Modifying an old PC power supply into a lead acid battery charger.

     I like driving my homemade e-scooter around and so does my younger brother and now also my friends, but the battery is weak and requires charging often. Up until now I've been charging my battery through an old UPS that I found thrown out. The UPS had a dead 12V lead acid battery so it was already set up to charge and maintain lead acid batteries. Sounds great, but the problem is that it charges at only 0.6A and to charge something like a 52Ah battery would take over 80 hours, which is forever. So I decided to make myself a battery charger.

    I've tried making a battery charger out of an ATX power supply before but the problem is that they're not meant to go far over 12V and if I remember right mine could barely reach 13.5V and to bulk charge a lead acid battery you want 14.4V. The reason why they can't go much over that voltage is that they're based on a push-pull topology, which means that the output voltage is limited by the transformer primary to secondary ratio. Ideally for this you want what's called a flyback topology, meaning that the power supply output voltage covers a much larger range and is not strictly limited by the transformer ratio. The drawback of flyback SMPS is that they're limited to about 150W most of the time, hence why more powerful power supplies go for push-pull. But in this case 150W is more than enough to charge a lead acid battery.

    The power supply I chose for my mod is an old PC power supply, that looks like an ATX power supply but does not meet the ATX form factor standard. Otherwise it has the usual outputs with the usual plugs and also a quirky 4kBit EEPROM chip that connects to the outside world with an 8-pin molex connector. But more important than that is that the power suppply is based on a flyback topology. Also as an added bonus it has the switching IC on the primary side and there's only one opto-coupler for voltage feedback. Usually ATX power supplies have an SMPS IC on the secondary and along with that have a complicated feedback and fault detection network that can be a pain when modifying the power supply to do something else. 

    To convert this power supply I needed to do two things. First I needed to increase the voltage from 12V to 14.4V and second I needed to add a current limiting circuit to not overload the power supply. To achieve this I decided to not even bother modifying the original feedback circuitry and to just cut the traces connecting the opto-coupler to the feedback circuit to connect my own feedback circuit to it. 



    The feedback circuit is very simple and uses an LM358 dual op-amp. It senses the current through a 0.02 Ohm current shunt on the low side and the voltage through a resistor divider. The current shunt voltage is compared to a 60mV reference voltage derived from the 2.5V voltage reference created by the TL431 reference IC. The voltage is measured by dividing and comparing the supply voltage to the 2.5V reference voltage. Both outputs of the opamp are joined through diodes to keep both outputs from interfering, along with 1n feedback capacitors to keep the feedback circuit from overcompensating and oscillating. D1 and D2 can be replaced with LED's in which case they would work as CC and CV indicator LED's. R3 should be connected to the other side of the current shunt for accurate voltage regulation, but in my case it was more convenient to connect it this way and the voltage drop was insignificant. The current limiting in this crcuit is set up for around 3A, which I think is a reasonable charge current. I think the current limit is important to keep the PSU from overloading. My previous ATX PSU battery charger would overload and shut down if I plugged in a battery that was too far discharged.

    When I tried the PSU out everything worked out perfectly fine and the voltage ended up at 14.3V, which is perfectly acceptable. The current would limit at about 2.9A. So all in all I think this charger was infinitely more succesful than anything I've attempted before, probably because it's actually usable this time. Now the scooter battery charges much quicker than it did with the UPS and I think this charger will come in handy many more times again in the future.


Anyways, that's it for now and thanks for visiting my blog.

Komentarai

Rašyti komentarą

Populiarūs šio tinklaraščio įrašai

Playing with junk: the SN29764 (LM1017) 4-bit binary 7-segment display driver.

Vaizdas
    While digging through my parts bins looking for my spare arduino nano I came across a boart that I was keeping solely for the 7-segment display that was attached to it. I noticed that there also was an IC marked SN29764 on the board and got curious. After some googling I came up with a LM1017 datasheet, which states that it's an equivalent to the SN29764.      Basically what this IC does is it takes a 4-bit binary number through pins 2, 3, 4 and 5, and displays that number on a 7-segment display. Since this IC was made to display channels on a TV it doesn't display the number 0 but it displays numbers from 1 to 16. Pulling pin 1 down sets the display to full brightness.     I find it quite interesting that there would be a dedicated IC for this application but the world proves to be quite large and technology has sure advanced since the old days. There isn't a date code I can read on the IC but it might very well be older than me even, well I'm not even that old so
Skaityti daugiau

Modding a laptop power brick into a lithium pack charger and some Manhattan style prototyping.

Vaizdas
      I've built a few 4S lithium battery packs but I didn't have a charger. So I decided to make one by modifying a switch-mode power supply. The power supply I used had a special plug on it with three pins. One of them being a sense pin to change the voltage. On this plug an adapter would connect, which would have a resistor in it that would change the voltage on the output. This power supply is nothing special, though. All this wire does is go straight to the REF pin on the TL431 reference IC inside. How do I know that? Because I accidentally blew the TL431 and had to pull the power supply apart to replace it. If you're willing to pull apart a laptop power supply you can add this wire to the TL431 reference and you'll have basically the same thing I have.     Opening them can be a pain. Mike from mikeselectricstuff shared in one of his youtube videos that whacking the weld/glue seam on the power supply will crack the joint and the power supply will come apart. But I&
Skaityti daugiau

Infrared RGB LED controller using Digispark (Attiny85).

Vaizdas
    While building a PC I thought I'd toss in some RGB to see how it looks. Problem is I had lost the receiver. Bummer! But I still had the remote, so I decided to make my own RGB controller. For this project I chose to use the Digispark, which is like an arduino but is based on the Attiny85. A regular arduino like the arduino nano, or even any other microcontroller like the STM32 "blue pill" can be used for this project. In this case I wanted a small microcontroller, something like an atmega328p would have been a bit of an overkill.     Before I could make my microcontroller do anything I had to find out what my remote was transmitting. To do this I used my OWON VDS1022 USB oscilloscope, with which I saved the samples of the waveform in a .txt file. The bitrate of the signal transmitted by the remote is quite low so something like a DIY audiojack oscilloscope (look it up) would work also. The .txt file then was passed through a matlab script I quickly put togeth
Skaityti daugiau