Need some help before I drive my ebike into the river.....

[Lenk42602]

Fellas, I am stumped. Need some help trouble shooting. Getting increasingly dis-interested in pedaling this bike around to the point of even thinking of selling..... :?


Completed the bike buld in 11/11 with the following set up:

24s A123/20ah pack/ 18X4115 Lyen controller / pre-disc 5303 with stock phase and hall wires.

After getting the bike set up, I had the controller set to about 40 amps max, with the LVC set accordingly for my lifepo4 pack.

From late november in 2011 until late january in 2012, I ran the bike with no problems back and forth to work, running the bike no longer than about 25 minutes at a time. During the this commute, I would do a little over 8 miles each way, and I am getting about 1.5 miles per amp hour.

In late january, I rode the bike on the first "long" ride - continuous fun ride. At around the 14-15 mile mark, I lost all power to the hub motor.
No power to the CA through the controller.
However, when turning the power on at the controller, current was reading properly at the phase wires.
After getting the bike home, I pulled the motor, checked the hall sensors for proper function. They were fine.
I then checked the wiring (hall/phase) for insulation issues. The wiring was fine.
Re assembled the motor, wired up
After some controller diagnosis, I determined tha a few fets were fried in the 18X4115.
Repaired the controller, reconnected everything, and immediately bike fired up and worked no load.
Took the bike for its first shake down run and it only made it about 1.25 miles on a WOT pull before shutting down.

Sent it back to LYEN for a 4110 controller.

Received the new controller, wired it up, everything has been working properly,except that I have yet to make it all the way home with out the battery pack cutting out.

The first time on the way home, I made it into my drive way and the pack cut out. I figured I made a mistake programming my LVC and made an adjustment.

The second to last attempt at commuting, on the way home, at about the 14.5 mile mark, I lose all power. the only way I regain power is by plugging my chargerback in to the battery charge port, thus powering up the BMS. had to pedal about 1.25 miles back to the house.

This happened two days ago on my "evening commute" vid. When I got home, I pulled the side panel, charged the pack to full voltage, pulled the balance leads off the bms for the subpacks, and measured full charge voltage in my 24s pack. All cells at ful charge were at 3.66 volts

Tonight, on the way home @ 14 ish miles, I again lost power and could not get the pack to power up the controller again until I got home, put the pack n the charger for a second, and my CA lit up reading 79 volts.

I again pulled the side panel off my bike, checked the balance leads on each cell. ALL cells were reading 3.28 volts.

I am losing all power at 14 miles, and about 8.6 ah used. This pack should be yeilding a full, usable 17-18 AH, and neither the BMS or the Controller should be cutting power due to LVC. Given that kind of capacity and my 1.5 miles/ah riding style, I should see close to 30 miles range before the LVC even gets close to acting on the controller or BMS. The BMS is set for 2.0 volts, the controller set for 2.2 volts per cell.

What else can it be? The only other thing I can thin is that either I missed or screwed up some insulation on the re-wire of the hub motor and it is causing a short that is somehow tripping the BMS, or the BMS is faulty.

Controller is good, hall sensors are good, all cell voltages are balanced in the pack. It seems like the same issue is still prevalent that existed from day one.

It is only after I plug the charger back into the pack that the controller powers up again. If I do not do this, then unplug/plug the power leads from the battery to the controller, there is no arcing or spark as you would normally see from an 80 volt pack. This indicates to me that the BMS is "off" not allowing current to pass through the power leads.

any help is appreciated. I need some fresh perspective....


Len

 

[auraslip]

I have a solution: throw the bms in the river.

edit: but seriously. I run everyday with out bms. No issues after a year and 5000 miles.

If you can't ditch the bms, you should at least make it so you can bypass it easily so you won't get stranded. IMHO, anyone with a bms should be able to do this!

Out of curiosity, what bms are you using?

 

[Lenk42602]

dunno what kind it is - Cell man put this sucker together for me. Its green and has black and red wires coming out of it with some little plastc doo hickys connected into a thing a ma bob deal type thing... :lol:

Seriously though, I am not sure if the BMS is cutting the pack power prematurely due to being faulty, or if a short could be causing the BMS to do what it does? Dunno...

Len

 

[auraslip]

Are you sure that you don't have a bad cell group? That'd be step one. Check to see if the bms isn't just doing it's job!

 

[Lenk42602]

already checked the cells after before and after charging with the BMS attached. They are all balanced immediately after BMS cuts out, and all cells are balanced immediately after full charge.

 

[Ypedal]

First , i'd run a capacity test on the pack, not using the bike... a bench test of my LBD setup with a calibrated CA or equivalent... rule out the bike side entirely.

Then if the pack can deliver 18 or so Ah on the bench, then start checking the bike, is the CA calibration correct ?..

you get 14 miles range at what speed before it cuts out ?

what is the pack voltage under load when the pack cuts out ? ( not at rest )

 

[Lenk42602]

First , i'd run a capacity test on the pack, not using the bike... a bench test of my LBD setup with a calibrated CA or equivalent... rule out the bike side entirely.

Then if the pack can deliver 18 or so Ah on the bench, then start checking the bike, is the CA calibration correct ?..

you get 14 miles range at what speed before it cuts out ?

what is the pack voltage under load when the pack cuts out ? ( not at rest )

CA is calibrated correctly - I think. All i did to it was set the rshunt value. Is there anything else I needed to do? Obviously I set the speedo stuff up...

around 14 miles the pack cuts out. this equates to less than half of the pack's capacity...

Pack voltage under load sagged to 74 volts right when it cut out. certainly not low, right?

you felling any better?

Len

 

[Philistine]

Just as an experiment have you tried using the CA to limit the maxamps to something well below your batteries capabality, like 20amps or such? The reason I say that, is that whilst I know nothing about lifepo4 batteries or BMSs, I was just always stunned at the degree to which my 5303 just pulled max amps, all the time (as in all the time). With all my other motors I would see changes in the amp draw throughout the power curve of the motor, but the 5303 just whores the amps the whole time. So maybe it is that the constant max amp draw on your BMS/battery is causing problems. Just a thought. If you wanted to test that theory, if you have some Lipo handy, try running the bike on an equivalent pack of lipo and see if you have the same problem, then you know it is the BMS/battery. Alternatively, as I said, drop the max amps down to 20amps or something and see if you get the same issues.

Personally I will never run a 5303 on anything less than a 24 Fet controller and Lipo (when going 40amps +), it is just such an amp whore.

 

[Lenk42602]

Thanks Phil,

I will attempt that . Currently my controller is set to max amps of 43 volts.

The most amperage ever drawn from the pack was 50.2 ampslast fall.

This A123 pack is able to handle 80 amps continous, 300a bursts...

I am kinda out of extra batteries, so we will give the bike a go by limiting the drawt to 20 amps....

len

 

[amberwolf]

When you measure the cells, are you doing it at the cells, or at the BMS board itself? It's possible there is a loose connection or bad wire that causes just enough resistance to cause the BMS to read lower than the cell actually is for one or more cells, and that's triggering the cutout.

If you have any way of monitoring per-cell voltages while riding, connected at the BMS PCB itself, I'd try that just to see.

 

[liveforphysics]

I suspect a false-positive hall combo.

Some people ride for years on a false-positive, putting huge stress on controllers and batteries, getting terrible Wh/mile, getting random part failures etc.


Measure your no-load full-speed current. Wheel off the ground, pin it, measure current drawn off the battery to do it. If it's over an amp or two, then you know you're running on a false positive.

 

[John in CR]

You're running a 4110 controller, probably with an R12 mod, so the LVC could be inaccurate and shutting the controller down at the wrong voltage. Set it a lot lower. You have a CA and a BMS anyway. You can test this by setting the LVC 1V below current pack voltage, and if it doesn't work, then that's the problem. Then if you really want LVC protection from controller, you have to find the relationship of Actual Voltage to Programmed Voltage.

I hope it's as simple as that for you. I got nibbled on by that exact issue, but run or not run at all just last week, so it reared its head at home, not on the road. I can imagine your frustration.

John

 

[itchynackers]

I agree with John. Still sounds like a LVC issue. Try this: run the bike until it shuts down. Then reprogram the LVC on the CA to VERY low (as an experiment, so it doesn't affect anything). Then reflash your controller with the LVC VERY low. See if the bike now runs again. If so, its just your LVC setting in the controller. Like John said, with those modified controllers, you can't take the LVC programmed value literally. For example, on my controller if I want an acual LVC of 86.4v, I have to set the software at (86.4 +2)/2 = 44.2V. Hope this helps.

 

[Lenk42602]

Fellas, Thanks for your input.

When you measure the cells, are you doing it at the cells, or at the BMS board itself? It's possible there is a loose connection or bad wire that causes just enough resistance to cause the BMS to read lower than the cell actually is for one or more cells, and that's triggering the cutout.
If you have any way of monitoring per-cell voltages while riding, connected at the BMS PCB itself, I'd try that just to see.

AW, I am measuring the voltages at the cells, off the balance/monitioring plugs that connect to the BMS.

I suspect a false-positive hall combo.

Some people ride for years on a false-positive, putting huge stress on controllers and batteries, getting terrible Wh/mile, getting random part failures etc.
Measure your no-load full-speed current. Wheel off the ground, pin it, measure current drawn off the battery to do it. If it's over an amp or two, then you know you're running on a false positive.

Thats some low hanging fruit. I will check that one first - I am thinking that I am running at 5 amps WOT, no load though. no good???... Also, Luke, I am seeing about 1.5 miles per AH, which is about where I should be given my set up???

BTW, My phase wires are wired all colors matching,

and hall : Mtr CTR
Y B
G G
B Y

You're running a 4110 controller, probably with an R12 mod, so the LVC could be inaccurate and shutting the controller down at the wrong voltage. Set it a lot lower. You have a CA and a BMS anyway. You can test this by setting the LVC 1V below current pack voltage, and if it doesn't work, then that's the problem. Then if you really want LVC protection from controller, you have to find the relationship of Actual Voltage to Programmed Voltage.

I hope it's as simple as that for you. I got nibbled on by that exact issue, but run or not run at all just last week, so it reared its head at home, not on the road. I can imagine your frustration.

John[/quote

I agree with John. Still sounds like a LVC issue. Try this: run the bike until it shuts down. Then reprogram the LVC on the CA to VERY low (as an experiment, so it doesn't affect anything). Then reflash your controller with the LVC VERY low. See if the bike now runs again. If so, its just your LVC setting in the controller. Like John said, with those modified controllers, you can't take the LVC programmed value literally. For example, on my controller if I want an acual LVC of 86.4v, I have to set the software at (86.4 +2)/2 = 44.2V. Hope this helps.

Itchy, John - LYEN gave me the "formula" for setting the LVC. Here's my set up currently:



Now, I do know that the LVC on the pack BMS is 2.0 Volts. According to the "formula" for the controller, I need to set LVC as: 2.0v * 24s = 48v, then (48V + 2)* 0.75

This gets me to 50 * 0.75, or 37.5

I will check no load, then the Ca LVC, Controller LVC. May just set both of those to zero and then take her out today before it rains.

Thanks guys.

Len

 

[Xanda2260]

5A no load! Something aint right there, thats rather high. Looking like Lukes nailed it...

What speed do you travel at? And what voltage does 24s of A123 come out as? Using the 74v that you cut out at, it looks like you're consuming around 50wh/mile. If you're shifting (35+) then its reasonable, but if you're only doing 25 or so, then its high.

Edit: just seen that you've switch halls, but not phase. Now I dont know if this applies to your crystalite hub but when I went to a Lyen contoller I had to switch the green and yellow halls, AND the green and yellow phase before it would work properly. This is a 9c though....

 

[Lenk42602]

5A no load! Something aint right there, thats rather high. Looking like Lukes nailed it...

What speed do you travel at? And what voltage does 24s of A123 come out as? Using the 74v that you cut out at, it looks like you're consuming around 50wh/mile. If you're shifting (35+) then its reasonable, but if you're only doing 25 or so, then its high.

Edit: just seen that you've switch halls, but not phase. Now I dont know if this applies to your crystalite hub but when I went to a Lyen contoller I had to switch the green and yellow halls, AND the green and yellow phase before it would work properly. This is a 9c though....

Actually, just came up from the shop after checking no load - motor peaks at 5A no load, but settles down to 3.93A @ WOT.
My lifetime average speed on this bike is 25mph, but I usually ride at a continuous 27-31mph.
24s fully charged is 87.6V, 79V nominal.

I will switch up the phase wires as indicated and report back.

I still don't see how this would affect the LVC cut off, unless it is sagging cells faster than the CA can read v min.

Thanks!

Len

 

[Lenk42602]

Edit: just seen that you've switch halls, but not phase. Now I dont know if this applies to your crystalite hub but when I went to a Lyen contoller I had to switch the green and yellow halls, AND the green and yellow phase before it would work properly. This is a 9c though....

So I am pretty sure i am hall/phase wired correctly. Checked out out my phase config. It is not what I listed above.

This is the actual hall combo

Mtr Ctr
B B
R R
Y Y
G B
B G

Motor is running properly, but stil at 3.9 amps WOT no load.

 

[dnmun]

a lot of people were not happy with that headway BMS because it would cut out for no reason. if you have to put it on the charger, the LVC sense could be screwed up and it cuts out at much higher than the 2.1V spec. the traces and the mosfet drivers are hidden under that black goop. maybe try using 5-6 volts of battery voltage from some C or D alkaline cells applied to the gate of the output FETs when it fails may be enuff to turn it on again and that would define the problem.

 

[Lenk42602]

dnum,

thanks for you input. you relentessly give people assistance with problems on this board.

My BMS is hardwired onto the pack, and two PCBs sandwhich the components in between. is there anyway I can diagnose its proper function with a meter, either while in operation or checking for continuity?

The BMS on this Cell Man Pack is a Headway? I can take some pics and post up shortly.

Len

 

[dnmun]

i try, it is always easier if it is someone else's problem. i have too many of my own here.

but i was thinking this is similar to the problems people complained about with the 12S headway BMS, cutting out for no reason. but the 12S headway BMS was made up of an 8S and 4S daughter board to make 12S. i was thinking then that there was a bad connection, cold solder joint maybe on the trace for the LVC that caused it to shut off. where the daughter board clipped into the motherboard. it may just be bad comparators for the cell level LVC, or even bad solder joints where they contact the trace for the LVC signal.

the idea is, if your BMS shuts off, to check to see what the gate voltage is on the output FETs, because if the output FETs are still turned on, the BMS is not causing the problem. but if the gate is off, then it woulda turned off because either it detected LVC, either correctly or falsely in your case, or it would be because it detected too much current through the shunt and that caused the comparator to turn off the output FET driver. but those are under the black goop.

once you open the BMS, you will have to unscrew the screws holding the heat sink down if you cannot reach the output mosfets, but they may be exposed from the side and you won't have to remove the heat sink plate. take some pictures when you get there, a lotta people have this BMS i think so they may have advice too. BOL

 

[Lenk42602]

okay thanks Dnum.

Gonna take a little step back for a few hours and the post up some pics later this afternoon.

Thank you,

len

 

[nicobie]

I'd try bypassing the BMS first now that you have eliminated having a false positive on the halls/phase wires. At least then you would know if the BMS is the problem. One thing at a time...

Good luck.

 

[Ypedal]

CA is calibrated correctly - I think. All i did to it was set the rshunt value. Is there anything else I needed to do? Obviously I set the speedo stuff up...

What is it set to atm ? and how did you obtain the proper value ?

If you had the hall/phase color combo wrong, the motor would be really.. REALLY hot at the end of a ride.. and i'd expect some occastional rough dead stop starts.. would also be higher than 4 amp no load. .. normally 1 or 2 amps no load but add a big fat heavy tire ( not perfectly ballanced ) and a scrubbing disk brake calipper it goes up.

 

[Lenk42602]

CA is calibrated correctly - I think. All i did to it was set the rshunt value. Is there anything else I needed to do? Obviously I set the speedo stuff up...

What is it set to atm ? and how did you obtain the proper value ?

If you had the hall/phase color combo wrong, the motor would be really.. REALLY hot at the end of a ride.. and i'd expect some occastional rough dead stop starts.. would also be higher than 4 amp no load. .. normally 1 or 2 amps no load but add a big fat heavy tire ( not perfectly ballanced ) and a scrubbing disk brake calipper it goes up.

hey Y,

the shunt value is set to 1.3333 mohm as per Lyen's input. I am certain the hall and phase wires are good. i have a moped rim and a motorcycle tire mounted, so it weighs a least twice as much as a regular rim/tire bicycle combo.

Also, this motor, wiring and controller never get hot. the motor only is slightly warm with the settings on my controller as indicated above, thats after continuos 25-30 amps at 30 mph for 5 minutes.

len

Len

 

[auraslip]

have you tried asking cellman?