, lower power usage as much as possible (disable RF) until mains appears, and then if still no supply, next level go to deep sleep and wake each 5 minute, to check if power appeared. It will be easy to implement additional functions, battery health monitoring, remaining runtime, programming some router cut-off for night time and etc.
I am considering to make it OSHW, if it will work fine.

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod