Battery? Controller? ...no... Jimmy Wu's dodgy BMS.

[oddjones]

Hi - was riding to work the other morning and I noticed a lack of power for a couple of seconds... "Hmm, funny" I thought. Then it happened again... "Must have a look at that when I get to work" I thought - "maybe my dodgy soldering/crimping inside the motor has come undone".

Then the lights went out on my throttle and everything stopped working.

Thing is, I've tried connecting the phase wires together and turning the wheel backwards as per last time and there's resistance there - it doesn't appear that the motor wire has come loose at all - I will take the cover off the motor but I don't think that's my problem - The fact that there are no lights on at all kinda points in the direction of the controller - where do I start diagnosing that?

I've looked inside my battery/controller box and I can't see anything obvious...( no broken/loose connections)... bah.

On a brighter note - I have a brand spanking new set of Shimano Deore Disc brakes waiting for me at the post office and another 36v 10ah LiFePO4 chugging its way very slowly from China.

Maybe next time I should add a £0.50 fuse to the £40 controller...

 

[oddjones]

Err - OK - so it wasn't the motor. It appears to be the battery - as it's no longer giving out any voltage whatsoever (I've tested it both sides of the BMS and there's nothing - there doesn't appear to be any sign of damage to the BMS circuit board or components)... What can go wrong with a battery?

I wasn't pushing the bike at all hard when this happened either - indeed I was going downhill when it stopped working - might a cell somehow have come disconnected inside the Duct-Tape Pack?

(What have I done this time!!!)

It's another Chinese purchase and the thought of being without bike for 2 months while I send the battery back and have it replaced is really depressing - where do I start guys?

 

[marty]

Try charging battery? If that don't work? Unwrap the duct tape. Is there a fuse in there? Don't work on a metal table. Take off all jewelry. Careful with any metal tools. Sparks and Smoke are bad.

 

[oddjones]

Try charging battery? If that don't work? Unwrap the duct tape. Is there a fuse in there? Don't work on a metal table. Take off all jewelry. Careful with any metal tools. Sparks and Smoke are bad.

Battery fully charged - is there usually a fuse in these things? Want to see how likely it is before I start unwrapping - It'll be hard to send it back once I've unwrapped it!

(BTW I am a motivated website designer)

 

[Johnbear]

I have unwrapped my battery and there is no fuse inside. Yours may be different. It is very likely that a cell tab is detached from your description. If it is a bad cell maybe you can find it and your vendor will send you a replacement? Check with them and see what they have to say. I had a bad cell on a Yesa pack and he sent me a replacement right away.

 

[fechter]

Try measuring voltage on the small wires coming from each cell. It sounds like maybe a connection between cells broke. By measuring the small wires, you might be able to figure out which cell connection broke without unwrapping the duct tape.

 

[oddjones]

Try measuring voltage on the small wires coming from each cell. It sounds like maybe a connection between cells broke. By measuring the small wires, you might be able to figure out which cell connection broke without unwrapping the duct tape.

Umm - I don't seem to be getting a voltage across any of 'em?

Oh fer Chrissakes!!! >>note to self - "Make Sure your meter is working before you trust it"<<

There's nothing wrong with the battery at all - I had a broken wire on my meter.

However - that doesn't solve my actual problem of course... doesn't seem to be the motor, and I now know it's not my battery soooo -- by a process of elimination, the creeping finger of blame is edging towards my controller - what do I test?

The thing that's really bugging me about this whole episode is that I can't think of anything I've actually done to cause it - I wasn't crashing the bike around, I wasn't using the motor to pull my fat a** up hill...I got the two "semi-cut-outs" when I was riding along the flat - incidentally, everything seemed to keep going at this point - I just lost power for a second or so - which is what suggested a loose connection to me, then everything cut off about 5 minutes later as I was coasting down a hill (although I did have the throttle applied)

 

[dogman dan]

never mind, that was a dumb comment.

 

[John in CR]

Sounds like another wiring problem somewhere. At least you know it's not a phase wire again, and not your expensive battery. You can't rule out your controller though, because I don't like the sound of a "battery/controller box". The controller casing is typically an aluminum heat sink, so they need to breath. Check your power and on/off switch wires first, since if you have no lights you know it's not a motor wire, so that narrows the field considerably. Since you didn't kill the controller with that first torture test, you'd think it could handle normal riding unless it overheated in an enclosed box.

John

 

[fechter]

What controller are you using? Does it have a LED indicator?

 

[oddjones]

What controller are you using? Does it have a LED indicator?

Ecrazyman PIC based 36v controller - I didn't see a LED when I took the lid off.

 

[fechter]

It's got one burried in the middle of the board. If you take the end plate off and it's sort of dark, you should be able to see it light up when you turn on the controller. It's green. The number of blinks will tell you the status. If the LED does not light up, then it could be the orange wire is not getting power or something went bad in the regulator.

 

[oddjones]

It's got one burried in the middle of the board. If you take the end plate off and it's sort of dark, you should be able to see it light up when you turn on the controller. It's green. The number of blinks will tell you the status. If the LED does not light up, then it could be the orange wire is not getting power or something went bad in the regulator.

Well - it looks like the controller to me -this LED doesn't light up at all. I've checked the connections and switches leading into the controller and they're all good - I get 36v into the contoller. When I connect the controller up and measure across it I get 1v - bizarre.

However - not the end of the world as I was buying a new controller anyway - unless there's a magic button I can press

 

[fechter]

If you are sure the orange wire is getting power but the LED is dark, I would check the voltage regulator parts. There are a couple of big power resistors that would be prime suspects. The LM78xx regulator chips next to them would also be suspect. Could be something as simple as a bad solder joint. If you can hang a test hook on the resistors, you can see if they are passing voltage.

 

[oddjones]

If you are sure the orange wire is getting power but the LED is dark, I would check the voltage regulator parts. There are a couple of big power resistors that would be prime suspects. The LM78xx regulator chips next to them would also be suspect. Could be something as simple as a bad solder joint. If you can hang a test hook on the resistors, you can see if they are passing voltage.
View attachment 1

OK you crazy engineering God - you have me hooked - I've got a few days while my new controller's being delivered anyway so - where am I looking? - I remember resistors from A level physics - they're the blue guys right? -- when you say they're prime suspects - what am I looking for? resistance across 'em? :lol:

LM78xx regulator chips - they're the big guys with three legs - what am I checking with them? - Bring it on!

 

[oddjones]

Err - my controller looks completely different to that...

I have tested voltage across the single big resistor on the board - it's 0.4v. Resistance across it is reading 200 on a 2000 ohm scale on my meter.

Connected to the heatsink are 6 MOSFETS and the rightmost item is a single voltage regulator; an LM317T (which means absolutely nothing to me)

 

[fechter]

Hmmm... that is different.

Anyway, put your meter neg. probe on the battery negative and make measurements from there.

I can see the orange wire is where your battery voltage comes into the controller and goes to the big 200 ohm resistor. With the battery connected and controller turned on, you should have full battery voltage on the orange wire end of the resistor. You should have somewhat less than that on the other end of the big resistor, I'd guess somewhere around 24v. The 24v (or whatever) then goes into the LM317T, and comes out somewhere around 12-15v. There should be another regulator that drops this down to 5v for the PIC. Could be one of the small transistor looking things, most likely a 78L05.



How is your orange wire connected on the battery end. Does it go through a switch?

 

[oddjones]

Hmmm... that is different.

Anyway, put your meter neg. probe on the battery negative and make measurements from there.

I can see the orange wire is where your battery voltage comes into the controller and goes to the big 200 ohm resistor. With the battery connected and controller turned on, you should have full battery voltage on the orange wire end of the resistor. You should have somewhat less than that on the other end of the big resistor, I'd guess somewhere around 24v. The 24v (or whatever) then goes into the LM317T, and comes out somewhere around 12-15v. There should be another regulator that drops this down to 5v for the PIC. Could be one of the small transistor looking things, most likely a 78L05.



How is your orange wire connected on the battery end. Does it go through a switch?

Orange and red controller wires both fed from battery + through a switch. voltage either side of the switch (without controller connected - ie simply testing voltage across switch) = 36v

with controller wired in measuring across battery terminals I get 36v with controller off, 1v with controller on.

With controller connected and turned on, voltage to orange wire end of resistor = 1v. Resistance end-to-end over orange wire seems ok (0.6 on a 200 ohm scale) - I can't seem to get more than 1v anywhere in the circuit (even directly across battery terminals) as soon as I wire the controller in.

 

[dnmun]

are you using the DC volts scale? not AC volts. if you have a 10A range for your VOM leads, you may want to measure the current flowing into the controller through the orange wire, and if you have the black wire coming out of the controller attached to the negative terminal of the battery, you can measure the voltage from switch all the way through the voltage regulators down to the PIC supply trace. attach the black lead of the voltmeter to the negative side of the battery, and use the red probe to measure DC volts. BOL, dm

ps: if you are new to the VOM, you could start out by measuring the voltage on a AA battery alkaline=1.5V, or nimh AA, which should be 1.2 volts. don't use the resistance scale or diode checker while the circuit is electrically active. if you use the VOM as an ammeter, always start on the highest setting first, and use the 10A (fused) lead plug on the front until you know how much current is present.

 

[fechter]

.....
with controller wired in measuring across battery terminals I get 36v with controller off, 1v with controller on.
.....
With controller connected and turned on, voltage to orange wire end of resistor = 1v. Resistance end-to-end over orange wire seems ok (0.6 on a 200 ohm scale) - I can't seem to get more than 1v anywhere in the circuit (even directly across battery terminals) as soon as I wire the controller in.

That would indicate a problem with the battery or BMS. You should have 36v with the controller on.

You could try removing the controller and connecting a 120v incandescent light bulb to the battery and see if the voltage stays at 36v.

 

[oddjones]

You could try removing the controller and connecting a 120v incandescent light bulb to the battery and see if the voltage stays at 36v.

Ah.

OK - connected up a (240v UK) 60W bulb across it, switched on and voltage went down to zero.

Dunno where to start really with the BMS - tested the resistance across all the big resistors - 35 ohm each

tested voltage across each terminal of the multi-pole connector - seemed to be an even rise from 2.56v in steps of 2.5v up to 36v

I'm in the (possibly) fortunate position that the same guy is actually sending me my next battery soon (in the next week) - it would be useful if I could clearly identify the problem before then - if it's the BMS I could maybe convince him to send another with the battery - Is it possible to diagnose with just a meter?

 

[dnmun]

with these ebay battery packs, a lot of people have had problems with the BMS failing to allow the battery to charge. you can bypass the BMS, to charge, by connecting the black wire from the charger directly to the negative pole of the battery. if you have the BMS open you can put a jumper across the charging fet, between C- and the black wire going from the top of the fet to the negative of the battery. once the battery has a charge, the BMS may work and allow you to charge normally, but this is a big problem for a lot of people. who did you buy your pack from? jimmy wu? cammyc? mp3hk? all the same scammer. he has ripped off a lot of people so far.

 

[oddjones]

with these ebay battery packs, a lot of people have had problems with the BMS failing to allow the battery to charge. you can bypass the BMS, to charge, by connecting the black wire from the charger directly to the negative pole of the battery. if you have the BMS open you can put a jumper across the charging fet, between C- and the black wire going from the top of the fet to the negative of the battery. once the battery has a charge, the BMS may work and allow you to charge normally, but this is a big problem for a lot of people. who did you buy your pack from? jimmy wu? cammyc? mp3hk? all the same scammer. he has ripped off a lot of people so far.

Ah you're joking?

My first battery was from jimmy wu, my second from mp3hk - I actually had not had a problem until after ordering my second battery from him - although I noticed that emails in both cases came from "Lau Chan" - you're sh*tt*ng me that he's a scam artist? - God I'm so naive - I feel an endless sphere campaign coming on - if he's ripping us off, we need to let the world know - then we need to let him know that we let the world know.

Can you please confirm what you mean by "jumpering" - I assume you mean between the two locations I've highlighted here...

What I need to know is the following -

my charger tells me that the battery is charged - is this false?

When you say "you can put a jumper across the charging fet, between C- and the black wire going from the top of the fet to the negative of the battery. once the battery has a charge, the BMS may work and allow you to charge normally" - do you mean I charge up the battery completely with it wired up this way or I just charge it for a while, then switch to the regular setup?

 

[dnmun]

sacman actually had the same problem, but a lot of others came out of the wood work:

http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=6034

you should be able to charge the battery though. maybe after it has mostly charged you could reinstall the correct connections from the charger to the BMS and see if continues to charge. do worry that the BMS has discharged your pack, try determining what the final pack voltage is, when charged. it could have bad cells already if the BMS has drained the pack past the reversed cell potential. even though it should not. the fets melting the solder is too much imo.

 

[oddjones]

OK - so it seems I have fallen victim not once, but twice, to the curse of Jimmy Wu and his wonky, dodgy BMS.

Whilst my battery nominally tells my meter it has 34.5v - there's really nothing in there - it doesn't seem to be charging - the charger tells me the battery is charged - I get a green light and it stops charging after 3o seconds - but of I touch the two contacts together (+ and - from the battery) there's no spark -nothing.

How do I charge this bastard up? Am I left with a £200 white elephant? What worries me is that I have already paid for another of these batteries. Come on guys - somebody on here must be able to fix one of these batteries - surely the cells themselves are ok?

For all those of us in a similar situation who have spent hundreds of dollars on his batteries - do we not have the skills here to either fix or replace his BMS? (hint I wish I did)

dnmun - can you please explain in greater detail what you mean by putting a jumper across the BMS - is this in order to charge or discharge? How long should the jumper remain in place?

please?