PDA

View Full Version : I need help!



Richard230
13 April 2011, 1531
Well, I installed four new batteries in the upper pack of my bike today. Before I did that I checked over all of the other 20 batteries on the bike and they all had 3.32 volts, plus or minus .01 volt. The new batteries all had 3.30 - 3.32 volts. Unfortunately, installing the new batteries just made things worse. :(

As soon as I hit the brakes while riding down the hill from my home, the power system shut off for the usual 3 minutes after I turned off the ignition switch. It did this every time I hit the brakes. This has been a problem for me ever since I installed the MiniBMS system, but it had been more random and would only occur when coming to a stop from more than 25 mph. So I was pretty sure that the problem was being caused by the regen system spiking the voltage and either the controller or the BMS was shutting off the power, until 270 seconds had expired (which seems to be programed into the controller when it detects a fault).

So I then defeated the regen system by pulling one of the wires to the relay (the black box in front of the DC-DC converter, preventing it from activating. That seemed to help for a while and I was able to brake without the system shutting down for several brake applications. Unfortunately, things didn't stay that way and the bike shut down again when I braked from about 25 mph and after traveling 2 miles. It then stopped one more time on the way home.

The other problem that I seem to have is that the pack voltage dropped from 80.6 volts to 77.5 volts after traveling only one mile at 25 mph on level ground, not using more that 50 amps at any one time and draining just over 1 amp hour. The at rest voltage stayed about there when I turned off the throttle. Upon arriving home, after using 6 Ah, I pulled off the seat and tank (the side fairings were already off) and by the time I had done that and checked all of the battery voltages, they were all over 3.30 volts and the system was showing 79.6 volts. There was no indication of a bad battery that I could find (nothing was swelled up, unlike two of the batteries that I had replaced and all the cells seemed to be well balanced).

Now I have a new problem. The charger will not stay on unless the ignition is on. I have to turn on the ignition switch to get it to charge, but as soon as I turn it off the charger relay light goes off and so does the charger. I have tried it several times over the past couple of hours, without success. Before I installed the four new batteries, it would charge normally - although the ignition had to be turned on and then off to get the charger relay indicator light to turn on so the charger would come on.

Since all of these problems started when I installed the MiniBMS, the only idea that I can think of is that I may have wired something incorrectly, when hooking up the head-end board. We did find a mangled wire connected to the ignition switch and replaced that, but doing so didn't seem to make any difference.

I was hoping that someone could look at the attached photos and let me know if something looks like it has been hooked up incorrectly. I would also appreciate any other ideas that you might have that could solve my headache. Thanks.

cycleguy
13 April 2011, 1600
Richard, the Mini BMS needs a continuous 12V supply, as well as a 12v ignition switch input. Without the continuous 12v input, it will not ground the charger relay to activate your charger. Refer to the linked user manual to double check your wiring.
http://minibms.mybigcommerce.com/template/files/MiniBMS%20User%20Manual%20-%20Distributed.pdf

BTW, how are you getting your 12v continuous power supply, aux battery?, dc/dc always on? or tapped into 4 series cells?

frodus
13 April 2011, 1605
Well, I installed four new batteries in the upper pack of my bike today. Before I did that I checked over all of the other 20 batteries on the bike and they all had 3.32 volts, plus or minus .01 volt. The new batteries all had 3.30 - 3.32 volts. Unfortunately, installing the new batteries just made things worse. :(
Voltage is not an indication of battery status/SOC or health. It could just be surface charge. A battery can settle to its nominal voltage fairly quickly. It doesn't tell you anything about what it does while loaded. I bet one (or more of them) are bad, but the balancers keep you from seeing that, because it tries to balance them.


As soon as I hit the brakes while riding down the hill from my home, the power system shut off for the usual 3 minutes after I turned off the ignition switch. It did this every time I hit the brakes. This has been a problem for me ever since I installed the MiniBMS system, but it had been more random and would only occur when coming to a stop from more than 25 mph. So I was pretty sure that the problem was being caused by the regen system spiking the voltage and either the controller or the BMS was shutting off the power, until 270 seconds had expired (which seems to be programed into the controller when it detects a fault).
I bet you're hitting LVC on one of the cells.


So I then defeated the regen system by pulling one of the wires to the relay (the black box in front of the DC-DC converter, preventing it from activating. That seemed to help for a while and I was able to brake without the system shutting down for several brake applications. Unfortunately, things didn't stay that way and the bike shut down again when I braked from about 25 mph and after traveling 2 miles. It then stopped one more time on the way home.
I bet you're hitting LVC on one of the cells.


The other problem that I seem to have is that the pack voltage dropped from 80.6 volts to 77.5 volts after traveling only one mile at 25 mph on level ground, not using more that 50 amps at any one time and draining just over 1 amp hour. The at rest voltage stayed about there when I turned off the throttle. Upon arriving home, after using 6 Ah, I pulled off the seat and tank (the side fairings were already off) and by the time I had done that and checked all of the battery voltages, they were all over 3.30 volts and the system was showing 79.6 volts. There was no indication of a bad battery that I could find (nothing was swelled up, unlike two of the batteries that I had replaced and all the cells seemed to be well balanced).

So you dropped to 3.22V a cell. I don't know whats you think is wrong with that. you'll ALWAYS have voltage sag. As soon as you load the batteries, they'll drop below nominal, which is 3.2ish anyway. 77.5V/24=3.229V. Well within normal limits. You should read up about voltage sag. Any time you load a battery, the voltage drops a little. It'll stay there until you release the load, at which time it comes back up. Your batteries are COMPLETELY normal.


Now I have a new problem. The charger will not stay on unless the ignition is on. I have to turn on the ignition switch to get it to charge, but as soon as I turn it off the charger relay light goes off and so does the charger. I have tried it several times over the past couple of hours, without success. Before I installed the four new batteries, it would charge normally - although the ignition had to be turned on and then off to get the charger relay indicator light to turn on so the charger would come on.

Didn't you say you're powering the BMS off the first 4 cells? Isn't that supplying the 12V to the miniBMS? The minibms controls the charger, right? So if its not getting power, does the charger stay on? Make sure you check this, because if the miniBMS doesn't work while charging, what's the point?



Please draw a schematic of your system. Pictures don't tell much, its hard to see. Much easier if you trace things out, and much easier for us to debug (since you've changed stuff on the bike from stock)

cycleguy
13 April 2011, 1638
As Travis said, you may be loosing some more cells and going below the min. controller voltage. The BMS would not cut off the throttle or the batteries, since, as you stated earlier, those BMS functions aren't yet enabled.
If one or more of those 4 cells that you are tapping to get 12V from are near death, you may not be getting full 12V to the BMS to trip your charger relay on.

podolefsky
13 April 2011, 1641
Man, that's frustrating. Sorry this bike has been such a pain.

Like Travis said, sounds like your batteries are all normal. My pack is about 80-82V fresh off the charger, settles in to about 77V after some driving, and floats back up to 80V or so after sitting for a few minutes. The fact that all your cells measure above 3.3V after some discharge means they might be fine. Or you might have a couple that have very little capacity, but you won't see it until you discharge them more. Lithium cells stay around 3.2-3.3V all the way down to 80% discharged. Even if one of your cells had only 20Ah left, you would need to use up 16Ah to start seeing the voltage drop a lot.

This sounds like something where you'll have to go through piece by piece and eliminate possible explanations.

System shut down on braking does sound like regen spiking the voltage, but you said it still did this after you disconnected the regen relay. So could be LVC, but not sure why that would happen on braking - if anything, it seems like it would happen when accelerating and the voltage sags.

Does the miniBMS have control over the main contactor, or some way to shut the system down? If you can temporarily rewire it so that the BMS can't shut it down, you could eliminate that as a possibility.

Also like Travis said, it is much easier to diagnose what's going on with a drawing of the system.

DaveAK
13 April 2011, 1648
If you're (maybe) hitting LVC under braking try disconnecting any brakelight switches. Obviously test this off road. I can't see why LVC, or the possibility of under powering the miniBMS could happen under braking unless something was creating a significant draw on the batteries. And the only thing I can think of that would effect voltage besides regen is the brake electrical system, which would have to be the bulb or a short.

DRZ400
13 April 2011, 1758
Sounds more like regen is causing the batts to trigger the HVC? Incorrectly wired at the HU would cause cutoff (look like LVC alarm)??

Richard230
13 April 2011, 1814
The 12V system, including the BMS, is powered by a DC-DC converter, which can be seen at the top of the attached photo. To my knowledge it is always on, as the charger light stays lit all the time, even with the ignition off and the AC plug disconnected from the wall outlet. The charging system was working OK last night and this morning. I took the bike out yesterday and used 10 Ah going 6 miles and it charged for 2 hours before turning off.

When braking with the regen on, I am seeing about 20 Amps when slowing from 25 mph and voltage spiked to 84 (as near as I could tell by glancing at the CA while I was coming to a stop). After I disconnected the regen, there was no voltage change, other than a slight rise with the throttle shut, as the battery pack returned to at rest voltage. Voltage sag under light throttle would bring the pack down to around 68 V, with a 50 amp draw. I did not try using full throttle, which will drag the pack down to 51 volts, as I didn't want to damage my new batteries until I had an idea how they were performing. The tail and headlights together draw about 0.3 amp, according to the CA, so that doesn't sound like a problem.

My personal problem is that I can't read an electrical schematic, much less draw one. There is no schematic for the GPR-S to my knowledge and in any case, my bike is different from the stock versions, as they do not use a sepex motor, regen or Sevcon controller. I also can't track the wiring very well and I have only a little understanding how the propulsion, charging and BMS systems work. While I do have the Sevcon controller and programmer manual and the MiniBMS manual, I really don't understand them enough to have enough confidence to make changes. I had a friend do the wiring work for me. He builds radio stations but has no experience with electric vehicles.

I was hoping someone who is familiar with the MiniBMS could offer an opinion by looking at the photos above if the head-end board was hooked up properly. My friend did it for me and I can't even tell if the HVC and LVC is activated.

To be honest, all I ever really wanted to do was to ride an electric motorcycle, I never wanted to build or repair one. That was why I bought a fully assembled DOT approved EV. Unfortunately, I really had no idea what I was getting myself into.

podolefsky
13 April 2011, 1918
Hey Richard, sorry this sucks. This might not be that comforting, but consider it an opportunity to learn a whole bunch about your bike (even if you didn't think you wanted to).

I'm not familiar with the miniBMS. I'm pretty sure the Sevcon has a high voltage limit of 84V, so that might explain the fault on regen.

Let's try to pin down some things:

- even with regen disconnected, you have a problem with the system shutting down when you brake, correct?

- does it shut down any other times, or just when braking?

- you said the lights only pull 0.3A. Is this with the brake light on, or just the tail lights? The brake light is on a different circuit, and if there's a short it could pull a lot more when you hit the brake lever. If the BMS is run from the DC-DC, it could be that it is losing power somehow (it could even be a bad DC-DC converter).

- is your DC-DC isolated or common ground (basically, does it have 4 wires or 3 wires)?

- is there a way you can eliminate the miniBMS from controlling anything so you can narrow down whether it is actually the BMS or something else?

That's a start - if you can answer those, we can go from there.

podolefsky
13 April 2011, 1921
Another side note. You must have some very weak cells somewhere if you are sagging to 68V at 50A. That's less than 1C. My GBS pack sags to 68V at 400A. At 50A, it hardly blinks (but my batteries are pretty new).

markcycle
13 April 2011, 1953
What is the voltage rating of the Sevcon controller?

podolefsky
13 April 2011, 2004
What is the voltage rating of the Sevcon controller?

80V link (http://www.sevcon.com/PDFs/Powerpak_PPK0106EN.pdf)

But manual says max is 100V, and default is 97.5V. And I think EMS and Thunderstruck (http://www.thunderstruck-ev.com/index.php?dispatch=products.view&product_id=31) set them to 84V to protect the motor (which is actually rated to 48V). That's why I thought the 84V on regen might be triggering a fault.

DaveAK
13 April 2011, 2005
What is the voltage rating of the Sevcon controller?
80V rated with an absolute max of 100V

magicsmoke
13 April 2011, 2010
Hi Richard, keep the faith. I'm sure this is something so obvious it's just being overlooked.
A couple of things I think to get us started nailing it down.

Firstly, the BMS head end is powered from the 4 cells nearest to it as we can see in your img 2/5 not the DC-DC.
Second, only one of the head end interrupt relays is wired (HVC) and only connects to the solid state relay screwed to the charger which I'm presuming therefore is the charger cutoff and if so can't shut the bike down. i.e. the bms has no bearing on the controller.

Staying with your charging issue for the moment, the minibms manual states that if any of the cells are below their LVC threshold then the ignition key needs to remain on until they have picked up. This overrides the 'loop' wire interrupt.
So it's possible that either you really did have one or more low cells or there is an intermittency in the loop wire.

As for the controller cutting out, what model is it and any chance you can take a photo of the wiring to it?

good luck
Rob

edit: sorry guys, I took so long typing that you've already got the controller info I was after :)

chef
13 April 2011, 2042
Sounds more like regen is causing the batts to trigger the HVC? Incorrectly wired at the HU would cause cutoff (look like LVC alarm)??
I second that. Regen would trip the HVC if one of your batteries is weak. With diminished capacity, its voltage would spike as it fills up quicker than the others. This would be exacerbated by running immediately after a full charge.


Hey Richard, sorry this sucks. This might not be that comforting, but consider it an opportunity to learn a whole bunch about your bike (even if you didn't think you wanted to).

I'm not familiar with the miniBMS. I'm pretty sure the Sevcon has a high voltage limit of 84V, so that might explain the fault on regen.
My Sevcon has operated well above 84v. The previous charger topped out at 96v which was killing the cells unfortunately. I ran it hot off the charger to no ill effect on the Sevcon.

DaveAK
13 April 2011, 2057
I concur with Rob. The miniBMS isn't connected to the controller in any way.

1. Your charging issue. As Rob says the system is designed such that a LVC and a HVC both open the normally closed loop. It makes a determination as to whether to look for LVC or HVC depending on whether the ignition is on or off. If you can only charge with the ignition on the BMS is noticing an immediate LVC or HVC condition. Either one or more cells is under or over the threshold or there's a wire fault. If it's a low cell then you must at some point after the cell has recovered switch the ignition off or the BMS will not be able to detect a HVC condition.

2. After the bike has cut out, is it possible to view the controller and see if the diagnostic LED is flashing, before switching off and restarting? If you have the calibrator try looking at the fault log. All the conditions are logged.

3. Typically the brake circuit is connected to the controller via a relay as it appears yours is, so I'm guessing it's nothing to do with the brakes vs. the controller as you've disconnected the relay. However, it is possible that the brakes are connected, (mine will be, because I'm atypical), so I would try with the brake swicth wires disconnected at the switch. Again, best not to do this on the road!

Barron
13 April 2011, 2058
--- deleted --- redundant

DaveAK
13 April 2011, 2105
Can you see the 12 pin connection to the controller? If we can determine what's connected that might shed some light on the intended operation.

podolefsky
13 April 2011, 2139
OK, I read over the miniBMS manual, and from what I can tell your BMS shouldn't be able to shut the bike down. You would need the BMS connected to the throttle, and it looks like you don't. (Even if it were connected, it would only reduce throttle, not shut it down completely...if I understand the manual right).

DaveAK
13 April 2011, 2153
OK, I read over the miniBMS manual, and from what I can tell your BMS shouldn't be able to shut the bike down. You would need the BMS connected to the throttle, and it looks like you don't. (Even if it were connected, it would only reduce throttle, not shut it down completely...if I understand the manual right).
Correct! Gold star for Noah! :D

podolefsky
13 April 2011, 2201
Correct! Gold star for Noah! :D

...and now I see you and Rob already basically said this. :rolleyes:

podolefsky
13 April 2011, 2202
--double post--

DaveAK
13 April 2011, 2205
Don't mind me Noah. I need help too. But more of the kind where you have to check in somewhere. :)

markcycle
14 April 2011, 0424
I'm not sure I have much to add here but I did have a few questions.
Its strange that with a few new cells and maybe a bus voltage a few volts higher, the problem is worse.
The Sevcon error log would be a huge help, even reading out the LED error code would be a big help

A obvious point did you check all the connections as related to the controller and motor (don't get mad at me if this question inspires frustration).
Next question is it possible that the back EMF of the motor is raising the controllers internal bus voltage to above the 84 volt set point and causing the controller to shut off, since this seems to be a no load coasting problem?

I don't see anywhere as much voltage rise as you do with Regen even at 25 amps of regen my pack stays at 3.3 volts per cell. Seems the Ri of the cells are high as has been mentioned, or the wiring is to thin and there is a high voltage drop across the leads or both.

Richard230
14 April 2011, 0836
I want to thank everyone for their response and comments.

I am going to try running the bike with the brake lights disconnected today and I am going to hook up the controller programmer to see what happens. If there are any LEDs on the controller, I can't see them. Nothing shows up when looking at the bike in the dark, other than the green lights on each BMS board and the red indicator light on the charger.

Although my friend used the programmer to check for faults when he was working on the bike and dialed back the maximum current from 250 amps to 150 amps, I have never used it and am not quite sure how it works. My friend told me that the programmer showed a intermittent interruption in the current to the controller and that was what was causing the power to be shut off for 270 seconds. So perhaps the brake light is what is causing the problem. However, the brake light does work fine with the ignition on in the garage and I would think that if there was a short, the light bulb would be burnt out. The other problem is that the "stalling" does not happen every time, just when braking from more than 25 mph - or so it seems.

Since the bike ran just fine before the original BMS master board failed and shorted out a couple of batteries, and all of my current problems have occurred since we pulled out all of the battery boxes and installed the individual MiniBMS boards, the logical conclusion is that something didn't get hooked up properly when everything was reassembled, or something was connected incorrectly, or a wire was damaged during the work. So far I haven't been able to locate a physical problem - although I did find a loose power cable connection yesterday, which when tightened didn't make any difference.

What has really got me puzzled is why I am not able to get the charger to work when it worked just fine before I replaced the four batteries in the top pack yesterday. I did my best to reinstall the wiring just the way it was before. I also looked at all of the BMS boards and everyone has a green light, which indicates that they are receiving current and have not cut out. When a battery would drop below 2.5 volts in the past, the green LED would turn off and the charger would not work until it came back on. But now all of the lights are on, all of the batteries have good voltage and yet it is not charging. I must have done something, but I don't know what.

So my plan today is to disconnect the brake light, go ride around my block for a while to see what happens and run the battery pack down some more. Maybe using more Ah will cause the charger to function again.

Thanks again for your advice and suggestions.

DaveAK
14 April 2011, 0911
Here's a quick check you can do. Check that you have continuity in the cell loop. The individual cell boards might show that the cells are fine, (green light), but if you have a loose connection between any of the boards this will show as an LVC or HVC event. Personally I'm not happy with my crimp connections and I'm expecting to have to redo them at some point.

Another test would be to check for a signal on the LVC terminals. This should show 0V under normal conditions or 12V when LVC has been triggered. If your cells are all showing 3.3ish volts, but you're getting an LVC signal it points to an issue with the BMS/wiring. This is for your charging issue, not the braking one.

billmi
14 April 2011, 0933
Richard, you’ve bought a quality BMS, but it is only partially installed. The result is that it’s not providing protection for your batteries. The MiniBMS isn’t wired in a way that could be causing your controller shut-down issues, those are another problem altogether that has been addressed in this thread (pack level LVC in the controller, other wiring issues, etc.)

The Mini BMS has three primary functions – protect your batteries from over-discharging (Low Voltage Cut-off or LVC), protect them from over-charging (high-voltage cut-off or HVC) and bring your pack into balance over multiple charging cycles. Of these three functions LVC and HVC are the two that protect your batteries, and as you are using your motorcycle now, you are not getting this protection. Over-discharging or over-charging will damage LiFePO4 cells.

The Mini BMS cell boards check for HVC or LVC conditions and report them to the main board, which takes steps to prevent over-discharge or charge damage if it is installed and used correctly, and if you pay attention to it.

The first problem I see is that your MiniBMS is only partially installed and because of this, it has no way to tell you to stop riding when cells hit LVC. Especially with cells of different age, batch, levels of abuse, etc., your pack is not likely well balanced, so there’s a pretty good chance of individual cells hitting dangerously low discharge levels (indicated by their low voltage) before the whole pack voltage is noticeably at the bottom.

The MiniBMS head board is designed to warn you about LVC with a buzzer, and by reducing your throttle, but you have neither of these systems hooked up. When you ride you have no warning that a cell has hit its lower voltage limit and is being damaged by continued use.

You’ve mentioned having to turn the ignition on to charge. The MiniBMS head board won’t allow the charger to charge if it is receiving an alert from the cell boards (it doesn’t know the difference between an LVC or HVC, it just knows there is an alert). Turning on the ignition bypasses the HVC cut-off for the charger and allows you to charge when some cells are below LVC levels (which it sounds like at the moment they aren’t since you can see each board is green). Does turning on the ignition allow charging right now? If so, the BMS head board and its link to the charger are OK – the problem is it’s seeing an alert, which would be caused by a cell board HVC or LVC (neither of which is happening) or a break in the loop circuit that links all the BMS modules together.

If you can charge with the ignition on but not when it's off, we know there’s a problem in the MiniBMS monitoring loop – don’t leave the charger running with the ignition on. If you leave the ignition on through the rest of the charge cycle, the bypass will still be in effect and you will have no protection from over-charging. There is a paragraph of hard-to-miss red text describing this in the MiniBMS manual.

What about the status LED on the relay between the charger and the MiniBMS headboard? Is that turning on, and if so, under what conditions? That LED means the MiniBMS is signaling the relay to run the charger.
Really, wiring in a buzzer and or an LED on the buzzer circuit would be huge, because then you’d be able to know when the MiniBMS is getting an alert signal, and be able to better react to what is causing it, or eliminate that as a cause for the charging problem.

The MiniBMS manual is quite thorough and clear in explaining how the MiniBMS works, and it sounds like taking the time to read the whole thing would help quite a bit.

DaveAK
14 April 2011, 0940
What Bill said. The manual describes what kind of buzzer to get and they're available at Radio Shack for a couple of dollars. Since you have the newer head end board with screw terminals it's simply a question of wiring in the buzzer.

harlan
14 April 2011, 0948
Yep. What they said.

Richard230
14 April 2011, 1314
I spent the morning reading the BMS manual, but other than the buzzer not being hooked up, I can't see any other areas that were connected incorrectly. If I could get the charger relay to come on and the charger to work, I could move on from there. I just can't figure out what changed when I pulled the old batteries out on the single box and replaced them with new batteries. That is when the charging system stopped working correctly. I did not short out any BMS boards, every one is showing a green light. However the red charging relay light comes on when the ignition is turned on and immediately goes off (turning off the charger, of course) when the ignition switch is turned off. I have not left the ignition on for more than a minute with the charger plugged into the wall outlet. I tried disconnecting the BMS loop wires, but that had no effect. Is there any way to activate the charger relay by bypassing the BMS board?

I did discover though, that disconnecting the stop light wires that connect to the brake lever switches prevents the bike from stalling. I removed the blue wire from the front brake assembly connector and left the wires connected to the rear brake lever. I could stop with no problems using the front brake, but as soon as I tried the rear brake the bike stalled for 3 minutes. So the problem seems to be in the stop light circuit, or the part of it that activates the regen system. This seems like something that can be fixed, but I need to get the batteries charged, otherwise I won't be able to ride enough to see if the stalling issue is resolved. I'll just have a lump of dead batteries. I figure I should have about another 15 Ah left in the pack before I run out of juice.

Thanks again for everyone's help.

magicsmoke
14 April 2011, 1326
Hi Richard, I know you're getting bombarded with suggestions and I'm trying not to add to that but here goes, one more :)
Can you disconnect the yellow 'loop' wires from the head end board and short the terminals at the pcb together with a jumper wire?.
This will bypass the individual BMS pcbs completely and will confirm the operation of the ignition / charger circuit.
This is obviously a debug measure and should not be left in position for unattended charging.

Rob

DaveAK
14 April 2011, 1326
You can activate the charger by supplying 12V to the charger relay in place of the BMS connections, but you won't have any HVC protection. If, and it's a pretty big if in your case, the batteries are reasonably in balance you might find the charger shuts down before a HVC event anyway, but I wouldn't trust that at all. So if you're going to bypass the BMS to get the charging started you're going to have to keep an eye on the voltage, not just the pack voltage but the individual cell voltages. You'll be able to see if any of them are shunting by watching for the red LED on the balance boards.

As for the brake issue I'm not sure what the next step is. I'd like to know how the Sevcon is wired, but it looks like the connector is buried in the bike out of sight. Any way to get a picture of the low voltage wiring connector?

Edit: Try Rob's suggestion of replacing the yellow loop wire with a jumper wire first. That's a quick and easy one.

Richard230
14 April 2011, 1458
I checked a couple of things during the past hour (in between mowing my lawn). I verified that the BMS loop has connectivity. Then I checked the LVC/HVC warning buzzer contacts on the head-end board with a voltmeter. I got a reading of 12.4 volts both with the ignition off and with it on. So it appears that if I had the buzzer hooked up it would be buzzing all the time and would be keeping me awake at night. I have a feeling that might be significant. But I still can't find any batteries that have a low voltage, nor do I see any BMS boards with a red light, or no light.

I would like to try Magicsmoke's suggestion. Is the "pcb" what I have been calling the "head-end" board that is stuck to the back of the upper battery box? I can do that.

Attached are the best photos of my controller that I could get without my camera becoming permanently stuck in the works.

I should also add that I verified that the BMS is being powered by the upper four batteries and that the voltage to the BMS head-end board is 13.10 volts.

DaveAK
14 April 2011, 1510
That 12.4V is very significant. It basically says that the BMS thinks there is a continuous LVC or HVC event. If the loop has connectivity then there are two possibilities I can think of for this:

1. There is a bad cell, and you're wrong.

2. You're right and the cells are fine, and the BMS is wrong - indicating a fault.

If we assume that since you can see all the lights are lit green and have measured the cell voltages, but the BMS is still showing an LVC/HVC event then there's a fault in the BMS. This could either be a faulty head end board, or it's incorrectly wired up. Anyone have any suggestions on how to test this? If you jumper across the cell loop terminals like Rob suggest then you definitely should not get a reading at the buzzer contacts. This would indicate a faulty board. (The PCB Rob refers to is the head end board.)

From your last picture I can immediately see some wiring that I wouldn't necessarily expect to see. Doesn't mean it's wrong, just that it's not wired as I would have expected. I don't have the manual with me right now to confirm the pins I can see. Give me 4 hours and I'll post some more thoughts on this.

magicsmoke
14 April 2011, 1513
Is the "pcb" what I have been calling the "head-end" board that is stuck to the back of the upper battery box?

Yep exactly, it is the head end. I was just making it clear that we want to short the screw terminals at the pcb rather than shorting the two unplugged yellow wire ends.

Rob

Richard230
14 April 2011, 1522
I just tried connecting the BMS loop wires together (the yellow wires going to cell loop terminals) and this had no effect. The charger relay did not come on nor did the charger stay on when I turned off the ignition. Connecting the wires together had no affect on the voltage reading at the buzzer terminals.

I wonder if I might have damaged the board when I replaced the batteries?

DaveAK
14 April 2011, 1527
Like Rob says, you need to put a jumper wire across the terminals on the board, not connect the two yellow wires together. Just remove the yellow wires from the head end board and replace them with one piece of wire, one end in one terminal and the other end in the other terminal. This will mimic the closed loop that you should have.

jpanichella
14 April 2011, 1528
I just tried connecting the BMS loop wires together (the yellow wires going to cell loop terminals) and this had no effect. The charger relay did not come on nor did the charger stay on when I turned off the ignition. Connecting the wires together had no affect on the voltage reading at the buzzer terminals.

I wonder if I might have damaged the board when I replaced the batteries?

Do you have an account over at DIYElectricCar? Dmitri (who created MiniBMS) posts over there and is always quick to reply to MiniBMS questions. He's a top notch guy.

Richard230
14 April 2011, 1532
Magicsmoke, I just re-read your last message and this time I shorted the terminals, instead of the wires. That turned on the charger and it is now charging with the ignition off.

Does this mean that the charger will not turn off? Or do I just need to keep any eye out for red lights? If there are any red lights on the BMS boards I they should be very visible tonight in my dark garage. Thanks for the tip. At least I can now charge my pack back up. Do you have any idea what this signifies? Could it be a bad component on the board and should I buy a new one, or should I do something else to fix this problem?

DaveAK
14 April 2011, 1541
Richard, sounds like we're getting somewhere. With your charger working in this fashion it does mean that it will not switch off. You must keep an eye on the process so as not to over charge any cell. If any of the red lights come on I suggest that you stop charging and check voltages. It sounds as if the BMS head unit is good and the fault lies in the cell loop.

Do you have a digital multi meter that has a continuity test? The cell loop doesn't actually supply any kind of voltage to the head end board. While the wires are disconnected can you check that you have continuity with a tester that will either light up or beep if the circuit is complete?

DaveAK
14 April 2011, 1548
Another question on the braking/controller issue. Can you hook up the programmer and tell us the Digtal IO setting under Traction Setup. When you hook up the programmer with the bike on you should get the screen that says Traction OK. Push the down arrow once, then push the right arrow until you get to the Traction Setup menu. It might skip a few options. When you get to the Traction Setup menu push the down arrow until you see Digital IO. It should be the second option. What number does it give you?

Richard230
14 April 2011, 1548
Hi Dave. My multimeter does have a continuity tester. I did check the continuity of the loop by connecting the tester across the wires and it beeped.

Fab man
14 April 2011, 1554
Richard please don't take this the wrong way, but I'm concerned about the tidiness of your install. In your first photo, near the center top, the connecting bars appear to be way too close. A bridge across this what appears to be less than a 1/4" gap would cause a dead short of two cells. Indeed, is that an arc nick in the upper left corner of the upper right bar?

In the same photo, I would be concerned about the copper traces on the bottom of the BMS circuit boards shorting out on the connecting bars.

I'm not the tidiest person myself. But, I quickly learned that batteries and other high power systems are places where you really want to be extra careful.

DaveAK
14 April 2011, 1558
Hi Dave. My multimeter does have a continuity tester. I did check the continuity of the loop by connecting the tester across the wires and it beeped.
Hmmm. Well that's got me stumped then. All you did by putting in the jumper wire was mimic a closed loop. And by testing the yellow wires you confirmed that you do have a closed loop. Therefore it should have worked. When you've got it halfway charged I'd try removing the jumper and replacing the cell loop wires and see if it's working again. If it doesn't then I'd definitely suggest contacting Dimitri over at DIYElectriccar.com for advice.

Richard230
14 April 2011, 1559
Another question on the braking/controller issue. Can you hook up the programmer and tell us the Digtal IO setting under Traction Setup. When you hook up the programmer with the bike on you should get the screen that says Traction OK. Push the down arrow once, then push the right arrow until you get to the Traction Setup menu. It might skip a few options. When you get to the Traction Setup menu push the down arrow until you see Digital IO. It should be the second option. What number does it give you?

Dave, this is what the programmer said: 1.6.1 Trc Setup Standalone.

I need to do some chores, but I'll be back in a couple of hours.

DaveAK
14 April 2011, 1602
Thanks Richard. I really need the next item 1.6.2 Digital IO. Should just be able to push the down arrow again to get there.

magicsmoke
14 April 2011, 1603
Totally agree Dave, definitely the loop. One slight gotcha with a continuity test is that I'm not sure what the 'on resistance' is for each of the loop 's opto switches. 24 in series may appear as an 'open' to some meters. dunno?
If so, another approach would be to reconnect both yellow wires to their original positions. You should now measure 12V ish across the yellow pair WHEN THE FAULT IS PRESENT. Keep your negative probe connected to one of the head end terminals (screw it in with the yellow wire if it'll fit) and progressively move around the loop with your positive probe measuring at each of the push on terminals, preferably the male connectors so as not to disturb any intermittency and so disguise the fault. Be careful not to slip and short the loop to any adjacent cell terminal.
When your meter reads 0V, then the terminal you're on or the one just before is the problem area.

edit : another issue with using the continuity function is that ironically, if the loop is faulty i.e. open circuit, then in this particular case, the head end is supplying 12V to the terminals which again might (probably would) make the test beeper beep.

Richard230
14 April 2011, 1756
Thanks Richard. I really need the next item 1.6.2 Digital IO. Should just be able to push the down arrow again to get there.

I don't get another screen after the 1.6.1 Trc Setup Standalone. When I push the down arrow button again, nothing happens.

I figure that my pack is about 90% charged. All of the batteries seem OK and I see nothing over 3.40 volts. The new batteries are showing 3.35 volts. No red LEDs are lit so far. I am charging again in the hopes of finding an over charging cell. Based upon my previous experience, I would expect to see the red light at 3.45 volts.

Fab Man, I agree that the wiring is not very pretty. Unfortunately, I have no wiring experience and am happy when something just works. I feel a lot differently about mechanical parts, but I just don't have the proper skills for wiring tasks. My son-in-law tried to show me how to solder but I was pretty messy and got solder all over the place. I had a real problem getting it off of the iron on to the wires. All I did on this job was to attach the BMS boards and tighten down the cell bolts. Most of the actual wiring was performed by a friend who builds and maintains commercial radio stations.

DaveAK
14 April 2011, 1825
Maybe after you've pushed the down arrow you need to push the right arrow to get to 1.6.2. There's a lot more settings. I just didn't have the calibrator in my hand to remember the exact steps.

Richard230
14 April 2011, 1838
I found it. It reads Digital I0 8.

I just finished checking all of my cells after two red lights came on. Two cells were at 3.33 V and the rest were at 3.32 or 3.31 V. So it looks to me like my batteries are fine.

In order to get to the 1.6.2 screen, I scrolled though a whole lot of stuff. I particularly liked that the "seat and pump" are on - whatever that means.

DaveAK
14 April 2011, 1926
Digital IO set to 8 for a standalone controller is the same as I have. I also think it's wired the same as I would have expected, but from your picture it's not clear if pins 6 & 12 are actually empty. If they are then it is as expected and I retract my earlier statement.

Seat & pump set to On is the default setting, but mine is set to off. I don't know what effect the difference makes, and I wouldn't suggest changing anything as it was working before.

If we can get the BMS working so that you can ride and charge without issue provided the brake switch is left disconnected then I think we're half way to solving your issues. You need to double check every connection in the cell loop. When we've got the BMS working reliably I'm not sure what to do about the brake issue remotely. That's going to be a much tougher one I think.

Please consider getting a buzzer for the LVC terminals, it really will help us troubleshoot for you.

Richard230
14 April 2011, 1953
Hi Dave. I plan to buy a buzzer this weekend from Radio Shack, but I don't plan on installing it until I can get the BMS under control so that it won't be buzzing all the time.

That seat and pump thing bothers me. My friend claims that the controller has a setting that causes it to shut off power if it stops getting a signal from somewhere. In a forklift, it would be from a switch under the seat. If the driver gets off the seat, the system shuts down until the driver gets back on the seat again. It might be that that feature should be off, but I agree that messing with the controller to solve the stalling issue when braking should be left for later. What I should probably do soon is to scroll through all of the programmer screens and make a note of each setting and see if anyone has any comments about things that should be changed.

DaveAK
14 April 2011, 2013
If you can make a list of all the settings I can certainly offer my opinion on them. Although there are several switch inputs, like the seat and pump one, there are also certain ways they get interpreted dependent on other settings. I'll certainly have a go at trying to sort out your setup and what it means. Nothing I've read anywhere indicates a three minute shutout period though. Something like that would typically be for overheating I would have thought.

Do you have access to the Fault Log, 1.5 on your menus?

DaveAK
14 April 2011, 2016
While you're testing the BMS you can hook up a 12V bulb to the BMS instead of the buzzer.

Richard230
15 April 2011, 0804
I just checked my bike this morning. All of the batteries are balanced out to 3.30 or 3.29 volts. All of the BMS boards are green. However, the buzzer circuit is showing 12.5 volts, so it is still showing a warning and the bike will not charge without jumping the yellow wires.

My best guess is that I damaged the head-end board when replacing the batteries the other day. When I removed the batteries I just removed the entire BMS boards as an assembly, without disconnecting them and replaced them the same way. Although I did not fry an individual board (which I have done before and I know when that happens) when connecting the power cable back up, perhaps I managed to send a voltage spike into the board that fried a component. Unless anyone has any better ideas, it would seem like I should just buy a new head-end board to replace the current one. Fortunately, my selection of the individual BMS board system was a good one for this replacement, as I seem to recall that the head-end board only costs $35 - just half the price of a new battery.

I'll try riding around again today and see if there is any way that I can keep the regen off to prevent stalling, while retaining my stop-light feature. However, I still suspect that changing the controller's programming somehow will solve the stalling issue.

My friend, who played with the programmer quite a bit (I kind of wonder if he made any changes that he forgot to tell me about - it is easy to push the plus or minus buttons by mistake) claims that there is a setting in the controller where you can select the number of seconds that the controller will cut power if there is an interruption in traction power. My son-in-law is a programmer/developer by trade and I will be seeing him soon. Maybe I can get him to read the Sevcon operating/programming manual and figure out what the various programs can do. Reading and understanding that 60-page manual is not all that easy, though.

I kind of wonder what I am going to do with all of this education that I have obtained at the School of Hard Knocks in the future?

DaveAK
15 April 2011, 0925
If it charges with the jumper wire and not the yellow cell loop wire I'd put money on the fact that your head end board is fine, and it's the cell loop that's the problem. This is the one part of the miniBMS design that I really don't like. The simplicity of a NC loop breaking and triggering an event is great, but the reality is, at least on my bike, that all those individual crimped connections are highly dubious and create multiple points of failure. I was quite surprised when it worked first time, but I'm fully expecting one or more connections to eventually fail.

My suggestion before trying another board is to check and double check every connection between the individual cell boards.

As for the controller - there are various settings for delays triggered by different events, but I didn't see anything that was in the order of 3 minutes. I'll check again though.

billmi
15 April 2011, 0935
Wow! That's a lot of progress in diagnosing the problems since yesterday afternoon. It is sounding a something related to regen is the culprit.

Dave - While that would work, it's important to note the amp limits (they are listed in the manual) for the buzzer circuit, and not use a bulb that exceeds that amp limit. A "12volt LED" (really an LED with a limiting resistor) would be a good choice here because their amp draw is next to nothing.

Richard - Perhaps I missed it, did you try disconnecting the alert loop from the head board, and just hooking up a single wire in its place, then checking the buzzer output for voltage (as well as the charger relay status)? That would be the sure diagnosis as to whether it's the head board. You're absolutely right about the head board being an inexpensive (relatively) component, but it's also pretty hard to fry it by doing something wrong while taking batteries in and out, since the cell level boards don't send any power to the head board, they just allow power to flow through the loop, or break the connection on an alert.


I kind of wonder what I am going to do with all of this education that I have obtained at the School of Hard Knocks in the future?

You'll be ready to help folks with problems on elmoto :)

DaveAK
15 April 2011, 1038
Dave - While that would work, it's important to note the amp limits (they are listed in the manual) for the buzzer circuit, and not use a bulb that exceeds that amp limit. A "12volt LED" (really an LED with a limiting resistor) would be a good choice here because their amp draw is next to nothing.
Absolutely correct. I didn't consider that.


Richard - Perhaps I missed it, did you try disconnecting the alert loop from the head board, and just hooking up a single wire in its place, then checking the buzzer output for voltage (as well as the charger relay status)? That would be the sure diagnosis as to whether it's the head board. You're absolutely right about the head board being an inexpensive (relatively) component, but it's also pretty hard to fry it by doing something wrong while taking batteries in and out, since the cell level boards don't send any power to the head board, they just allow power to flow through the loop, or break the connection on an alert.
He has done exactly that. Used a jumper and got the charger charging. As far as I'm aware the LVC and HVC trigger the same output, i.e. the buzzer will sound when HVC is hit as well - at least it does on my slightly older design board.

Richard230
15 April 2011, 1454
I went out for a ride today. I connected the regen and the brake light and tried about 10 hard stops and the system did not shut down. Then while slowing down for a stop sign, the bike went dead. I immediately hooked up the programmer and it showed: Traction Line Cont. o/c According to the diagnostics section of the controller manual that means that the "Line Contactor did not close". So today's questions would be what is a "Line Contactor", what does it do, where might I find it and can it be repaired?

The power went off twice again on the way home, which was only a mile away. I rode 3 miles and 10 stops before it shut down and one mile, with 4 stops and two shut downs on the way home.

I skimmed the manual and also could find nothing about a 3-minute delay for the controller to reset itself. But it must have this feature, as it takes exactly 3 minutes with the ignition off before the traction power will come on again. I really need to ask my friend where he got this 270 second stuff, or if he just made it up based upon the time it took for the controller to reboot.

I just checked the buzzer circuit again. With the jumper wire in place and the charger relay red light on (but not charging), the buzzer circuit is off. I then removed the jumper wire and the buzzer circuit came on. I also removed both loop wires from the head-end board and the buzzer stayed on. I then checked the continuity of the loop again and my multimeter buzzed and displayed the number "502", whatever that means.

frodus
15 April 2011, 1527
sounds like your contactor might be disengaging..... the line contactor is the main "relay" that connects the pack to the controller. It makes a big CLICK when you turn the bike on.

HighlanderMWC
15 April 2011, 1536
When my contactor got corroded it would shut off relatively randomly, although it was worse on hard acceleration. Eventually I wasn't able to get it to engage at all.

DaveAK
15 April 2011, 1541
The Line Contactor is a big relay that sits in between your battery pack postive and the controller. Typically with the Sevcon it is opened and closed by the controller. Pin 8 of the Sevcon provides 24V to the contactor to close it. The contactor will have some heavy gauge and light gauge wires to it. Check the light gauge connections as these are the ones that open and close the contactor. Trace the light gauge wires to check for problems other than just at the contactor ends.

The buzzer circuit is behaving as expected. With the jumper in place the head board sees a closed loop and thinks everything is fine. As soon as you remove the jumper the loop opens, the head board thinks there's a problem and activates the buzzer. The yellow loop wires that you are connecting are not being seen as a closed loop by the head board. There's something wrong with the loop even if your tester is showing continuity. This is really stretching my limited electrical knowledge, but I wonder if the gauge of wire you're using is having an effect. Those yellow wires seem pretty substantial to the gauge I'm using. Hopefully someone else can chime in, because beyond actually going through the loop connection by comnnection, cell by cell, board by board I'm not sure what else can be done.

Richard230
15 April 2011, 1554
I wonder why it would only shut down when the regen is activated? That loud click is one thing that I always wait for before activating the kill switch to start off. Now the odd thing is that when the bike "stalls", I can still hear the loud click from the contactor when I turn the ignition switch back on. But the motor system will not activate (all I see is 0.1 amps, which I assume is the draw from the lights), until exactly three minutes have elapsed since I turned off the ignition switch. If I turn it back on before the three minutes are over, I can hear the contactor click, but it will not start again unless I wait another 3 minutes with the ignition off. It sure sounds like the controller is involved in the process and doing something. I can only surmise that the regen is kicking the contactor off randomly.

Perhaps the contactor is corroded as Highlander mentions. Can it be taken apart and cleaned? I was going to ask what it looks like and where it was located, then I turned on the ignition switch and instantly found the location from the loud click. It is located behind the right side panel, covered with a removable rubber flap and has two power cords attached. Photo attached.

DaveAK
15 April 2011, 1602
Sounds like the contactor is operating correctly.

I suppose that there's the possibility that the controller is faulty, but that's an expensive lump to replace so lets assume that it's OK. I need to read up on the regen settings to see if there's anything there that we can adjust. I'll do some research and then perhaps have you confirm some settings.

Richard230
15 April 2011, 1617
Thanks for your help, Dave. If nothing else works, I'll just defeat the regen circuit and try to bypass the stop light circuit that is probably connected to the regen relay. I think my son-in-law may be stopping by this weekend. He just loves to (try) fixing broken stuff and is a programmer by profession. It wouldn't bother me much to loose the regen if I knew the best way to do that. Perhaps there is a setting in the controller that will turn it off or ignore its signal.

I ordered a new MiniBMS head-end board, as I am better at replacing things than repairing them. Since the system was working OK before I replaced the batteries, including the signal loop, the most likely problem in my mind is the head-end board. In any case I will double check the connections on each BMS board.

DaveAK
15 April 2011, 1636
To disable regen in the controller set 1.1.6 Neutral Regen Current to zero. Although simply pulling the lead to the regen relay should disable it too.

Out of curiosity, what is it currently set at, and what is 1.1.44 High Voltage Cutback set at? These two settings are both tied to how regen works.

frodus
15 April 2011, 1643
Sounds like the contactor is operating correctly.

how'd you come to that conclusion?

If its sensing an intermittent connection (i.e. the connection either drops out, or it momentarily drops out), it'l throw a fault.

DaveAK
15 April 2011, 1654
how'd you come to that conclusion?

If its sensing an intermittent connection (i.e. the connection either drops out, or it momentarily drops out), it'l throw a fault.
Because that would be the controller, or the wiring, not the contactor itself. i.e. I wouldn't rush to replace the contactor or as Richard asks if it can be opened and cleaned.

Now it could be an intermittent wiring problem that only occurs when the bike is in motion, or something within the controller in the way of either a fault or an errant setting. If, like Richard says, it clicks when the ignition is switched on then it's doing what it's supposed to do and I would classify it as good. I'd still suggest checking all the wiring though.

I have little knowledge of the internals of a contactor and perhaps there is something within that is causing intermittent operation under motion, but from what Richard has described I'll stick with my conclusion that it "sounds like" the contactor is operating correctly. It's by far from a definitive statement.

HighlanderMWC
15 April 2011, 1729
Is it possible to hook up the contactor so that the controller is taken out of the equation?

DaveAK
15 April 2011, 1754
Is it possible to hook up the contactor so that the controller is taken out of the equation?
Theoretically yes, but it would take some messing about with the wiring, which I'm sure Richard doesn't want to do.

Another thing about contactors is that some of them require a certain voltage to close, but less voltage to remain closed. I'm not sure if this applies to this particular model, or if the Sevcon is the type of controller that reduces the line contactor voltage. That could be another possibility, but would likely have been an issue from day one.

Richard230
15 April 2011, 1803
I just checked 1.1.44, Hi Voltage Cutoff. It says 94.0V. I have never seen anything like that voltage on the CA. About the highest I have ever seen was 86V coming down a steep hill at high speed.

I think I will try to disable the regen in the controller this weekend and see what happens. That is a lot easier and safer than messing with big power wires, the controller or some other thing that I might damage. I'll report back what happens when I try Dave's suggestion and cut the regen at the controller.

Thanks everyone for your comments and suggestions.

DaveAK
15 April 2011, 1815
94 is good. Just wanted to make sure it hadn't been set to something ridiculously low. Let us know how it goes with the other setting.

jpanichella
15 April 2011, 1921
This is a really great thread, simply because it provides a sort of troubleshooting play by play which could help a lot of people down the road. Good luck Richard, I hope we get this all sorted.

Richard230
16 April 2011, 0801
This is a really great thread, simply because it provides a sort of troubleshooting play by play which could help a lot of people down the road. Good luck Richard, I hope we get this all sorted.

You and me both. Sorry this is taking so long to figure out.

However, I am really fortunate that I am retired and don't have to depend upon my electric motorcycle for commuting to work. I am also fortunate that I have five IC motorcycles as a back-up to the GPR-S, but I sure hate to use them for short around-town trips. When it is running, the GPR-S doesn't require any warm-up, is perfectly smooth, completely quiet and I really enjoy watching the gas prices at the local gas station going up each day - until I have to fill the tank of my BMW R1200R with premium fuel, that is. Then I do like everyone else at the station does and just close my eyes and avoid looking at the price read-out. I really feel sorry for independent truckers. I don't think they are going to have any options to replace diesel any time soon.

magicsmoke
16 April 2011, 0930
C'mon Richard, enough of this don't know much about electrics and being retired nonsense. Sort yourself out young man! :)
You're sooo close to figuring out the charging issue at least.

Please try this one more time before fitting the new headend board.

Disconnect both of the two yellow wires from the headend screw terminals and connect one of your continuity tester probes (doesn't matter which) to either one (not both) of the yellow wire ends.
Now, with your other tester probe, follow that same yellow wire around the pack and probe at the individual pcbs, ideally right on the soldered male part of the push on connectors so as not to disturb a faulty connection and disguise the fault.
When you get an 'open' or 'high resistance' or 'no beep' depending on how your tester works, then it is the previous section of the 'loop' that's at fault.

I hesitate to add this bit, 'cause I don't want to confuse the issue, but if the above test is inconclusive then repeat it exactly except with your meter set to resistance. Perhaps 2K or thereabouts. The reason is that each of the individual pcb 'switches' is really just a low resistance, not a short, and the successive addition of these resistances in series may add up to more than your continuity tester considers a short and will therefore falsely indicate an open.
So just repeat the test and confirm that the meter reading increments by an approximately equal amount as you progress around the loop.

Remember though that the fault might be in the first link you test!. Don't let it fool ya.

One more thing, some of the push ons look a little close to the battery terminals. Just ease them away if so.

Good luck.

Richard230
16 April 2011, 1547
Thanks Magicsmoke. I'll give that a try tomorrow. Right now I have the fairing on the bike and I am too tired to pull it off this afternoon. I thought my son-in-law was coming over to help me check things over, but he is working on some video game projects and bailed out.

Today I defeated the regen at the controller, as Dave suggested. Unfortunately, the bike died again at my second stop. I had my programmer with me and it showed the usual "contactor" fault.

I really did think that the regen was the problem. However, I have decided that I prefer riding without the regen as the braking just feels more natural, so I plan to just leave it defeated. Also, I noted that there was a lot less power fluctuation showing on the CA screen when I would come to a stop. With the regen activated, the CA would show 20 or more amps going into the pack and then just before coming to a stop, it would show 12 amps going out. Now I don't see any of that drama on the screen.

However, I did continue to ride down to the foggy beach and back and experienced no more power stalls, so I am not sure what is going on. You would think that if there is a short or something, the problem would occur upon every application of the brakes. This time it only occurred once in about 15 stops.

My CA suffered some sort of melt-down just now. I went to turn on the ignition to verify that the both brakes still activated the stop light, and the CA showed 71.4 volts. I just about had a heart attack, as I had just completed charging the pack and I was getting 80.1 volts when I checked it last. So I tried one more time and the screen showed 80.1 volts again. I checked it with my voltmeter and it read 79.6 volts. I think I need the Ghost Busters.

chef
16 April 2011, 1605
My CA suffered some sort of melt-down just now. I went to turn on the ignition to verify that the both brakes still activated the stop light, and the CA showed 71.4 volts. I just about had a heart attack, as I had just completed charging the pack and I was getting 80.1 volts when I checked it last. So I tried one more time and the screen showed 80.1 volts again. I checked it with my voltmeter and it read 79.6 volts. I think I need the Ghost Busters.
The CA is a reliable instrument from my experience. Perhaps the 71.4v reading is a clue that might help with your diagnosis. Although unlikely, I wonder if you have some cells being driven to 0v or you have a short that's bypassing a couple cells. If there is a short that cuts out cells in your pack, such a sudden drop in voltage might make the controller go into emergency cutoff mode.

Richard230
16 April 2011, 1815
That is the first thing that occurred to me. That momentary malfunction of the CA I am calling just a fluke right now. I never saw anything like that when riding, or at any other time in the past. I am not going to worry about it unless it happens again. In the meantime I will keep my fingers crossed about the CA and perform a tour of the MiniBMS board connections as Magicsmoke and others have recommended. However, if he could see how jammed against the frame and upper battery boxes the BMS boards are in the lower packs, he would be really concerned. If the shutoff occurred when going over a bump, I would be going right to those boards. However, I think I will check the battery boxes and make sure that they can not move forward when braking. I should also ride around more with the regen turned off to see if I can determine any sort of pattern, other than braking, that causes the contactor fault. I was surprised that I was able to travel 6 miles today with only one cutout. That is an improvement.

Fab man
16 April 2011, 1855
I agree with magic smoke. Take a break . Hit it again later, maybe with another set of eyes to help. You've come this far, I think your getting close to fixing the problem. I still have a concern about the proximity of the backs of the BMS circuit boards to the connecting bars. I see circuit board/connecting bar overlap in one of your other pics, too. It looks like when the battery post bolts were tightened, the BMS ring connectors twisted slightly-pulling the BMS boards close or on to the connecting bars. The intermittent nature of your problem could be something as insidious as the batteries shifting forward when you brake, causing a short.

If you or someone else can do it safely and carefully, gently pry back the boards from the bars and/or insert insulating tape between the two. Maybe this will solve your problem. If you pull off the BMS units, check for rubbing and/or arcing on the back near the bars.

Richard230
18 April 2011, 0819
It looks like this thread is not going to wind up anytime soon. I feel like I am sinking into quicksand.

So last night I checked out the GPR-S again. I turned on the ignition and the CA showed 67 volts!. I then turned it off and back on again and it said 71 volts (I am leaving off the tenths). I then used my voltmeter to check the pack voltage at the terminals next to the contactor and it showed 75 volts. Then I went around to every individual battery and got a reading of 3.30 or 3.29 volts. Unfortunately, I did discover a BMS board without a green light during the checking process, but that battery showed 3.30 volts. I have one more board left and I should replace that one (which is going to be a project, as it is located in my bottom pack under two other layers of batteries), but I can't see how that would cause my problems. Would that result in the continual BMS warning that the head-end board is receiving?

This morning I checked the bike again and the CA showed 79.6 volts, while my voltmeter was reading 75 volts again. I then dragged out my Extech multimeter (which costs 10 times what I paid for my pocket voltmeter) and it showed a pack voltage of 79.8. My old voltmeter is 25 years old and has the original batteries, could it be that the batteries need to be changed? The LCD screen is still sharp.

Besides the checking the voltage readings, I tried checking the BMS loop with my voltmeter's continuity setting, first by just sticking one probe into the BMS head-board junction with one of the wires and then probing the individual board contacts (just because it was easy to do - I was going to follow magicsmoke's recommendation to remove the wire from the board today). That was when I got a strange result. As I went around the boards on the batteries in the upper pack (the one where I replaced the batteries and removed and replaced the boards), I got a sound from each board's sensor wire connection, but when I tested the last contact at the right side of the pack, where the loop wire goes to another pack, the head-end board clicked and the charging relay came on. But when I moved the probe to the other side of that board, the head-end board clicked off. As I suspected, it appears that there is a problem with the BMS sensing system across the upper pack, but I don't know what to do about it. I also don't know if the "dead" BMS board would cause the problems that I am having. There is just too much weird stuff going on for me to grapple with, but I wouldn't be at all surprised to hear that all of my issues are due to wiring and assembly errors.

Today's plan is to follow magicsmoke's recommendation to remove one sensor wire from the head-end board, connect it to a continuity probe and use the other probe to check each BMS sensor board wiring post. Any other recommendations will be appreciated.

billmi
18 April 2011, 0934
The Cycle Analyst uses a voltage divider to step down the battery voltage to a range that it can measure. I don't have the manual in front of me, but if I remember correctly, it can be off by +/- 2% in normal operation. Being 4 volts off is a little over double that, but don't expect high-precision out of the CA's voltmeter.

Richard230
18 April 2011, 1359
Well, I used my continuity tester to test the BMS control cable wiring this morning. It didn't get past the "dead" BMS board, so I bypassed it with a jumper wire as a temporary measure while I wait for to see my friend who has my last extra MiniBMS board, which I left in his shop. As near as I can tell, the control wire system now has continuity around the battery pack, although there were several battery cells where I couldn't reach my probe to the soldered connections because of the very tight clearances on top of some of the batteries. I did straighten out the boards and wires on my upper pack - the one where I had replaced the batteries and made the mistake of using a power ratchet to re-tighten the battery bolts. That was what caused the wire and board twisting. However, the warning circuit is still energized, but maybe that is normal when you bypass a board. My pack voltage is holding at 79.6 volts, according to both the CA and my voltmeter.

Today's question is: should I try test riding and charging my bike with the dead BMS board bypassed, or should I wait until I can replace it?

frodus
18 April 2011, 1409
got a single cell charger? Maybe charge that one up before you go out, and ensure its not going to LVC.

And don't go for a full ride, just a mile or two to try to work out the regen/restart issue.

Richard230
18 April 2011, 1801
I'll give that a try tomorrow. I don't have a single cell charger, but I will take the bike for a short test ride and see what happens. Hopefully there will not be any more earthquakes and power outages tomorrow, like there were today. We got hit with an earthquake just two miles from my house and the power went out for the afternoon and just came back on.

chef
19 April 2011, 0930
Richard - you might want to pre-balance the cells instead of relying on the BMS to do it. The Mini BMS has limited shunting ability and wouldn't be able to handle large differences in cell SOC. As others have noted, you can't determine cell SOC on an LFP based on voltage alone. Two cells at 30% & 90% could both read 3.3v.

Richard230
19 April 2011, 1536
Richard - you might want to pre-balance the cells instead of relying on the BMS to do it. The Mini BMS has limited shunting ability and wouldn't be able to handle large differences in cell SOC. As others have noted, you can't determine cell SOC on an LFP based on voltage alone. Two cells at 30% & 90% could both read 3.3v.

I wasn't aware of that, chef. However, I am sure you are correct and I am just as sure that my cells are out of balance due to their age and hard life. The four new cells that I installed recently seem to be much more stable than the rest of the cells in my pack. When charging, most of my batteries will fairly rapidly rise to 3.45 volts, while the new cells will stay around 3.32 volts during the charging process. It is clear that I need new batteries, but I am not going to invest in new cells until I can get my problem with the bike stalling for three minutes when coming to a stop, occasionally. Plus, the BMS still will not allow the charger to operate unless I jump the board and the warning circuit continues to be active even when all of my batteries are within a couple of hundredths of a volt. (When my batteries hit a 3.45 volts and I shut down the charging system, they rapidly balance out to 3.30 volts.) So the problem with the BMS also needs to be sorted out and repaired. No use getting new batteries when the bike has these issues.

So today I took frodus's advice and rode my bike for two miles. It stalled once on the third stop and after 1/4 mile. I had the regen disconnected so I can't blame that for causing the problem. When I left my pack voltage was showing 79.6 volts and when I returned it showed 77.5 volts and then hit 78.4 volts after resting for a minute. I had used just over 3 Ah and after 30 minutes of charging, the BMS board red lights starting coming on all over the place, pretty much all at once and I pulled the plug.

I have pretty much run out of ideas or things that I am qualified to check. Unless anyone else has any other suggestions for me to try, it would seem that what is needed is to take all of the batteries out of the bike, check the installation of the BMS boards, the signal wiring, the power cord wiring and put everything back together slowly. When the MiniBMS boards were installed there were three of us working on it and everyone was in a big rush, as it was getting late and we all wanted to go home. Even I know what can happen under those conditions.

My next plan is to try to get my son-in-law to take the bike over to his house and let him perform the reconstruction work. He sold the Aprilia chassis that I gave him and now he has some room in the garage for my GPR-S.

frodus
19 April 2011, 1554
please stop, get a single cell charger immediately, and balance charge Every Single Battery. Maybe get 2, because it might take a while.

If you start with an unbalanced pack and just throw the batteries in there as they arrived, you'll have problems.

Richard230
19 April 2011, 1753
That sounds like a good idea. Where can I buy a single-cell charger? I have a couple of 12V motorcycle smart chargers and I thought that maybe they could charge up four cells at once, but I have never tried it.

chef
19 April 2011, 2013
Ed had some Thundersky single-cell chargers. They're packaged as regular looking wall warts but with alligator clips on the connector end. The end voltage probably is too high though (3.65v).
You might consider investing in a variable DC power supply. They come in very handy for different cell chemistries and general electronics use. This one supports 1.5 - 15v @ 2A and costs $40 (http://www.amazon.com/Tekpower-Variable-Supply-1-5-15-HY152A/dp/B000RO8J98). I've got an ancient variable DC supply which I've used to [very slowly] pre-balance cells. It works, just gotta be patient since it's relatively low current.

jpanichella
19 April 2011, 2029
Could all the cells be top balanced at once by paralleling the whole pack and using a single cell charger?

chef
19 April 2011, 2044
I've read instructions by one of the manufacturers that doesn't recommend paralleling batteries. In the case of LFPs it wouldn't work unless they're at a knee of the voltage curve, in which case they're already balanced. Same reason as above: LFPs have a long flat part of the voltage curve. With no potential difference, no current would flow between cells at 30% and 90% SOC for example.

Edit: Sorry I misread the post and thought it was about placing them in parallel without a charger and letting them self-balance. Charging in a massive parallel configuration should work.

podolefsky
19 April 2011, 2101
I've read instructions by one of the manufacturers that doesn't recommend paralleling batteries. In the case of LFPs it wouldn't work unless they're at a knee of the voltage curve, in which case they're already balanced. Same reason as above: LFPs have a long flat part of the voltage curve. With no potential difference, no current would flow between cells at 30% and 90% SOC for example.

Is this specifically in regard to charging, or just not paralleling in general? I think that is due to concerns about parallel cells if one of them fails to short. But it's done all the time (discussed in another thread, consensus was that it's OK since failures are very rare).

[edit] I removed what I just said - I'm actually not sure what will happen if the cells have very different capacities, but I think it's like this: Normally, with cells in series, they all see the same current and so the low capacity ones fill up first. But if they're in parallel, they'll see different currents as they charge, so even if the low capacity ones fill up, the large capacity ones might keep going. Then when you put them back in series, you'll kill the low capacity ones too early.

It might also work kinda funny because the charger is expecting to see the load of a single cell, but it will instead see the load of several cells. Not sure if this will mess up when the charger knows to stop charging.

jpanichella
20 April 2011, 0541
Requoting from Frodus at DIYElectricCar

"get a single cell charger, parallel them all, and fully charge them all. Use copper wire to connect all of the + terminals together and connect to the + of a single cell charger for lifepo4. Then connect the - terminals together and connect that to the - of a single cell charger.

This will slowly charge them to the same level and insure they're equalized. Then you can take them out of parallel, and wire as your project requires (I'm assuming series). This will make sure they're all the same voltage and on the top end of the charge, they'll be balanced."

Richard230
20 April 2011, 0737
Requoting from Frodus at DIYElectricCar

"get a single cell charger, parallel them all, and fully charge them all. Use copper wire to connect all of the + terminals together and connect to the + of a single cell charger for lifepo4. Then connect the - terminals together and connect that to the - of a single cell charger.

This will slowly charge them to the same level and insure they're equalized. Then you can take them out of parallel, and wire as your project requires (I'm assuming series). This will make sure they're all the same voltage and on the top end of the charge, they'll be balanced."

Is this a procedure that could be performed with a group of random-condition cells like I have, or is this type of balancing/conditioning more useful with a set of new batteries? And if it can be used for batteries of different ages and capacities, would you discharge them to a certain level first?

Chef, thanks for that tip on the charger than can be adjusted for different battery chemistries. That looks like a very useful device. I will have to get one.

frodus
20 April 2011, 0815
That sounds like a good idea. Where can I buy a single-cell charger? I have a couple of 12V motorcycle smart chargers and I thought that maybe they could charge up four cells at once, but I have never tried it.

get a single cell from here:
http://www.batteryspace.com/Smart-Charger-6.0A-for-3.2V-1cells-LiFePO4-Battery-Pack-100-240VAC-CE.aspx

frodus
20 April 2011, 0816
Could all the cells be top balanced at once by paralleling the whole pack and using a single cell charger?

Sure, that's the best, to parallel and also charged fully. But I was trying to keep him from having to rip all the batteries out of the bike.

frodus
20 April 2011, 0819
It might also work kinda funny because the charger is expecting to see the load of a single cell, but it will instead see the load of several cells. Not sure if this will mess up when the charger knows to stop charging.

They'll just current limit to the limit of the charger.... i.e. if its a 6A charger, it'l put out 6A, and thats what the batteries get, no matter how much they want. The charger I lined to is smart enough to go CC/CV, where CC is up to and including 6A.

Richard230
20 April 2011, 0821
Thanks for the tip, Frodus. However, I had just ordered the variable-voltage charger that chef recommended. It seemed that it would be useful for use when charging other cell types that might come along in the future. I got the last one in stock, it appears.

frodus
20 April 2011, 0822
Is this a procedure that could be performed with a group of random-condition cells like I have, or is this type of balancing/conditioning more useful with a set of new batteries? And if it can be used for batteries of different ages and capacities, would you discharge them to a certain level first?

Chef, thanks for that tip on the charger than can be adjusted for different battery chemistries. That looks like a very useful device. I will have to get one.

Doesn't really matter. They'll all be fully charged to the same level, and since they're parallel, the current goes to the ones that have "lower" charge. I'd still single cell charge them first, then parallel them, because if one is really low, and one is fully charged, you will get a spark and things may get hot. There is nothing to "control" current when you wire batteries in parallel. Its best to start as close to the same charge as possible.

Now, you could just charge each one fully, one at a time, and it'l do pretty much the same thing. I was trying to keep you from tearing the batteries out. If its fairly easy to do, then its not a bad idea.



Thanks for the requote jpanichella!

Richard230
20 April 2011, 0827
I learned that lesson about attaching parallel batteries that are not all at the same voltage yesteday. I was planning to hook up four of my old batteries in series and try charging them with a 12v charger just to see what would happen. I got the polarity mixed up and connected the first two in parallel. I got a big spark, quickly disconnected them and decided not to play with the batteries any more.