What does this charger behavior indicate?

General Discussion about electric vehicles.
Post Reply
User avatar
Chalo   100 GW

100 GW
Posts: 7961
Joined: Apr 29 2009 11:29pm
Location: Austin, Texas

What does this charger behavior indicate?

Post by Chalo » Mar 02 2019 5:18pm

Last night, my cheap generic 48V 4A lithium ion charger started displaying a behavior I haven't observed before. After a couple days of using the bike and running through perhaps half of its battery capacity, I plugged in the charger and it began charging normally, with the red charge indicator LED lit and fan running. It did that for a couple of hours, as per normal, before going quiet. When I went to unplug the charger, I found that the green indicator LED wasn't steadily lit like it normally is, but blinking briefly, over and over. I assumed that the charger had failed in some way, so I got my brand new spare charger and plugged it in-- but it started doing the same blinking green light routine.

It reminded me of the hiccup mode on an old SLA charger, where it oscillates between charge and float modes as the battery approaches maximum voltage. But in this case the red charge indicator wasn't involved at all, and the green light made it look like the charger was restarting over and over.

The fact that both my chargers, one of them brand new, exhibited the same anomalous behavior at the same time makes me think the BMS must be involved. Does anybody have an idea what might be the matter? Thanks.
This is to express my gratitude to Justin of Grin Technologies for his extraordinary measures to save this forum for the benefit of all.

User avatar
amberwolf   100 GW

100 GW
Posts: 30765
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: What does this charger behavior indicate?

Post by amberwolf » Mar 02 2019 5:33pm

If the BMS stopped battery charging before it was near full, and then began pulsing the charge FETs on and off, it could possibly do what you see. You can test for that by checking the FET gate signals.

Some time ago, I saw some odd behavior (that this thread reminds me of, but might not be the same as) with a BMS on an old Luna mini pack, that actually caused Power Faults on a Satiator charger, with some repetitive fast-cycling behavior beforehand. It would also stop charging before the pack was fulll, and required draining the pack down some before it would restart charging. I never found any cell problems that could cause it, so I assume it's a BMS fault. Pack works, just randomly does this sort of thing.

markz   100 GW

100 GW
Posts: 10088
Joined: Jan 09 2014 11:38pm
Location: Alberta Canada

Re: What does this charger behavior indicate?

Post by markz » Mar 02 2019 7:13pm

Time to open up that battery and have a peek at 'errrrr, usual visual checks all around then take the digital multi meter to 'errrrr and start probing 'errrrr. I would place the probes on the main output of the BMS, plug it in to the charger and check 'errrrr out. Then if you can open 'errrrr up even more and start probing 'errrrr right down to each individual cell, and visuals.

Nothing out of the ordinary lately, like a crash while riding, bicycle fell while it was standing up, moisture ingress, riding in the rain/river.

I doubt the chargers are at fault, hopefully nothing went wrong pluggin in the new one. BMS all the way mate!

User avatar
Matador   10 kW

10 kW
Posts: 918
Joined: Jun 29 2016 10:00pm

Re: What does this charger behavior indicate?

Post by Matador » Mar 02 2019 7:44pm

Chalo wrote:
Mar 02 2019 5:18pm
Last night, my cheap generic 48V 4A lithium ion charger started displaying a behavior I haven't observed before. After a couple days of using the bike and running through perhaps half of its battery capacity, I plugged in the charger and it began charging normally, with the red charge indicator LED lit and fan running. It did that for a couple of hours, as per normal, before going quiet. When I went to unplug the charger, I found that the green indicator LED wasn't steadily lit like it normally is, but blinking briefly, over and over. I assumed that the charger had failed in some way, so I got my brand new spare charger and plugged it in-- but it started doing the same blinking green light routine.

It reminded me of the hiccup mode on an old SLA charger, where it oscillates between charge and float modes as the battery approaches maximum voltage. But in this case the red charge indicator wasn't involved at all, and the green light made it look like the charger was restarting over and over.

The fact that both my chargers, one of them brand new, exhibited the same anomalous behavior at the same time makes me think the BMS must be involved. Does anybody have an idea what might be the matter? Thanks.
Had the same thing happen to me on the last charge i put on my eskate battery before on cell group died. Your charger is fine... The blinking is just a reflect of the BMS trying to do more work tan usual, which means your battery could had a cell group that is out of whack.check individual cell group voltages... Before you kill your battery like i did.

User avatar
Chalo   100 GW

100 GW
Posts: 7961
Joined: Apr 29 2009 11:29pm
Location: Austin, Texas

Re: What does this charger behavior indicate?

Post by Chalo » Mar 02 2019 9:12pm

Thank y'all for the insights.

There's an update: After the charger anomalies last night, I unplugged the bike. This afternoon, I rode it to the bike shop and it performed normally. On returning, I plugged in the charger and it began charging normally. I probably will have to roll out for tonight's band gig before I have an answer to the question of whether it's going to repeat last night's weirdness. But tomorrow after rehearsal, I'll go to the machine shop and start probing the battery for faults.
This is to express my gratitude to Justin of Grin Technologies for his extraordinary measures to save this forum for the benefit of all.

User avatar
amberwolf   100 GW

100 GW
Posts: 30765
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: What does this charger behavior indicate?

Post by amberwolf » Mar 03 2019 2:37am

Chalo wrote:
Mar 02 2019 9:12pm
This afternoon, I rode it to the bike shop and it performed normally. On returning, I plugged in the charger and it began charging normally.
That's essentially the behavior I had on the BMS of the mini pack (whose cells are fine).

User avatar
emmetbrown   10 W

10 W
Posts: 99
Joined: Oct 28 2010 1:38pm
Location: Germany

Re: What does this charger behavior indicate?

Post by emmetbrown » Mar 04 2019 2:09am

This is the norm behavior of expensive bike packs and chargers if the bms has found
a problem in cell balancing, so perhaps that is what you should investigate.


Emmet

Voltron   10 MW

10 MW
Posts: 2237
Joined: May 02 2013 4:53pm
Location: Santa Barbara CA

Re: What does this charger behavior indicate?

Post by Voltron » Mar 04 2019 10:47pm

On one of mine that has a cheap BMS, it doesn't seem to actually balance. Instead, when a group hits 4.18, but the total voltage is too low, it goes into that cutting the charge and flashing the green light mode. Maybe it was balancing, but so imperceptibly slowly it would have taken a week to even out.

Voltron   10 MW

10 MW
Posts: 2237
Joined: May 02 2013 4:53pm
Location: Santa Barbara CA

Re: What does this charger behavior indicate?

Post by Voltron » Mar 04 2019 11:53pm

I ended up manually balancing the battery and the charger went back to operating normally.

User avatar
Chalo   100 GW

100 GW
Posts: 7961
Joined: Apr 29 2009 11:29pm
Location: Austin, Texas

Re: What does this charger behavior indicate?

Post by Chalo » Mar 05 2019 12:18am

A further update: this evening, and yesterday evening, the charge cycle completed in the usual way, with a steady green indicator. I'm still going to investigate the battery in greater depth, but this is an encouraging development.

The BMS board I used is supposed to balance the cells at the top of charge with up to 60mA per cell discharge. Now I wonder whether it's doing that consistently.
This is to express my gratitude to Justin of Grin Technologies for his extraordinary measures to save this forum for the benefit of all.

User avatar
Matador   10 kW

10 kW
Posts: 918
Joined: Jun 29 2016 10:00pm

Re: What does this charger behavior indicate?

Post by Matador » Mar 05 2019 8:33pm

Chalo wrote:
Mar 05 2019 12:18am
A further update: this evening, and yesterday evening, the charge cycle completed in the usual way, with a steady green indicator. I'm still going to investigate the battery in greater depth, but this is an encouraging development.

The BMS board I used is supposed to balance the cells at the top of charge with up to 60mA per cell discharge. Now I wonder whether it's doing that consistently.
Hopedully yes. In my case one cell was on its way out. When my pack died, the dcir of on cell was way yoo high. It would charge to almost 4,2, but as soon as i would pit a load, that particular cell would drop nellow 2V. All other cells were fine, but i think the flashing red green could be bms balancing related... the weakest cell hitting 4,2v the first..

User avatar
Chalo   100 GW

100 GW
Posts: 7961
Joined: Apr 29 2009 11:29pm
Location: Austin, Texas

Re: What does this charger behavior indicate?

Post by Chalo » Mar 05 2019 10:00pm

I've been too ambushed with work (feh) to dig into my battery. I did notice that the charger (after having gone to its happy solid green state last night) resumed blinking green/off when I turned on the switched outlet it was plugged into. Boo.

It's a 13S 1P pack of 25Ah cells, so if it has a bunk cell, I can simply replace it. I'm half hoping it's the BMS. That's easy to replace, and an opportunity to add programmability.
This is to express my gratitude to Justin of Grin Technologies for his extraordinary measures to save this forum for the benefit of all.

John in CR   100 GW

100 GW
Posts: 14392
Joined: May 20 2008 12:58am
Location: Paradise

Re: What does this charger behavior indicate?

Post by John in CR » Mar 07 2019 7:09pm

The only chargers I've had that behaved similarly were LiFePo4 chargers. Those have a tendency to greater imbalance at the top, so the chargers are built to kick back on after the BMS burns off some surface charge of the high cells. 3.6V and 3.7V nominal cells don't have the same behavior, so the chargers generally just go green and turn off at the top with no restart.

User avatar
Chalo   100 GW

100 GW
Posts: 7961
Joined: Apr 29 2009 11:29pm
Location: Austin, Texas

Re: What does this charger behavior indicate?

Post by Chalo » Mar 23 2019 1:00pm

Here's an update:

I probed my battery for cell voltage and found them all very close to each other. There were no apparent problems with interconnects or plugs.

My charger went back to normal behavior for a while (steady red indicator with fan when charging, steady green/no fan when full/balancing).

A few days ago, I found my charger dark and quiet and cold as if dead, even though I had plugged in the battery just a couple of hours earlier. I unplugged it and resumed my work, then found that it went back to charging when I tried it again at the end of my work day. Hmm.

I plugged it in again later that evening and it began charging normally, but then later I found it manifesting a previously unseen malfunction: It was cycling on and off with a frequency of about one on/off per second, alternating between red indicator + fan and no indicator/no fan. I switched to my other charger and it did the same. I measured the battery voltage, which was about 54.1V out of a fully charged 54.6V.

The next day, I rode to work, taking care to use lots of throttle and put the battery under load to help manifest any imbalances. It sat for a few hours while I worked, and then I opened the battery again and checked cell voltages. They all measured between 4.076V and 4.081V.

The battery shows no signs of problems. Both the chargers misbehave from time to time, but both in the same way when I check them against each other. So I guess that means the problem is likely related to the BMS board. I have installing a new programmable BMS on my to-do list now.

What I don't understand is how something that happens in the BMS could result in my chargers turning themselves off, or on and off cyclically? It's like they're invoking short-circuit protection or something else that shouldn't be applicable in this case.
This is to express my gratitude to Justin of Grin Technologies for his extraordinary measures to save this forum for the benefit of all.

User avatar
amberwolf   100 GW

100 GW
Posts: 30765
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: What does this charger behavior indicate?

Post by amberwolf » Mar 23 2019 11:30pm

I had that behavior with the Satiator trying to charge that Luna pack, and AFAICR it's the same thing that would sometimes happen to the pack's original owner with the charger that came with it. It could cause a power fault error in the Satiator, too.

When I hooked the Satiator to the discharge port instead of the charge port, that didnt' happen, it would charge normally (though of course no cutoff during balancing), and similarly if i hooked both ports together, it would not happen.

I never got to checking with an oscilloscope, but based on the wattmeter I stuck between the charger and BMS charger port, current was cycling on and off at the rate that the Satiator screen was, so the BMS was, apparently, turning the charge port on and off and on and off.

*Why* it was doing that I don't know, but I suspect an intermittent problem with either the actual BMS electronics (or PCB), or an intermittent problem with a sense wire to a cell, if the BMS is designed to verify the cells are actually connected (many don't care, and won't give any error and instead will kill cells by overcharge or overdischarge).


Either way, since the chargers (good ones anyway) should turn off if they detect current dropping below a certain level (in case BMS doesn't shut off, or pack isn't protected by one), then when the BMS shuts off charge input port the charger should also shut off. When the BMS turns the port back on, the charger should detect a drop in voltage on it's output, and turn itself back on. Etc.

User avatar
tomjasz   10 GW

10 GW
Posts: 4234
Joined: Mar 29 2014 1:45pm
Location: Out riding, MN USA
Contact:

Re: What does this charger behavior indicate?

Post by tomjasz » Mar 24 2019 2:17am

I apologize for the unedited response, but iPad and I aren’t getting along.
If you have one of the that Luna sold

Chalo wrote:
Mar 23 2019 1:00pm
Here's an update:

I probed my battery for cell voltage and found them all very close to each other. There were no apparent problems with interconnects or plugs.

My charger went back to normal behavior for a while (steady red indicator with fan when charging, steady green/no fan when full/balancing).

A few days ago, I found my charger dark and quiet and cold as if dead, even though I had plugged in the battery just a couple of hours earlier. I unplugged it and resumed my work, then found that it went back to charging when I tried it again at the end of my work day. Hmm.

I plugged it in again later that evening and it began charging normally, but then later I found it manifesting a previously unseen malfunction: It was cycling on and off with a frequency of about one on/off per second, alternating between red indicator + fan and no indicator/no fan. I switched to my other charger and it did the same. I measured the battery voltage, which was about 54.1V out of a fully charged 54.6V.

The next day, I rode to work, taking care to use lots of throttle and put the battery under load to help manifest any imbalances. It sat for a few hours while I worked, and then I opened the battery again and checked cell voltages. They all measured between 4.076V and 4.081V.

The battery shows no signs of problems. Both the chargers misbehave from time to time, but both in the same way when I check them against each other. So I guess that means the problem is likely related to the BMS board. I have installing a new programmable BMS on my to-do list now.

What I don't understand is how something that happens in the BMS could result in my chargers turning themselves off, or on and off cyclically? It's like they're invoking short-circuit protection or something else that shouldn't be applicable in this case.
A $12 BMS, retail price, used in those packs have a 4% or higher fail rate. Always spec a spare duplicate BMS. IMO & IME
Thanks Justin_le we're here thanks to you. All the best to the mods for their tireless work keeping it on an even keel.

User avatar
Chalo   100 GW

100 GW
Posts: 7961
Joined: Apr 29 2009 11:29pm
Location: Austin, Texas

Re: What does this charger behavior indicate?

Post by Chalo » Mar 24 2019 2:43am

I built my own pack from automotive cells. There are few physical points of failure, but there is a relatively inexpensive BMS card that I didn't make.

1 in 25 fault rate for a component that's made in at least a semi-automated manner would be anomalous in my world. The thing I installed was all SMD, and appeared to be wave soldered. It would take both a sloppy process and nonexistent inspection/QA to achieve such a gross failure rate.

When I find a suitable replacement, I'll replace it and see how it goes.

I'm still at a loss for what makes a charger that has two normal states (red/green) to exhibit at least three abnormal states (flashing green/shutdown/flashing red) when there's nothing observably wrong with the battery. What could the BMS be doing to induce this weirdness? If it were only mediating the connection to the battery, the charger would either be displaying steady red (sub 54.X volts) or steady green (above 54.X volts or open circuit). But that's not what I'm getting.

The battery is a 1P series of high current cells, 25Ah nominal. The charger is nominally 4A. Is there any likelihood that some "bottomless pit" quality of the cells could be spoofing the modest capacity charger into a fault state?
This is to express my gratitude to Justin of Grin Technologies for his extraordinary measures to save this forum for the benefit of all.

User avatar
amberwolf   100 GW

100 GW
Posts: 30765
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: What does this charger behavior indicate?

Post by amberwolf » Mar 24 2019 2:45am

Unlikely. But if the current switching was happening at a high enough rate at the BMS port, perhaps it would cause the charger to display unintended and unexpected behavior?

It's part of why I wanted to scope the BMS port; especially the gate controls, and the cell sense wires.

markz   100 GW

100 GW
Posts: 10088
Joined: Jan 09 2014 11:38pm
Location: Alberta Canada

Re: What does this charger behavior indicate?

Post by markz » Mar 26 2019 11:02pm

I am shocked :shock: that Chalo got stumped. No one can know everything, except only the elite gurus of E.S. even then they can't know it all, thats why you gotta surround yourself with smart people, that is why I come to Endless Sphere.

User avatar
Chalo   100 GW

100 GW
Posts: 7961
Joined: Apr 29 2009 11:29pm
Location: Austin, Texas

Re: What does this charger behavior indicate?

Post by Chalo » Apr 07 2019 2:59pm

After more anomalous charger behavior, last night I opened up my battery to replace the BMS with an identical spare. On probing the cell voltages, I found two adjacent cells that were 70mV higher than the others (others were within about 5mV).

On closer examination, the balance lead common to that cell pair was broken. I soldered it back to its tab. By that point, I had already desoldered the big power leads from the old BMS, so I replaced it with a new one anyway. My hope is that the problem is now fixed (though I haven't recharged the battery since last night). I guess I'll find out soon enough.

I'm curious why a BMS with such a major fault state would still allow charging and discharging?

And I'm still on the lookout for a programmable BMS with Bluetooth cell monitoring. I need 13S, minimum 35A capacity but I wouldn't mind more. I appreciate any suggestions.
This is to express my gratitude to Justin of Grin Technologies for his extraordinary measures to save this forum for the benefit of all.

User avatar
amberwolf   100 GW

100 GW
Posts: 30765
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: What does this charger behavior indicate?

Post by amberwolf » Apr 07 2019 10:26pm

Chalo wrote:
Apr 07 2019 2:59pm
I'm curious why a BMS with such a major fault state would still allow charging and discharging?
None of the few BMSs I've had direct personal experience with, nor most of the ones observed in pack-problem threads, appear to have any way to detect that a balance lead is disconnected, and apparently are designed in a way that lets any disconnected lead "float" to a voltage that doesn't activate HVC *or* LVC, so the user has no way to know cells are no longer being monitored by the BMS.

Thus, we get packs with severely overcharged or overdischarged cells (groups), as the above problem allows growing imbalance over time, with no notice to the user. :(

Between that and balance channel transistors failing "on" (draining cell groups constantly), it's not a big leap to see why some people call them "battery murdering systems". (I don't, but can sympathize, and I don't use a BMS on my EIG packs, but I haven't taken the ones off the Luna pack or EM3EV pack).

I don't know which (if any) of the commonly available small-pack BMSs actually know that leads are disconnected. The only ones that would be useful to notify the user in this case are those able to display individual group voltages on either a built in or remote display of some kind (like the BT BMSs that can use your phone as a display).

There's a long thread for "bluetooth BMS" that has at least two different ones in it, I forget where the second newer one starts in the thread though).
https://endless-sphere.com/forums/viewt ... 14&t=88676

Post Reply