What The Hell Just Happened?

[Hunka Hunka Burning Love]

Better get that hand attended to.  Got some flamazine?

[esaj]

1 minute ago, Rehab1 said:

So connect both the batteries along with the inter/  battery connectors? What are the inter- battery connector used for?

I have no idea how the batteries on the ACM connect together, so can't tell... @KingSong69, do you know the wiring? Are these some datalines, charge wires, discharge wires..?

Most packs usually have just one set of wires for charging and other (usually thicker, and with something like the XT60-connector) for discharge, and with multiple packs, there's some sort of connecting-wire set to connect them all together:

EC3 -> EC5 Y-connector between the pack EC3's and the mainboard EC5:

 

[Vik's]

2 hours ago, Rehab1 said:

Omg I just tried to reconnect the batteries on my ACM 1600 to give them a charge and all hell broke loose! Burned my hand, melted the connector!

All of the cale connectors are simple plug and play. What the hell happened? 

....

 

Wow!

Doesn't sound good! Glad you're OK!

A little reminder for all of us to respect those strong battery packs.

[US69]

These "extra" wires are new, and only existing since ACM , msuperv3...

As far as i know they are for balancing/communication between the BMS's of the packs.

so there are 2 charge wires, 2 power wires, and 4 communication wires.....

[US69]

Here you can see the amount of wires...unfortunately not the plugs

[Jrkline "Wheel Whisperer"]

2 hours ago, Rehab1 said:

Omg I just tried to reconnect the batteries on my ACM 1600 to give them a charge and all hell broke loose! Burned my hand, melted the connector!

All of the cale connectors are simple plug and play. What the hell happened? 

 

Having a lot of experience paralleling EUC batteries(8 of the 10 wheels I currently own have modded batteries),that looks like a dead short,not just a connection of 2 batteries at different voltage levels,unless one of the batteries is fully charged and the other is at 0 volts.I have connected batteries inadvertently at different voltage levels,one time I connected a fully charged battery(67V) to another at (58V) and it sparked when connected,but there was no arc flash like you experienced.The spark was similiar to when you plug your charging cable into the EUC when the charger isn't connected to grid power.

[Rehab1]

1 hour ago, Hunka Hunka Burning Love said:

Better get that hand attended to.  Got some flamazine?

Thanks! Not an open wound, just forming a nice blister. Good idea though. My wife is a nurse so she is keeping an eye in it.

Now what to do about my face?

 

 

25 minutes ago, jrkline said:

Having a lot of experience paralleling EUC batteries(8 of the 10 wheels I currently own have modded batteries),that looks like a dead short...

Not sure where the short would have transpired. All of the connectors are male/ female so it would be hard to connect the wires in such a fashion to create a short circuit.

[Tilmann]

If we exclude EMP as a likely cause for a minute, one of those should help nailing the culprit: https://www.amazon.com/Epsilont-EPS1000MM-Multimeter-Auto-Ranging-instrument/dp/B01GQRXJGW/ref=sr_1_26?ie=UTF8&qid=1493516938&sr=8-26&keywords=multimeter 

Can we persuade you to put something like that in your shopping basket, @Rehab1?

[esaj]

It's not totally unheard of that the connectors fry when paralleling packs or sticking the charge connector into the charge port, although I don't immediately remember this badly burned connectors. Fault with the pack itself cannot be totally ruled out, I'd at the very least try to measure the voltages of the packs to see if there's large enough difference to explain this.

It doesn't take very large voltage difference over a suitably small (yet higher than "normal" connection) resistance to produce a very large current and power dissipation at the resistance. Many small spot-welders for battery work and such can use voltages as low as couple of volts. Here's a old simulation from the one I've been meaning to make with 10 paralleled 47000uF capacitor bank, 9 milliohm ESR per cap, charged to around 12V (which probably is too much, but anyway, I was looking at peak values). The discharge-side consists of 5 paralleled IRF2804, about 2 milliohms per mosfet (so roughly 0.4 milliohms there).  Two 1 millisecond pulses over a 10 milliohm resistance, 1 millisecond apart, power dissipation at the 10 milliohms resistance (simulating the welding prongs and the pieces to be welded, of course in most cases the resistance will likely be larger):

The total energy amount isn't that large, as the pulses are so short, but the momentary power goes above 10kW (10000 watts) and the amperage is >1000A. With a smaller bank and higher ESR-caps (much less current and the voltage drops faster), at 12V starting voltage I could already cut (as in partially vaporize) thinner component legs.

Unlike with a capacitor bank, that discharges and drops it's voltage fast with high currents (and of course in this case, the discharges are controlled), the batteries can sustain the current at high values for much longer...

[Rehab1]

2 minutes ago, Tilmann said:

If we exclude EMP as a likely cause for a minute, one of those should help nailing the culprit: https://www.amazon.com/Epsilont-EPS1000MM-Multimeter-Auto-Ranging-instrument/dp/B01GQRXJGW/ref=sr_1_26?ie=UTF8&qid=1493516938&sr=8-26&keywords=multimeter 

Can we persuade you to put something like that in your shopping basket, @Rehab1?

Thanks but my current meter is a slight upgrade to this model so no need to add to my Amazon account. 

EMP....Ugh...did my government just perform a test in preparation for bringing down another North Korean missle? 

 

[Rehab1]

2 minutes ago, esaj said:

Fault with the pack itself cannot be totally ruled out, I'd at the very least try to measure the voltages of the packs to see if there's large enough difference to explain this.

Consider it done tomorrow!

 

4 minutes ago, esaj said:

It doesn't take very large voltage difference over a suitably small (yet higher than "normal" connection) resistance to produce a very large current and power dissipation at the resistance. Many small spot-welders for battery work and such can use voltages as low as couple of volts.

That is a good analysis! My hand feels like it came in contact with an arc welder!

 

7 minutes ago, esaj said:

The total energy amount isn't that large, as the pulses are so short, but the momentary power goes above 10kW (10000 watts) and the amperage is >1000A. With a smaller bank and higher ESR-caps, at 12V I could already cut (as in partially vaporize) component legs

Amazing amount of power!

[Carlos E Rodriguez]

I am so sorry for your ordeal. One of your packs must have been low. Don't know how low would cause such an arc. 

Let say one is at 4.2 and the other is at 3.2. I don't know what is the internal resistance of these batteries. 

Connecting gthebpack to the board would make a small spark because of the big capacitors. These would have been ok. 

Connecting one pack to board then connecting second pack assuming you had them alreadybparallel on the prior try does not explain to me why this occurred. 

You indicated first try you parallel both packs first. If the batteries were uneven then you should have gotten the fireworks the first time.  Maybe I got it wrong. 

Ibsupposentherebis no way to mistakenly reverse the connections?

[Carlos E Rodriguez]

My final theory is than when you attempted to connect the second battery you tried to plug backwards. Now I understand the connectors are keyed but I thing it is possible to have one of the male pin reach and touch one of the female pin when plugging backwards and cocking the connector at an angle. 

That would explain the arc flash event. I just can not see an arc flash from uneven voltage. Uneven voltage could cause a fire due to overheated batteries discharging into the other. But not an arc flash. 

I believe you managed to short by wiggling trying to connect backwards. 

 

[Rehab1]

9 minutes ago, Carlos E Rodriguez said:

Connecting gthebpack to the board would make a small spark because of the big capacitors. These would have been ok. 

I should have continued that route as both packs were already connected together including the secondary inter- battery connectors. When a small arc occurred as I connected the battery to the main board I immediately back away!

 

13 minutes ago, Carlos E Rodriguez said:

Ibsupposentherebis no way to mistakenly reverse the connections?

Only if there is a difference between the 2 battery packs. Both packs are identical in size and number of connectors but I cannot say for certain which pack was mounted on which side of the wheel.

[Rehab1]

2 minutes ago, Carlos E Rodriguez said:

I believe you managed to short by wiggling trying to connect backwards. 

 

I will gladly check on your therory tomorrow but it would seem inconceivable based of how the female and male connectors are injection molded.  Each connector has a key and slot along with a different geometric shape at each end making it virtually impossible to join the connectors improperly that would create a short.

 

 

[TremF]

That looked nasty  I'm glad it wasn't worse than it was.

I will be taking all the innards out of my current Monster case and putting them into a new one soon. I'm almost scared to do it now 

[anims]

I am glad you are fine.
I have experienced something similar to my MCM4.. but it was my bad

It is amazing see how they melted, like nothing.. Gotway should start using XT90 for these batteries packs!

So, what is the correct procedure for plug bat?
1. Check Voltage on both packs
2. Then connect small/thin wires together
3. Connect XT60 on both batteries together
4. Connect to motherboard

??

[esaj]

9 hours ago, Carlos E Rodriguez said:

My final theory is than when you attempted to connect the second battery you tried to plug backwards. Now I understand the connectors are keyed but I thing it is possible to have one of the male pin reach and touch one of the female pin when plugging backwards and cocking the connector at an angle. 

That would explain the arc flash event. I just can not see an arc flash from uneven voltage. Uneven voltage could cause a fire due to overheated batteries discharging into the other. But not an arc flash. 

I believe you managed to short by wiggling trying to connect backwards. 

Connecting XT60's backwards doesn't seem plausible due to the shape of the connectors, they're designed so that you shouldn't be able to do that...

I doubt there was an actual (electric) arc, although Rehab1 said "My hand feels like it came in contact with an arc welder!". When electricity "arcs", it actually jumps a small gap over the air (insulator breakdown, strike through), and this needs a lot of voltage (kilovolt-range?). In this case, I'd think it's just large power loss at the moment of connection, although arcing might occur inside the connector once the copper starts to melt away, as much lower voltage is needed for strike through (compare that to arcing inside a blowing fuse, as the super-heated air turns into plasma or something along those lines). Still, when the connector sparks, I'd think most of the sparks you see are actually really tiny pieces of hot/molten copper flying, rather than actual electrical arcs.

When the connector is being connected, at the moment the connector terminals start touching, current will start to flow if there's a voltage difference between the connector. A good high-current connector will have very low resistance (like <1 milliohm) when properly connected, but when you're still pushing them together, the connection has small surface-area touching, and thus higher resistance for a while, and the resistance vs. voltage causes a large current to flow through the connector, while the resistance at the connector is still relatively high, taking the power dissipation momentarily into very high numbers.

XT60 has around 0.3 milliohms of resistance, measured by zlymex here: 

 

It's supposedly rated for 60A (like the name suggests), although zlymex said 25A is safe. Using that manufacturer 60A value, and knowing the internal resistance, the continuous power the connector is then supposed to withstand should be around 

P = I² * R   =>   (60A)² * 0.0003ohms = 1.08W

That actually sounds a bit high for a piece of plastic and small amount of copper (the connector weighs < 3 grams, or about 1/10th of an ounce) if wasted continuously, maybe that's why zlymex suggests around 25A for continuous use (25² * 0.0003 = 0.1875W). And of course you won't be using that high current all the time while riding. But still those are continuous values, and the heating and high power loss happen very fast in a situation where the connectors are being plugged together and actually get damaged (ie. the copper melts or at least deforms), so the wattage must be very high for a short duration of time.

As a simplified example, I think it goes something like this: For a voltage-difference of "just" 5V, and assuming that the voltages of the batteries themselves don't change that fast when the current starts flowing (ie. the difference stays the same, which it won't but for a very small time), and leaving out battery internal resistances to keep things simpler, over a 100 milliohm (0.1ohm) connector-resistance, you get around 50A and 250W of power. For 10 milliohms and that same voltage, the current and power will be 500A and 2500W, respectively.

Copper has a specific heat (c) of about 0.385J/(g*C). If the heat would spread evenly over, say 2 grams of copper (a single connector weighing about 3 grams, including the plastic case and both copper conductors), and the 2500W dissipation would last for 1/10th of a second (100 milliseconds), the copper temperature change could be calculated as

T_delta = E / m * S      Temperature change (Celcius degrees) equals energy (joules) divided by mass (grams) times specific heat (joules per gram celcius)

and

E = P * t     Energy (joules) equals power (watts) times time (seconds), or joules equals watt seconds

2500W over 0.1 seconds releases 250 joules of energy. Putting that into the temperature change -equation for 2 grams of copper, you get a change of

250 J / (2g * 0.385J/(g*C)) = 324 degrees celcius

Which, while hot, is not enough to melt copper. However, the heating occurs in a much smaller volume and mass than the entire connector. Again, pulling values out of my ass, let's say the part that takes most of the heat energy only had 0.1g (100 milligrams) of mass, the temperature then goes up by

250 J / (0.1g * 0.385J/(g*C)) = 6493.5 degrees celcius

Which is easily enough to melt that portion of the copper (pure copper melts around 1085 degrees celcius), unless I made any mistakes in the equations. You can imagine (or calculate) how much more or less heating occurs with different voltage difference s, currents, powers, time-frames... Of course the reality is even much more complex and changing situation over the fraction of a second when you're pushing the connector in, heat flowing away from the point in the copper, part of the copper around the contacting point(s) melting away and cutting the connection, then re-connecting as the connector is pushed further, battery voltages changing with current and internal resistances etc. etc.

However, the idea is very similar to how spot-welding works: a relatively large power is heating a very small point ("spot") for a small time-frame, and even though the total energy amount released isn't large (the total energy released in the 10kW / 1000A spot-welder simulation I showed earlier is less than 10J), it heats that small volume (small mass) of metal so fast that the heat can't flow away fast enough, and it melts.

[Rehab1]

40 minutes ago, esaj said:

Connecting XT60's backwards doesn't seem plausible due to the shape of the connectors, they're designed so that you shouldn't be able to do that...

I checked that theory in the video and it is impossible due to the unique geometric shape of the connectors.

 

40 minutes ago, esaj said:

When the connector is being connected, at the moment the connector terminals start touching, current will start to flow if there's a voltage difference between the connector. A good high-current connector will have very low resistance (like <1 milliohm) when properly connected, but when you're still pushing them together, the connection has small surface-area touching, and thus higher resistance for a while, and the resistance vs. voltage causes a large current to flow through the connector, while the resistance at the connector is still relatively high, taking the power dissipation momentarily into very high numbers.

So is there a sequence for connecting the cables together?  Like I stated earlier initially both packs were connected together with no issues but when I made the final connection to the V+ and V- pigtail coming off of the CB there was a small arc and sparks, but nothing in comparison to the second attempt where one battery pack was connected to the CB first and at the moment I connect the second pack the sparks flew.

In this video I checked the voltages and both packs are equal at 81.8 v. 

 

[esaj]

40 minutes ago, Rehab1 said:

So is there a sequence for connecting the cables together?  Like I stated earlier initially both packs were connected together with no issues but when I made the final connection to the V+ and V- pigtail coming off of the CB there was a small arc and sparks, but nothing in comparison to the second attempt where one battery pack was connected to the CB first and at the moment I connect the second pack the sparks flew.

In this video I checked the voltages and both packs are equal at 81.8 v.

Can't say for sure, if the packs were also at equal voltage yesterday, and the mainboard capacitors wer already charged, there shouldn't have been sparks at all. Like I stated before, some sparking when connecting to the mainboard can happen, because the bypass capacitors will draw a large current for very small time, but charge up fast to the same voltage as the packs. When two packs at equal voltages are connected together, no sparks should fly. So that then leaves the question why would the packs have been at different voltages when you connected the two in the latter case (first battery pack already connected to mainboard, then second pack is added in parallel). I've usually gone the route of first connecting the packs together and then to the mainboard, but if everything's ok, it shouldn't make that much of a difference.

Could the mainboard have been damaged at the first attempt (or even before), and it's drawing some current through a (partial) short-circuit, ie. small resistance path to ground? That would explain why the first pack already connected to the mainboard could have been at lower voltage when you went to connect the second pack in parallel.

Here's one way to try this out (without the batteries):

- Set the multimeter to VOLTAGE measurement (from the video it seems yours is an autoranging one, but if not or using another meter, set it to 100V or above DC)

- Measure the VOLTAGE from the pins of the connector that comes from the mainboard and to which the batteries should be connected to (ie. not the batteries, but the mainboard just on its own)

You should get a 0V reading. The point of this step is to check that there is no charge left in the capacitors, as if there ism and it's high voltage, it could damage your meter in the next step. Probably they've discharged a long time ago since yesterday. Although I'd hazard a guess that most meters would withstand high(ish) voltages in resistance-measurement -mode, just to be safe, If there is more than a couple of volts of voltage there, do not proceed to next step:

- Set the multimeter to RESISTANCE measurement

- Measure the RESISTANCE between the same connector you did check the voltage from

You should see a raising resistance value (it WILL start at 0, because resistance measurement applies a small voltage and current to the circuit, and will show 0 ohms or low values when the caps will start to charge, climbing up to the actual resistance once the caps have charged to same voltage as what the meter uses for measurement). I tested this with one of my Firewheel mainboards, and the resistance settled to around 183K ohm (183000 ohms). If the value stays at 0 or doesn't climb to something like several kilo-ohms (I'd expect at least something like that?), there's a short or at least low-resistance path to ground in the wiring or on the mainboard itself.

 

[Rehab1]

27 minutes ago, esaj said:

Here's one way to try this out (without the batteries):

- Set the multimeter to VOLTAGE measurement (from the video it seems yours is an autoranging one, but if not or using another meter, set it to 100V or above DC)

- Measure the VOLTAGE from the pins of the connector that comes from the mainboard and to which the batteries should be connected to (ie. not the batteries, but the mainboard just on its own)

60 volts are showing

 

[esaj]

1 minute ago, Rehab1 said:

60 volts are showing

Well, that was unexpected, if the mainboard has been disconnected from the batteries since yesterday...   Good thing I mentioned to check the voltage first though, eh?   If you have the power button connected to the board, you could try pressing that (the mainboard should try to boot up and probably discharges the caps fairly fast). If that's not an option, placing a resistor between the connector pins should discharge the capacitors. For 60V and a normal 0.25W resistor, something like

P  = U² / R  <=>  R = U² / P   =>   R = (60V)² / 0.25W  =  14400 ohms

To be on the safe side, I'd suggest going higher, like 20-47k or such, but probably the caps should discharge pretty fast, so the power dissipation on the resistor shouldn't matter that much, unless going to really low values (a 0.25W can withstand higher power loss for a short while). I don't know the capacities of the capacitors, so I can't calculate how long it takes to discharge, but keep it there for a little while (10-20 seconds?), then check again what the voltage is. Once it's close to 0V, you can proceed with measuring the resistance.

Probably there is a high resistance on the board, as otherwise the caps themselves should have discharged a long time ago... unfortunately then we're back to square one, what caused that huge sparking in the first place?  

[Tilmann]

47 minutes ago, esaj said:

.. unfortunately then we're back to square one, what caused that huge sparking in the first place?  

Geez, this is puzzling. Gotway electrons flow in mysterious ways, I guess. What's left besides EMP? The Russians?? Kim Jong Un???

[esaj]

9 minutes ago, Tilmann said:

Geez, this is puzzling. Gotway electrons flow in mysterious ways, I guess. What's left besides EMP? The Russians?? Kim Jong Un???

Bad luck maybe? I've had the connectors sometimes spark and sometimes not, although never as violently as in Rehab's case. The worst I've had was some multimeter probe tip vaporizing , but that was with a full short-circuit:

"Funny" thing, even though the current must have been several tens or hundreds of amperes (with a voltage drop of around 60V), the spike was so short-lived it didn't even burn the 10A fuse in the meter, apparently the BMS of the pack cut the power after the initial spike or it was cut once the tip had blown up and no contact was made anymore. Yeah, that's what you get when you don't check which jacks you've connected the probes to... was meant to measure pack voltage, had the right range selected on the meter, but the probes were left in the current-measuring jacks -> instant short circuit when the other tip touched the connector.

[Carlos E Rodriguez]

12 hours ago, Rehab1 said:

I will gladly check on your therory tomorrow but it would seem inconceivable based of how the female and male connectors are injection molded.  Each connector has a key and slot along with a different geometric shape at each end making it virtually impossible to join the connectors improperly that would create a short.

 

 

If you look there is no damage on the female connector. The male puns evaporated. Both male pins evaporated indicating that they shorted together. 

I believe my theory is that somehow during connection attemp the two male pins shorted physically. 

Another theory is that a smal arc ionized the air around the two male pins and that cascaded into a full blow short arc flash between the two pins. 

Tonprecent this, connectors need to be plugged in with commitment. If you do not slowly the arc would start and posibly cause the ionized air short. If you do it firmly the connector will be fully seated before it can form a full short.