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

[Lenk42602]

have you tried asking cellman?

I contacted Paul via email yesterday, and sent him a link to this thread. I am assuming he will respond when he sees the email and follows to this thread.

It is one of the reasons I am holding off opening up the bms to do any diagnosing. I want to hear his input before I do anything.

It just seems odd that the common cut off for the BMS is after a certain distance/ah used. Not excessive amps during battery cut off, and not low voltage.

Guess we have to wait and see. In the mean time I will be changing the LVCs on the controller and CA to rule them out as being the source of the issue.

Len

 

[Philistine]

Xanda2260 wrote: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 phase wire combo

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

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

I am confused by how you are explaining your hall and phase combo (you seem to be suggesting your are connecting red to black?). Isn't this a Lyen controller to an X5? Just to be clear for anyone reading, the only combo I have ever seen work anywhere for a Lyen to any Crystallite motor, is:

Halls:
Mtr CTR

B->B
R->R
Y->Y
B->G
G->B

PHASE:
Mtr: CTR

G->G
B->Y
Y->B

As far as I know that is the only combo to ever work between a Lyen and Crystalite motor, and it is all I have ever used.

A 9C is totally different with a Lyen, a 9C is the same on the halls, but with the phases the Yellow matches and others swap.

Cheers, Phil.

 

[Lenk42602]

Halls:
Mtr CTR

B->B
R->R
Y->Y
B->G
G->B

PHASE:
Mtr: CTR

G->G
B->Y
Y->B

As far as I know that is the only combo to ever work between a Lyen and Crystalite motor, and it is all I have ever used.

A 9C is totally different with a Lyen, a 9C is the same on the halls, but with the phases the Yellow matches and others swap.

Cheers, Phil.

jesus my brain is deteriorating - Phil I had (and have) my hall and phase wires connected as you've listed above. Not what i typed. need to do a bit more proofing before i post. I edited mine above to avoid confusion..

 

[Lenk42602]

gotta work tomorrow, but here's my plan for proceeding:

I have set the controller LVC to the equivalent of 30 volts.

I have set the CA LVC to 30 volts.

The Battery BMS should cut off at 48 Volts.

I rechecked the battery voltages right now - all cells are reading 3.61 volts, except for 1 cell, @ 3.48 volts.. I thought I checked each suffiicently, but I think I missed one with the probes this afternoon.

Does this mean I have a bad cell? Could this be the cause? Perhaps the BMS is in fact working - If this cell is bad and tripping the battery BMS LVC during discharge, it would all make sense, no?

Gonna leave my charger running on the bike until I get home tomorrow night and see if the BMS can balance the cells on full charge.

Will post up my results tomorrow evening. On Monday, if the weather cooperates I plan on riding the bike into work, and am preparing to pedal my 120 pound bike the last 1-2 miles in the event that the bike ceases to function again. That has been rather consistent.

I will be go pro'ing the operation - could be an ES first - almost like getting a UFO on video tape due to prior knowledge of its scheduled appearance!

Thanks to all who have helped so far. Perhaps Paul will provide some additional insight tomorrow.

Len

 

[dnmun]

take the charger to work. it is better for the battery to have plenty of margin if it is causing problems. charge up before you leave each end. 3.48 means it will take longer to catch up. not sure what the shunt current is on that BMS or how long it has charged.

 

[Xanda2260]

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.

Okydoke, not that then
Hope you get this sorted, I dont run a BMS so I'm afraid I'm out of ideas.
Good luck fella

 

[Lenk42602]

take the charger to work. it is better for the battery to have plenty of margin if it is causing problems. charge up before you leave each end. 3.48 means it will take longer to catch up. not sure what the shunt current is on that BMS or how long it has charged.

Good morning dnum.

I rode my pedal bike in today. (here at work checking in before the day gets started)The ebike is still at home on the charger. Gonna leave it charging all day today and tonight.

When I get home tonight I will post up the individual cell voltages.

Tomorrow (monday) will be the ebike ride-to-work-test.

Len

 

[SamTexas]

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.

3.93A is still too high for a no-load current at 79V nominal (310 watts). My 9c 2807 has a no load current of about 1.7A at 74V nominal (126 watts) at a wheel speed of around 45mph. I assume you have already ensured that there is no brake drag on the wheel.

I would first verify the accuracy of your amp meter. In this no load test where the current is relatively low, you can use a common multi meter to measure the current. Most of them supports up to 10A for 60 seconds. In short, let's first try to eliminate any possible problem associated with your current amp meter.

 

[John in CR]

Len,

Don't worry about the wiring. You have it right. I don't think the old X5's even have false positives due to the much lower pole count, and if it did it would be quick obvious and your motor would be long dead.

Don't sweat that lower voltage on one cell, it could be meaningless since voltage above that is just surface charge on the others and quickly disappears. It wasn't the BMS tripping from over-current, so it's really down to the LVC being too high or the BMS doing it's job, but since the controller is the new part of the equation my money's on the LVC. The alternative is a bad cell or BMS, so we're all pulling for the LVC, which you can check with a few different LVC settings before even setting out.

At least you picked up some extra exercise today. Your pedal bike has probably been feeling neglected. I almost glad I don't have one. 8)

John

 

[Lenk42602]

Len,

.... so it's really down to the LVC being too high or the BMS doing it's job, but since the controller is the new part of the equation my money's on the LVC. The alternative is a bad cell or BMS, so we're all pulling for the LVC, which you can check with a few different LVC settings before even setting out.

At least you picked up some extra exercise today. Your pedal bike has probably been feeling neglected. I almost glad I don't have one. 8)

John

I am pretty confident the shut down issue is either:
- bad cell
- LVC settings, either CA or controller
- Bad BMS

In order to address each of these:
- will be checking the "top charge" of all cells when I get home today
- I have already set the CA and controller LVC's lower than the BMS LVC of 48 Volts
- If the above two yeild no conclusions, then I am going to lean on Paul (cell man) for the proper steps in diagnosing the BMS.

Just FYI, I remembered vaguely some discussions on the stealth thread regarding issues with pack capacity, particularly not recognizing the range that should be available with the pack sizes. I looked over the thread last night ; in one instance it was in fact a bad cell. In another instance it was a bad BMS. The way the bad BMS was diagnosed was by taking a voltage reading off the battery leads and charge leads. if the weren't the same, the BMS was bad.

With my pack, on a full charge, the BMS yeilds the same voltage, battery leads and charge leads.

Tomorrow, after after operation Pedal my 120 pound bike home is underway, if the problem continues, I will then take a voltage reading on the charger and battery leads.

By the way, I have been pedaling my pedal - only bicycles much more this year than I have in the past several. About 15 pounds lighter than I was a this time last year.

Damn e-bike just isn't fully dialed yet..... :?

len

 

[nicobie]

The easiest way to check if the problem is your BMS is when the bike fails the next time jumper between the two big black wires on the BMS. One of them goes to the neg. on the battery and the other one goes to your controller. If the bike fires up, you have either a bad BMS or a bad cell. You'll want to check your cell voltages after the bike quits and before you charge.

I would also unplug your CA when testing the bike Monday.

Don't worry, You will get this solved sooner or later... :wink:

 

[auraslip]

just a note - you should be checking for a bad cell while the pack is at least partially discharged or better yet when it hits lvc. It's a pita on the road. I wish all battey pack makers included taps for cell-logs on their packs!

 

[Lenk42602]

ran the charger for 24 hours on the pack, hoping the bms would balance if there are any cell voltage differences.

cell voltages, after full charge:

3.58
3.57
3.57
3.33
3.57
3.57
3.57
3.58
3.57

3.57
3.57
3.58
3.57
3.57
3.57
3.58
3.58
3.58
3.57
3.57
3.58
3.58
3.57
3.58

Suspect cell?

Moving forward with the plan to isolate the BMS on the pack by essentially disabling the controller and CA LVC, and running the bike until it shuts down. I will then check cell voltages once BMS shuts down.

Unfortunately it looks like rain all day tomorrow so pedal bike again. Will have to wait to load test until weather clears up.

Perhaps Paul (Cell Man) will respond to the email I sent and post up here tomorrow.

Len

 

[amberwolf]

Do you have a single-cell charger? Or access to a lab-type power supply, that has adjustable voltage and adjustable current limit? If so, you could just try recharging that single cell, and see what happens.

 

[John in CR]

I'm with AW. See how much that low cell will take with a single cell charger. 3.33V after full balance charging doesn't sound good to me. 3.48V you mentioned before seemed like no problem, but 3.33V seems low even for a slightly weak cell. If it takes a nice chunk of ah to charge to 3.65 , then something is amiss with the balance charging, so a BMS issue. If it takes just a little to get there, then it seems like a cell issue. I don't like coincidences though. New controller and suddenly can't go the distance, made me think it was just the controller LVC, so I'm still pulling for just a setting issue. That 3.33V made me cross my fingers for you.

 

[Lenk42602]

Do you have a single-cell charger? Or access to a lab-type power supply, that has adjustable voltage and adjustable current limit? If so, you could just try recharging that single cell, and see what happens.

thanks aw. i could find a single cell charger pretty easily i guess, but as I stated befre, i do not want to start ripping apart the pack before Paul (Cell Man) responds to this discussion. This was an extremely expensive battery pack ,and I do not want to take any additional action on the pack itself in order to reduce the variables that can be attributed to its malfunction.

I don't like coincidences though. New controller and suddenly can't go the distance, made me think it was just the controller LVC, so I'm still pulling for just a setting issue. That 3.33V made me cross my fingers for you.

hey john,

I did in fact change the controller, however, I think I was experienced the same pack behavior before hand. On my first long ride to test capacity(late january 2012) although I had a different controller, the bike's power cut off at about 14 miles, and the same amount of AH used. I go pro'd that ride, posted on my build thread, and can make out the reading on the cycle analyst right at the point of failure.

very similar AH used and distance traveled, then no power.

regardless, I am still planning on the shake down as soon as the roads dry up here.

Hopefully I can eliminate a few more suspects after that.

I am gonna guess that it was Professor Plum, in the Kitchen, with a knife.....



Thanks,

Len

 

[amberwolf]

thanks aw. i could find a single cell charger pretty easily i guess, but as I stated befre, i do not want to start ripping apart the pack before Paul (Cell Man) responds to this discussion. This was an extremely expensive battery pack ,and I do not want to take any additional action on the pack itself in order to reduce the variables that can be attributed to its malfunction.

Well, no need to tear anything apart to single-cell charge. Just connect the charger to that cell only, thru it's balance wires.

But I understand wanting to wait till Paul chimes in.

 

[Dauntless]

There's something mala en se about all that money tied up in an unreliable battery system. For the time being I think I'll stick to the cheaper, more stable, less sexy approach. Although I'm hoping someone I know will jump in to electrics and just have to go LiPo, so I can learn all these lessons with him paying for it.

If you MUST ride your bike into the river, save it for the Red Bull Flugtag. There's always a way to salvage some fun out of ANYTHING.

http://www.funvblog.com/2010/10/30/25-cool-photos-of-red-bull-flugtag/

 

[Lenk42602]

Alright folks, test diagnose run is complete. Have to be at work @ 2:30 pm.

CA LVC is Set to 0 volts

Lyen Controller LVC is set to 24 volts (lowest volt cutoff capable of controller)

Upon departure,cells all reading 3.47~ 3.48 volts, except for the one cell @ 3.33

Go pro will be running, pointed at the CA.

 

[Lenk42602]

Test results:

lost power @ 8.82s AH consumed
@ 13.81 miles

Cell voltages, immediately upon power loss:
All cells @ 3.27~3.28, runt @ 2.55

Bolted Side panel back up, pedaled back home.

Checked voltages across all cells, all cells @ 3.27~3.28, but the runt was @ 2.78.

(video to be insereted here in a bit.)

 

[Ypedal]

yep, that 2.55 is your guilty party.. time to top it off with a single cell charger ( use a cell ph charger if you have to.. pull the plug at 3.7v ) and time how long it takes to get there.

a " runt " can still function if it just needs topping off once in a while, but if there is a very bad cell in that group ( is you pack made from 26650 cells ? ) it may need surgery.

also possible the BMS is causing a drain on that channel leading to the problem, not the cells themselves.. that requires no-fun testing.

 

[liveforphysics]

Its possible that a tab weld got jostled loose off one of the parallel cells in that group.

 

[Lenk42602]

Y,

The pack was constrcted with the newer 20ah prismatic cells from Cell Man. AMP20M1HD-A, I think?

Amberwolf suggested I charge right through the balance wire? Should I soure a connector that I can use to just plug in to the corresponding leads in the balance plug?
What do you reckon it will take to bring the cell up to 3.7v, time wise?

Is this going to be an ongoing issue, or will single cell chraging bring the cell into balance with the rest of the pack?

Luke,

wouldn't the voltage of the pack reflect 23s rather than 24s if that were the case?

BTW, I reveiwed my go pro footage, but its not really compelling wrt the issue at hand. Not gonna post it.
Thank you,

Len

 

[Ypedal]

being a single 20ah cell is easier to deal with than multiple 26650 cells in a parallel group.. however it's a bigger cost/cell to replace if it's a dud..

a low and slow self discharge cell happens, it can still work.. but you need to figure out if it will take energy or not. ie : permanent loss in capacity vs slow discharge you can deal with using a single cell charger.

yes, using the ballance leads is fine.. no more than 2 amps tho. * Edit .. depends on gauge of wire used on ballance leads.. but usually fairly small stuff...

edit 2, as luke pointed out, a bad connection on a ballance lead can show as a low voltage reading.. worth checking.

 

[Lenk42602]

a low and slow self discharge cell happens, it can still work.. but you need to figure out if it will take energy or not. ie : permanent loss in capacity vs slow discharge you can deal with using a single cell charger.

yes, using the ballance leads is fine.. no more than 2 amps tho.

ah, I see. so i need to now determine if the cell is even capable of being brought up to full charge. Assuming the BMS is working, given that the BMS has successfully charged all the other cells to full capacity (all measured 3.57~3.58 V at full charge, while the runt measured 3.33V) wouldn't we have already etablished that the runt has a permanent loss in capacity?

feels like movement towards a solution is unfolding.......

Len