48v lifepo4 charger = 55v need repair help

[dnmun]

why don't you just list the cell voltages from 1 at the bottom to 16 at the top?

do you have a single cell charger? with that you could charge up the cells that are dead, but i suspect you were riding on them while they were below 2.1V so they are toast.

if you have some money you can just email travis(is it tgintz @evcomponents.com ?) and buy some replacement cells for the dead ones and then put it back together and try to charge it wih your charger then. you need to look at each of the sense wires and make sure they are connected to the cells and did not break in the accident like luke said. if they did you need to fix that too, but first list the cell voltages from low to high.

i couldn't find travis new email address at evcomponents.com but you could log on to their site and order or send a pm to 'frodus' on the sphere.

 

[webfootguy]

If the cells are that low, they will not likely hold a full 10ah anymore and will limit the usefulness of your pack. If you have the ability to measure their discharge capability (CBA II or such), I would do so. Otherwise, I would just replace them. Best prices around are: http://www.evcomponents.com/ProductDetails.asp?ProductCode=38120S

 

[dnmun]

wow, i had never been to their site. that is cheap, that is cheap...!

same price as what we paid in the group buy and they don't have to wait 4 months.

it's not fair. we were looking at paying $24 each back when we thought we would airfreight them from headway. wow.

 

[Capripower]

So the question is why did the voltage drop on only those 2 cells ? because they where never balanced all the way ?
Is it therefore advisable to check all the cells of a new battery to make sure that all of them are at the same level?
who many times do I have to charge a new pack before all the cells are balanced ? 2 cells of my second pack are at 1.5v with
the rest being at 3.1 after 7h of charging so I will have to buy a few new cells and thanks for the info .
Thanks dnmun
Felix

 

[dnmun]

sorry, unless we know where the low cells are located and what all the others look like around it, there is little anybody can do for you since we can't even tell you where to look. i know this is the 4th or 5th thread where we tried to help you, luke told you what to look for even. if they are dead there is nothing to do but buy some new ones. but if your sense wire is broken the new ones will end up dead too, except it won't discharge anyway if the sense wire is broken. BOL.

 

[Subfightr]

Sorry it took so long to get the pictures to ya'll. here they are.

If it takes up too much room lemme know and I'll remove the post. Keep in mind, I am a noob to all this. But willing to do whatever anyone says. (regarding the charger )

Under side of charger case

Top of charger Case

Front of case

here is some tiny fuse looking thing

 

[Subfightr]

I dont know why it wont accept my images... any idea? I think I put the command correct.

Here is a link to a gallery of pictures. Please let me know if i can give any more information. http://img268.imageshack.us/gal.php?g=dscf0010bk.jpg

 

[dnmun]

do you have the images on your computer? are they big files? you can postup VGA size files. i think the rule is about 100k, but i think my camera is 150k or so. an ancient almost free, $35, sony 3.2meg, but i did get canon 3.2Meg for free, and even got the CCD replaced on recall so it is like new, but i prefer the sony.

anyway you can go delete those x's and try 'upload attachment' directly from your disk. then 'add the file'. but first go back and get two full sized VGA images in focus from the front and back, then put them right on top of each other like leslie did, but you can see how the traces are built up on the output side, but i could not follow the layout to the oscillator, try just two full sized VGA images to replace those.

this is the charger that is bad? doesn't light up or the voltage did not go high enuff? do we know the pack is balanced?

i wanna take credit for fixing mike's headway charger tonight too, but it actually just started working on its own, i don't have clue. the other one i thought had the best chance is totally dead with no power led at all.

this is more of murphie's inverse law, but i bet when i send it back, it will stop working too, so i will wait and test it again when i finish my battery.

 

[Subfightr]

I uploaded all the files to imageshack.com and had it cut the file size in half, and change the resolution so it would be suited for msg boards. Were you able to follow the second link? I just copied each link to the pictures and pasted them inside the img command. While my girlfriends camera is also 3.2 mega pixels, the camera would not focus for crap on the object up close, what you see in the second link is as clear as I could get it. I'll play around and edit the above posts links till somethin works just so I know for future reference at the very least. I figured it out, the file size for most is like 106 kb, over by just a few kb.. I can redo it if you like to fit the 100kb limit or you can just follow the below link http://img268.imageshack.us/gal.php?g=dscf0010bk.jpg the pictures within the link can be clicked and made larger

Yes, this is the 48v charger that outputs only 54 volts not 59-60 as it should. I notice it states 220v... does this mean it is to go into a 220v wall jack/plug? As apposed to the 110's that it is currently plugged into?

This charger does light up, just not correctly, once plugged in, the chargers green light is to turn orange to indicate it is charging and the red light is to go off. But no, both red and green lights remain on endlessly and never turn orange, the fan does come on and never shuts off.

Is my battery pack balanced? I have no idea, how does one find out? I have owned this pack for about a month and the entire time it has had to be charged with a 48v charger designed for SLA batteries. I would assume it is not balanced but yeah, I donno what Im talkin about...

 

[amberwolf]

Yes, this is the 48v charger that outputs only 54 volts not 59-60 as it should. I notice it states 220v... does this mean it is to go into a 220v wall jack/plug? As apposed to the 110's that it is currently plugged into?

That could indeed be the problem.
I can't read most of the (Chinese?) label here:
http://img62.imageshack.us/i/dscf0026.jpg/

but the numbers are in English, and it's likely that the 220V is not a "max" input voltage, but a "nominal" one. Do you have a 220V single phase outlet anywhere in your house?

Also, it is labelled as 50hz, rather than the 60hz used in the USA (if that's where you are). Don't know what problems that might create.

As for the pack being balanced, you'd probably have to measure the individual cells inside the pack, unless the BMS has an indicator per cell on it somehow.

 

[317537]

Here is the top view of the lm358 op amp,

View attachment lm385.JPG

As you can see the this works as a flip flop or frequency generator. The when reference voltage reaches, eg full charge and we will speak of it this way for the lamen on the inverting input, the output sways from either a or b depending which voltage is higher on the inputs.

In my case I found out how this works the hard way, I melted the LM ic and my charger stays in full charge mode by default and when the reference voltage is under 51v via a resitor and or diode it stays in float. This LM358 ic is hooked up to both leds and the transitor switch for the fan and I suspect its tied to that fet aswell.

Ive had a dry joint on this for some time and when moved i could here some arcing. I found the dry joint and soldered it down and proceeded to tin a few more joins to make it more sound. I did this without my glasses and was doing this all blind. I cross soldered a high voltage to the input of this lm358 and it blew in record time.

I'm confident that a replacement 358 chip will get this charger on its way again. It still charges in full charge mode but it doesnt switch back to float without the lm358 chip. all the resistors are in good working order to the input, I will post some more stuff and pics of the broken chip.

The reason chargers "working" stay in float mode when not connected is that the only reference voltage is the the one the highest voltage. It needs to be brought down by a pack to switch to full charge mode.

 

[Subfightr]

Thank you for the response AMBERWOLF, I wrote the seller and asked him/her to translate it for me. I donno if I have a 220v, if I do... I guess I'll try it, no idea what will happen once it receives twice the mount of volts... I hope my home owners insurance will cover the fire damage

Sub

 

[317537]

That 450v cap look like its not been converted to US standard but the hertz in this circuit is not important. As it goes through the rectifier diodes and comes out DC. I would expect the right input voltage and correct resistors in order to make this thing give up the right current.

I'm not sure about how this would effect the charge modes. But your bike still could have battery issues. The charger will not blow up or cause a fire. Some switch modes work on either voltages.

 

[317537]

Hows this for sharp focus.

This is a typical burnt out IC.

View attachment lm358.jpg

In this case pin 8 the supply of my lm358 chip has been overvolted. Similar to a hall sensor supply these things need to be powered by a supply. These can take up to 32v so I must of givin it at least 60v, @ 220v these things explode and blow out a chunk of shrapnel.

Watch your eyes when messing with electronics you will blind yourself with the easies mistake. Caps can be the worst when overvolted.

http://www.youtube.com/watch?v=[youtube]v5v4_AlDQTo[/youtube]

Whoahhhh. Imaging a big sucker rated at 16v going off on a 220v mains.

 

[dnmun]

220 on the case just means it can be used up to 250V, and the 350-400V caps are correct since the spikes are in the 350V range for 120 V operation and they add another cap in series when it is running on 220V. you should see an empty spot for a cap then on the input side on the 120V chargers.

the lm358 does not get involved in the switching at all. the input section on the switcher where your power resistor and the oscillator chip all drive the power mosfet are separated by the inductor from the output stage where the lm358 lives. the only connection is the feedback through the phototransistor which provides the feedback to the oscillator chip at pin #2 on mine, about 1.84V appears on pin #2.

the lm358 is driven from a 5V source in the output section that comes outa the inductor separate from the floating ground and the main charge current path, which you can see if you go through and follow where the traces are coming out of the inductor on the output or 'high' side. the output section on mine has 3 legs going into the floating ground, 2 legs to the output current path, where the traces are built up for extra current capacity, and then one more leg supplies the 5V regulated voltage for the circuits, and it goes through a diode from the inductor leg to the 5V rail. except it's not 5V i think.

the 358 output drives the charging led and the power FET across the output that connects the charger to the battery when the battery is plugged in, and has an input from the voltage delta created by the charge current flowing through all those shunt resistors, 4 in parallel with a big 1 ohm resistor on top, that is hard to follow on the boards though, i have to go through and mark the floating ground and the output + sides, and the 5V on the output, with a sharpie, there are jumpers to carry the 5V and the ground all over the place where the 358 is, but your board looks much simpler than the headway charger i have, but this is a two stage charger, CCCV so it has another 8 pin SOP IC, a fairchild knockoff? F9444, which controls how the charger works when it reaches final voltage.

you should go to my other charger thread where i have the link to the hififever website where they have the datasheet for the OB2269CP. that is so educational about how the switching mode feedback works, then it will be obvious when you are looking at the board where the voltage regulator is on your board, with the little cap for stability, between the output + and the ground in a resistor bridge, tied back to the phototransistor, so then you will kinda see the layout of the parts on the board and it is easier to follow that way.

still not sure how to figure out what is wrong on these. one of the chargers i got from mike started working last night, i have no idea why, but i will take credit for it anyway. the other one is now totally dead, and it was the one i thought had the best chance of working, go figure.

 

[317537]

Yes that 358 chip traces back through the el817 phototransistor to pins on the uc3842 chip which does the pwm. I think its a case or the 358 controlling the leds fan and Pwm.

Edit. Its not the main switch like you said but it the logic behind it.

 

[dnmun]

yes exactly. the output from the opamp in the lm358 turns on the led for green, and it may shut off the output FET too, and turns on the fan and the 'bias' to the phototransitor led that makes the phototransistor go fast and that turns on the oscillator more(longer pwm) through the feedback pin #2. you can see the voltage go up and down on the feedback pin as the output current from the charger goes up and down.

make sense?

 

[317537]

Ok I bought 3 LM358n op amps today and a saddle, installed two 50v 220uf with series with divider resistors bridging both terminals to equal one 110uf 100v i vented last night, the divider resistors are to keep the caps balanced. Two caps will give much better heat displacment. Found the dry joint by tapping the board with a plastic screwdriver butt while plugged in when a low light environment, I and saw the arc.

It works


I figured I might have some grief so I installed a saddle for the 358 chip so I could replace them.




In float mode.


View attachment full mode.jpg
Full charge mode.



Leave enough alone now. I've had my fun and have this charger doing 3.1 amps soldered in solid state. I just had to have some patients and await the morning till the shops opened.

Oiled the fan and she's good to go.

 

[Subfightr]

Very impressive Les/#'s Good for you man, that is awesome.

 

[317537]

Very impressive Les/#'s Good for you man, that is awesome.

ES is awesome an place to put your thoughts into action and to share learning about electronics. 6mths ago I would of been lost. I have been doing low level electronics for years and years. I figured its time to get my hands dirty.

Dnmun has been a great source of information to me in this pushing me to learn a little more.

 

[dnmun]

hahahahaha

gotta quit while your ahead. i mentioned your power resistor hack in my headway charger thread.

my power resistor is 2 ohm so i could only find a 12 ohm that was 1W in my junk. i added the 12 to the 2 ohm in parallel, so it is a 17% increase in output current. i even added solder to the traces to carry the current ,

but it went from 3 to 3.5A, and that's worth something.

now i wanna figure out how to adjust the final voltage. it's in those resistors. between the output + and the ground, and tied to the input to the phototransistor so i think that is where to change them. figure where to tack on an extra resistor in parallel.

i found three different types of oscillators on ebay, that is so funny, buying spare opamps for when they blow up. hahahaha

 

[317537]

The charger shut off at 59v on the dot, however Im having a time balancing these gel agm batts so ive added a 5amp rectifier diode to bring it down to 58.3v to give these new SLA batteries plates a chance to cycle in and gain some surface area. They float at 13.9v each and with the diode rest at 13.72 which is in spec. They should balance in float and charge to max capacity overnight.

 

[dnmun]

when my headway pack is pulsing the power lead is also almost going out as it pulses. so the front end is shut down, has to be by the pwm controller.

the power to the led comes off the inductor and provides the feedback VSS reference inside the pwm controller. and VDD, and that could mean the pulsing is starting in the front end, or the front end is being shut down by some kinda feedback from my ICs. these chargers have the CCCV stages programmed in and i think it is not shifting to CV .

i have also noticed that the chinese always use blue for the charger positive terminal and ping uses blue for his charger lead too.

 

[317537]

hahahahaha


i found three different types of oscillators on ebay, that is so funny, buying spare opamps for when they blow up. hahahaha

Well Ive got some lm317ts and a lm338 I could experiment with. That op amp seems handy you know.

I think on my charger it gives out around 63v thats why i needed a 100v cap in there and there is a zener tied to the op amp shuts it off at 59v, you may see two series diodes facing opposite polarity near the edge of my board, i think its handy to have the extra overvolt potential so it shuts off and not stays in full charge mode if you get my drift. Undervolt or overvolting the circuit will probably give you no results other than lessen or increase efficiency of the current draw you have got going now, and if you go too hard you will increase the resistance. Try looking at the logic feedback. I need to find your thread.

Can you pop me a link to your thread?

My charger seems easy peazy and the headway might be fun to learn some.

 

[dnmun]

here is the link, the charger #1 is now the steady light, and #2 is the one pulsing.

also this is where the link is for the oscillator in my headways and they have the best schematic for understanding the pwm feedback circuit and how the phototransistor controls the feedback in the oscillator.

http://endless-sphere.com/forums/viewtopic.php?f=14&t=13060