Why does this keep happening (XT60 connector spark)

[xorbe]

1 hour ago, Shabba said:

69v and 1v

69V/20 = 3.45V per cell, so that pack is less than half full (for 84V model).

Someone else will have to advise what to do with the "1V" pack, clearly something is wrong, perhaps BMS related.

Edited May 2, 2020 by xorbe

[svenomous]

OK, clearly the left pack (as per orientation in the photos) is non-functional.  Let's explore some possibilities:

1. Wires and connector, which seems a bit unlikely if you're getting a steady 1V reading.  If it was wiring/connector, I'd expect to see intermittent readings affected by wiggling stuff around, or just 0V.
2. One ore more dead cells.  Does anyone know what the cell layout is in these packs?  Given the voltage of the other pack, these are 20s (84V, fully charged), but I don't know how many series per pack.  Can a single bad cell drive a whole pack down to 1V, especially if there are at least 2 series in the pack?
3. The pack really got depleted to almost nothing.  Maybe a short circuit.  If so, this pack is dead and must be replaced, as all of its cells would be below their minimum voltage, and won't take a charge.
4. The BMS in the pack has gone bad. Or, could it have cut off due to a short?  If it did a cut-off for short-circuit protection, is there a way to "reset" it?

For 1, a repair would be possible as long as the problem is external to the pack (in the connector pigtail coming out of the pack.  For a wiring issue inside the pack, or for 2, a more complex surgery would be required that involves cutting open the wrapping and doing a full forensic examination including checking each cell's voltage with the voltmeter.  For 4, theoretically the BMS could be replaced, but again this is delicate surgery while dealing with potentially enormous electrical and chemical energy if something's not done right.  For 3, nothing can be done.

It seems increasingly likely that you'll have to replace the pack by buying a new one, unless someone has a suggestion.  Connecting this pack to the harness and hooking up a charger to the wheel would probably not work, as a) the charger usually won't initiate charge if it detects a voltage that is too low (part of its safety features), and b) the charge into a pack apparently so deeply depleted could cause a very nasty reaction, including bursting/outgassing...unless the BMS has simply gone into short-circuit protection cut-off, anyway.

I suggest you do nothing for the moment (except maybe finding out what a replacement pack would cost), while we see if anyone else with good battery/BMS knowledge has input to give.  How about @Chriull for example?

[xorbe]

2 hours ago, svenomous said:

OK, clearly the left pack (as per orientation in the photos) is non-functional.  Let's explore some possibilities:

1. Wires and connector, which seems a bit unlikely if you're getting a steady 1V reading.  If it was wiring/connector, I'd expect to see intermittent readings affected by wiggling stuff around, or just 0V.
2. One ore more dead cells.  Does anyone know what the cell layout is in these packs?  Given the voltage of the other pack, these are 20s (84V, fully charged), but I don't know how many series per pack.  Can a single bad cell drive a whole pack down to 1V, especially if there are at least 2 series in the pack?
3. The pack really got depleted to almost nothing.  Maybe a short circuit.  If so, this pack is dead and must be replaced, as all of its cells would be below their minimum voltage, and won't take a charge.
4. The BMS in the pack has gone bad. Or, could it have cut off due to a short?  If it did a cut-off for short-circuit protection, is there a way to "reset" it?

For 1, a repair would be possible as long as the problem is external to the pack (in the connector pigtail coming out of the pack.  For a wiring issue inside the pack, or for 2, a more complex surgery would be required that involves cutting open the wrapping and doing a full forensic examination including checking each cell's voltage with the voltmeter.  For 4, theoretically the BMS could be replaced, but again this is delicate surgery while dealing with potentially enormous electrical and chemical energy if something's not done right.  For 3, nothing can be done.

It seems increasingly likely that you'll have to replace the pack by buying a new one, unless someone has a suggestion.  Connecting this pack to the harness and hooking up a charger to the wheel would probably not work, as a) the charger usually won't initiate charge if it detects a voltage that is too low (part of its safety features), and b) the charge into a pack apparently so deeply depleted could cause a very nasty reaction, including bursting/outgassing...unless the BMS has simply gone into short-circuit protection cut-off, anyway.

I suggest you do nothing for the moment (except maybe finding out what a replacement pack would cost), while we see if anyone else with good battery/BMS knowledge has input to give.  How about @Chriull for example?

1. Unlikely, for the reasons you gave.

2. Unlikely.  Either down 4.2V per dead group, or 0V due to total circuit break.

3. Unlikely.  It takes a long time to drive cells down to 0.05V such that it becomes their maximum voltage under no load.  But technically possible.  Hmm, I do see in the post history that the wheel wasn't ridden for some months.  Maybe if one pack was left in and wheel left on ... maybe ...

4. I was under the impression that the power leads were directly connected to the batteries, and the BMS could only cut the incoming charge current.  However I'm at a loss of any other explanation, so maybe there is some huge switching mosfet in there for protection.

Unless someone else knows for sure, I suspect this involves opening the battery pack for diagnosis, with direct measurement of cell groups.

Edited May 3, 2020 by xorbe

[Chriull]

Bad BMS sounds very probable.

If this is a GW BMS they have no protection circuitry at the output - it's directly connected to the battery cells.

So there could also some crazy polarity changing been going on.

Did i miss something about the history of the pack or was just nothing written?

Anyway - it has to be opened and thouroughly examined! Especially cell voltage measuremend should be made not only to look at the absolute voltage, but the polarity too!

With quite some safety measures - best at some place it can safely burn out, just in case.

... or just safely dispose it...

Depending on "the history" the other pack should be charged (attended) as long as the wheel is still open to see if it reaches 84V...

[rS22]

I made a simple pigtail tool that solves the arcing problem for me. In the S22, I don’t really think there’s room to add XT90’s. So I made a cable with male and female XT60’s on the ends, through wires on the two small pins, and an XT90 in the middle. When reassembling the wheel, I connect the first battery using the pigtail, with the XT90 as the last connection. That allows the capacitors on the main board to charge back up through the anti-spark XT90. I then remove the pigtail, and reconnect the batteries without sparks. I hope this is helpful to someone.

https://drive.google.com/file/d/1JuJ3yFjxjQ7W-7zuGIQ07cULpY3rxmnz/view?usp=drivesdk

[alcatraz]

On 7/7/2023 at 1:20 AM, rS22 said:

I made a simple pigtail tool that solves the arcing problem for me. In the S22, I don’t really think there’s room to add XT90’s. So I made a cable with male and female XT60’s on the ends, through wires on the two small pins, and an XT90 in the middle. When reassembling the wheel, I connect the first battery using the pigtail, with the XT90 as the last connection. That allows the capacitors on the main board to charge back up through the anti-spark XT90. I then remove the pigtail, and reconnect the batteries without sparks. I hope this is helpful to someone.

https://drive.google.com/file/d/1JuJ3yFjxjQ7W-7zuGIQ07cULpY3rxmnz/view?usp=drivesdk

Wheels should really come with these pigtails. The cost is near nothing.

I bet you could use a special one to equalize pack voltages too. That solves a lot of issues too.

[mrelwood]

 

1 hour ago, alcatraz said:

I bet you could use a special one to equalize pack voltages too.

The caps are fast to charge, since they hold very little energy (like we learned from the GodMod saga). Batteries though, they hold a huge amount of energy. Bleeding a 100V 1800Wh pack even for a single volt would take a reasonably sized 50W resistor something like 1.5 hours. If the resistor is attached to a heat sink that is.

The included chargers are much faster and more convenient for balancing packs. Besides, how often do you have to balance them manually anyway?

[alcatraz]

It's an easy safety measure to use a pigtail before connecting something together. If the pigtail gets hot then do NOT proceed. It's a simple straightforward step that would prevent a couple of failures.

I promise the companies would earn back the cost of it by fewer support inquiries etc.

I wouldn't be surprised if you get a bit longer lifespan out of the caps too by using one. Also, if you have fuses on the battery inputs, do you really want to run a split second current rush through them, again and again, and expect them to stay within spec? Seems like an unnecessary risk just to avoid one simple step.

Edited July 9, 2023 by alcatraz

[mrelwood]

As per the quote in my post, I only commented about having a “pigtail” that would equalize the pack voltages.

1 hour ago, alcatraz said:

if you have fuses on the battery inputs, do you really want to run a split second current rush through them, again and again

And how many times is this “again and again” for an average user?

[alcatraz]

3 hours ago, mrelwood said:

As per the quote in my post, I only commented about having a “pigtail” that would equalize the pack voltages.

And how many times is this “again and again” for an average user?

No worries. I'm just brainstorming. 

I wouldn't be surprised if the automotive industry requires the use of such devices when working on cars. Like the larger the caps, the greater the need. 

Wheels are only going to get more powerful. Might come in handy. Just saying... 

[alcatraz]

More times than with a pigtail. Is it dangerous, maybe not. Will it make a difference for a 2-10 year old wheel out of warranty? Possibly. 

[rS22]

KingSong just put out an invitation on Instagram to email feedback to their CEO. So I sent the following (sorry if some of the info strays from the subject of this thread). I’ll post any reply I get here. Who knows what effect, if any, it will have, but you gotta try, right?

Dear Sir or Madame,

I am a happy owner of a KingSong S22 Pro. I have logged about 380 miles on my wheel, and I offer you the following feedback on the product:

1) I have decided that I will disassemble the wheel every 300 miles to clean the moving parts and check torque on all fasteners. I would appreciate having a reference online for torque specs, and a suggested preventive maintenance schedule, including rebuild information for the shock (similar to what is offered for high-end mountain bikes). 

My wheel, after 350 miles, developed a shudder or jitter during hard braking, which turned out to be caused by all sixteen of the screws holding the roller rails to the wheel frame being loose enough to allow movement. Tightening these screws eliminated the jitter.

2) Please consider using sparkless connectors (e.g., XT90) for battery disconnection. Since the main board must be disconnected to perform disassembly of the wheel, the user is required to repeatedly disconnect/connect these connectors. The initial connection causes arcing, which degrades the connectors. I believe there isn’t sufficient room in the wheel to add these connectors, so I have built a tool (a cable assembly, pictured below) that I use for the initial reconnection, which uses an XT90 to charge the capacitors on the main board. Once this is done, I remove the tool, and reconnect the batteries without arcing. I believe the sudden inrush of current to the main board, and the arcing at the connector should be prevented, and if prevented might decrease failures. A tool like this could be included with the wheel, or made available as an accessory.


Thank you for your consideration.
 

[alcatraz]

Are XT90 sparkless? I just wonder how that would work?

[rS22]

Just now, alcatraz said:

Are XT90 sparkless? I just wonder how that would work?

Yeah they are. It was new to me. One of the female pins has two contacts: one shallow, with a plastic spacer, and one deep (like an audio connector). The shallow one is tied to the deep one with a resistor that limits the current as you’re plugging in, and as you push the connector all the way home, you make contact with the second contact. So it prevents the arc at the shallow contact.

[0000]

Can also confirm this works. Something I did with my RS when I had to do a deepclean after a submersion. Harness came out in pieces, so it was convenient to replace the XT-90 with an XT-90S connector off the wheel. A modification you'd never want to perform while connected to any batteries for anyone reading btw.

The one caveat is I don't know if the embedded resistor is rated for high voltage, so it may suffer a decreased lifespan in the number of connections it survives. Probably fine though as it's not something you're disconnecting/reconnecting every ride.

Edited July 15, 2023 by Vanturion