Another MSuper V3+ Wiring failure - Ug

[Radislav]

That's how they fix motor wires:

 

[esaj]

48 minutes ago, Radislav said:

That's how they fix motor wires:

Nice, if only they'd have done it like that from the beginning   Nice (hydraulic/pneumatic?) hammer-thingamabob for crimping, it would suck to try to crimp that big connector by hand with a pair of pliers  

[LuckySXM]

I've done some tests with my FLIR camera of the heat generated at the connectors when doing some forced motion, both by hand, and standing on the wheel.

Below are the pics with temps. I had just put shrink tubing on two of the connectors to hold them together, and left one connector bare with metal visible (and full of danger for my calf and body integrity).

Pic 1 : wheel turned off. (ambient temperature 28 degrees Celsius).

Pic 2 : wheel turned on, idle.

Pic 3 : forced back and forth motion by hand on handle.

Pic 4 : 30 seconds of forced back and forth motion while standing on the wheel (bear in mind I had to get down of the wheel, grab phone, aim and take picture, which took about 10 seconds, during which time the temperature decreased by a good amount).

I'm guessing temperature could go as high as 100 degrees when climbing under hot weather, which is water boiling temperature...

Sick.

A few weeks ago I saw more than 70 degrees in the Gotway app after a huge climb under hot weather (it's always hot here anyway), but since I don't know where the sensor is placed, I don't know how it relates to the temperatures I registered today on the connectors.

As a heads up, in the interval between the time I ripped open the black plastic connector covers to check the welding, and today when I put shrink tubing, I had secured the connectors with some general purpose tape that's given for withstanding 93 degrees Celsius.

I did only one ride with that temporary setup.

When I took the tape off today, it was partly melted at a few small spots...

Scary.

[Marty Backe]

27 minutes ago, LuckySXM said:

I've done some tests with my FLIR camera of the heat generated at the connectors when doing some forced motion, both by hand, and standing on the wheel.

Below are the pics with temps. I had just put shrink tubing on two of the connectors to hold them together, and left one connector bare with metal visible (and full of danger for my calf and body integrity).

Pic 1 : wheel turned off. (ambient temperature 28 degrees Celsius).

Pic 2 : wheel turned on, idle.

Pic 3 : forced back and forth motion by hand on handle.

Pic 4 : 30 seconds of forced back and forth motion while standing on the wheel (bear in mind I had to get down of the wheel, grab phone, aim and take picture, which took about 10 seconds, during which time the temperature decreased by a good amount).

I'm guessing temperature could go as high as 100 degrees when climbing under hot weather, which is water boiling temperature...

Sick.

A few weeks ago I saw more than 70 degrees in the Gotway app after a huge climb under hot weather (it's always hot here anyway), but since I don't know where the sensor is placed, I don't know how it relates to the temperatures I registered today on the connectors.

As a heads up, in the interval between the time I ripped open the black plastic connector covers to check the welding, and today when I put shrink tubing, I had secured the connectors with some general purpose tape that's given for withstanding 93 degrees Celsius.

I did only one ride with that temporary setup.

When I took the tape off today, it was partly melted at a few small spots...

Scary.

Very cool. I am surprised the temps would get that hot. But boiling water temperature is still not melting soldering temperature.

I've thought about this a little. If the concern is that the solder is going to melt at the connector, why aren't people concerned that the solder will melt where the wire feeds into the circuit board.

The failure mechanism is probably due to a cold solder joint that fails with excessive vibration. But I'm just guessing like everyone else. Hopefully the silicon applied across my connectors will save me if the connections are marginal.

[Hunka Hunka Burning Love]

I would be concerned at both ends of the wire so that's why I think some sort of locking connector to the control board would be best.  All connectors should be crimped and soldered along with heatshrink tubing for extra assurance.  Waterproofing would be a bonus.  Hopefully the wires inside the motor that connect to the copper branches are well secured as that could be another area of concern.  It really doesn't take a huge amount of extra effort to weed out these little bugs, but hindsight is always 20 20.  Hopefully these issues will turn into standards that all manufactures will follow.

Maybe it could be that the connector metal could be retaining the heat more in that area like a heat well so as it builds up it melts the solder there whereas the control board is plastic with thin copper traces so there is no large mass of metal present.  Just a guess, of course. 

@Pingouin I would think that epoxying / applying silicon sealant to the connector / wire should hopefully be enough if the solder becomes molten, but maybe check to see if your connectors are in the affected bunch or not.  As long as there is no opportunity for physical disconnection during vibrations things should be okay I'd imagine.  Molten solder still conducts electricity.  If you were a heavier rider doing some serious climbing, I'd recommend cutting those connectors off and replacing them with new crimped and soldered ones with heatshrink tubing.

[LuckySXM]

49 minutes ago, Marty Backe said:

I've thought about this a little. If the concern is that the solder is going to melt at the connector, why aren't people concerned that the solder will melt where the wire feeds into the circuit board.

I am. While working on my wheel, my eyes kept being drawn to that area...

When I opened up my Inmotion V8, for other reasons, I saw big, solid, heavy connectors on the motherboard, with cables leading directly to the inside of the motor. No bullshit design. Heavy duty stuff.

I don't see why Gotway can't use similar design. After all, we're talking of potentially life threatening weakness here.

Mind blown.

[Marty Backe]

2 minutes ago, LuckySXM said:

I am. While working on my wheel, my eyes kept being drawn to that area...

When I opened up my Inmotion V8, for other reasons, I saw big, solid, heavy connectors on the motherboard, with cables leading directly to the inside of the motor. No bullshit design. Heavy duty stuff.

I don't see why Gotway can't use similar design. After all, we're talking of potentially life threatening weakness here.

Mind blown.

This has already probably been posted elsewhere so forgive me if I'm being redundant here. Doing a search for the melting point of various solders shows that that on the low end (cheapest?) solder melts at 183c   I just can't see that happening. Your temperature readings were exceedingly far from this temperature. That's a good thing.

[Hunka Hunka Burning Love]

That might be true for commonly found commercial solder from a reputable source.  When you start adding in different blends of metals that melting temperature can vary quite a bit.  If you read that pdf I posted about low melting solder you can see some interesting melting points.  Without having a sample of the Gotway solder to send to a commercial testing lab, it's going to be pretty much heresay, but there must be something that made Gotway decide to switch to a crimped and soldered connector rather than say using a hotter soldering iron and telling the soldering guy to heat up both the connector and wire extra lots.  They also use heat insulation tubing near the connectors, and if temperatures weren't getting up there why would they bother adding those?  I've only seen those used around halogen lighting wiring sometimes and other areas like range elements where heat is a concern.

I'm no expert, but from what I have seen, I would think a cold solder joint would pull out of the connector with a solder blob that is shaped to match the connector or the wire would pull off leaving the bulk of solder on the connector.  If there is solder on the connector and a blob of solder on the wire, I'm thinking the solder must have gotten molten enough for the wire to pop out with vibration.  Even if I wanted to, I think it would be difficult to intentionally form a lot of cold solder joints where things fall apart that easily considering the weak forces on the wire from riding.  

[Marty Backe]

21 minutes ago, Hunka Hunka Burning Love said:

<snip>
 They also use heat insulation tubing near the connectors, and if temperatures weren't getting up there why would they bother adding those?  I've only seen those used around halogen lighting wiring sometimes and other areas like range elements where heat is a concern.
<snip>

Good stuff. Regarding the insulation, Gotway seems to be all over the place. On the Monster there was no insulation tubing on the individual motor wires. Then today I did my MSuper V3 (manufactured earlier than the Monster) and it did have insulation tubing on the individual wires.  Maybe they let each assembler decide how to package the wires based on how they feel that day 

[US69]

8 hours ago, Marty Backe said:

This has already probably been posted elsewhere so forgive me if I'm being redundant here. Doing a search for the melting point of various solders shows that that on the low end (cheapest?) solder melts at 183c   I just can't see that happening. Your temperature readings were exceedingly far from this temperature. That's a good thing.

In the original thread "suggestion to all gotway users" from december...it was the user zlymex who pointed out that the heat on the connection came from the "resistance" that sits on that connection!

on a nice and good soldered connection that would be only about 0,03 milli-Ohm...

On the motor wires are up to 120Amps running....a bit higher resistance would then generate a good amount of Watts on the soldered connection, and thats when it melts!

And now the scary thing:

a "crimped" connection, like GW is suggesting now in its "solution" even has a much, much higher resistance, so that would generate even more heat on the connection , so i dont know if it is as professional as it looks...

Just take a look what the heat can get done here:

 

[Marty Backe]

35 minutes ago, KingSong69 said:

In the original thread "suggestion to all gotway users" from december...it was the user zlymex who pointed out that the heat on the connection came from the "resistance" that sits on that connection!

on a nice and good soldered connection that would be only about 0,03 Ohm...

On the motor wires are up to 120Amps running....a bit higher resistance would then generate a good amount of Watts on the soldered connection, and thats when it melts!

And now the scary thing:

a "crimped" connection, like GW is suggesting now in its "solution" even has a much, much higher resistance, so that would generate even more heat on the connection , so i dont know if it is as professional as it looks...

Just take a look what the heat can get done here:

 

That's a good thread to re-read. Did you ever solder your wires?

I think I've exhausted all of my available mental energy on this subject. It's a very deep rabbit hole. I've restrained my wires/connectors with silicon and sticking with statistics for now.

[Hunka Hunka Burning Love]

2 hours ago, Marty Backe said:

Good stuff. Regarding the insulation, Gotway seems to be all over the place. On the Monster there was no insulation tubing on the individual motor wires. Then today I did my MSuper V3 (manufactured earlier than the Monster) and it did have insulation tubing on the individual wires.  Maybe they let each assembler decide how to package the wires based on how they feel that day 

It could be the result of product refinement over time or just a response to specific models?  Hard to say.  I'm not even sure why they have that knitted heat sleeving on the wires.  What's it's purpose?  Maybe they want to keep the wires seperate and insulated so they don't melt the wires themselves?  Dunno.  You would think they would have some standard practices that are universal.  Maybe some wheels aren't as prone to this disconnect issue.  Hard to say.

It sounds like for the majority of people who are of average weight for their height who aren't climbing excessively steep hills these problems aren't very common so that's good.

[Marty Backe]

7 minutes ago, Hunka Hunka Burning Love said:

<snip>

It sounds like for the majority of people who are of average weight for their height who aren't climbing excessively steep hills these problems aren't very common so that's good.

Yeah, but I do climb excessively steep hills 

[Hunka Hunka Burning Love]

Maybe I could have worded that better...   If you're waaay overweight for your BMI index and are climbing steep hills there could be trouble.  If you have a normal BMI and climb steep hills everything likely will be fine.  Gotway is just encouraging their riders in a roundabout sort of way to maintain a healthy weight while riding uphill!  

[US69]

2 hours ago, Marty Backe said:

That's a good thread to re-read. Did you ever solder your wires?

I think I've exhausted all of my available mental energy on this subject. It's a very deep rabbit hole. I've restrained my wires/connectors with silicon and sticking with statistics for now.

not yet...will have to this weekend...as every steep hill with a bit of pace makes me nervous as hell now.

And i have taken a look in my wheel, and the connections look all, but not good :-(

 

so i guess i will solder them direcly with a wire around....or redo the connector on motor side nicely!

unfortunatly my solder skills are as good as Gotways...could be better :-)

[Pingouin]

6 hours ago, Hunka Hunka Burning Love said:

@Pingouin I would think that epoxying / applying silicon sealant to the connector / wire should hopefully be enough if the solder becomes molten, but maybe check to see if your connectors are in the affected bunch or not.  As long as there is no opportunity for physical disconnection during vibrations things should be okay I'd imagine.  Molten solder still conducts electricity.  If you were a heavier rider doing some serious climbing, I'd recommend cutting those connectors off and replacing them with new crimped and soldered ones with heatshrink tubing.

I did try to pull on them before adding some epoxy, they didn't seem weak at all, after looking at it following a high power demanding ride, the epoxy didn't melt and the connectors where still firmly attached, I think I'll check on them every 1000km, or after a very steep climb, I don't feel like resoldering the cables as it might void my warranty :/

Gotway Kebye doesn't seem to give a shit about it when I mailed them they didn't even reply to my question about how to make things secure, they only replied to my other question ignoring the "disturbing" part...

[esaj]

8 hours ago, Marty Backe said:

I've thought about this a little. If the concern is that the solder is going to melt at the connector, why aren't people concerned that the solder will melt where the wire feeds into the circuit board.

The failure mechanism is probably due to a cold solder joint that fails with excessive vibration. But I'm just guessing like everyone else. Hopefully the silicon applied across my connectors will save me if the connections are marginal.

Connectors always have (relatively speaking) much higher resistance than wires or soldered joints. If the connectors indeed heat high enough to melt solder (around 183 Celsius for eutectic Sn/Pb, ie. leaded solder and something like 220-240 Celsius for Sn/Ag and other non-leaded compositions), the heating happens at the connectors. Or a very badly done solder joint (higher resistance). When I say "higher resistance", it's still in the low milliohms-range, not full ohms.

If you have a resistive circuit (not taking into account any AC-stuff, like reactances from capacitances or inductances or such), the point(s) which will "waste" most power into heat will always be the highest resistance(s). Of course in the wheels most of the power goes into the motor and thus motion, but in any electrical circuit (that's not superconductive) power is wasted as heat at least at some point, even high efficiency switching power supplies have an efficiency of up to 90-97% or thereabouts, the rest from the "100%" goes to heat in power losses. The reason linear power supplies are big and heavy is that they have low efficiency and waste a lot of power as heat when they drop the voltage and thus large heatsinks and lots of thermal mass is needed inside the unit (my 100W linear-PSU weighs about 10kg or so).

How it relates to this is that there's always some resistance between the joined male- and female-connector, the connection is never "perfect". @zlymex posted some measurements of different connector resistances (it continues for more posts if you scroll that down):

 

 

 

Now, let's say the connector is like that "Gotway V3 Motor connector" in Zlymex's picture. 1.3 milliohms for the crimped end (let's say that was the badly soldered wire instead, ok ), plus 1.0 milliohms for the actual connection between the male- and female-connector plus 0.03 milliohms for the soldered end, for a total of 2.33 milliohms (0.00233 ohms). Doesn't sound like much, does it? Shoot a 50A of current through it, and you get a voltage drop of

U = R*I   =>  U = 0.00233ohm * 50A = 0.1165V

(116.5 millivolts) over it. Still doesn't sound too bad. How much power is it wasting there?

P = U*I = P = 0.1165V * 50A = 5.825W

Or, more directly (no need to calculate the actual voltage drop by replacing U in P=U*I with the equation for U):

P = U*I = R*I*I = R*I² 

Which shows us that power loss is relative to resistance and the square of the current. Double the resistance, the power loss doubles. Double the current and the waste heat quadruples.

So, is 5.825W a lot? Yes and no, it then comes down to a lot of different variables, how fast the copper can shed the heat to the surrounding air, how much mass the connector has, how long does that amount of current run through it (it takes time to build the heat up) etc. Copper is a great thermal conductor, which is why it's preferred in heatsinks vs. aluminum (and then aluminum is used because it's much cheaper than copper ).

Now, if you double the current to 100A, you get some more serious wattage for that small thermal mass: 23.3W. There are soldering irons that have less wattage (15W for example, although of course they take time to get up to high heat, but then again there's probably more thermal mass in a soldering iron tip than that connector).

Maybe I was too quick to dismiss the idea that the solder tin could melt inside the connector. With bad connection (ie. "high" resistance like several milliohms) accompanied with high current draw for an extended period of time could actually heat the connector above the solder melting point. The way to fix that is to use proper connectors with very low contact-resistance (high quality bullet types come to mind) and properly soldered wire <-> connector -joints (to help with mechanical stress crimping could also be applied, like Gotway seems to be now doing).

 

Another example how hot bad connections and high current can make things:

(the red part is a metal nut glowing very hot, not see-through plastic or such )

 

[LuckySXM]

So that explains the advice from the guy from French reseller Urban360 : "while you're at it, use a very fine blade to spread the four parts of the connector further away so as to make better contact between the two parts of the connectors when plugin them together, to reduce resistance from bad contact, and thus reduce heat".

This guy does the maintenance for all wheels sold by them. He knows what he is talking about, it seems.

[esaj]

24 minutes ago, LuckySXM said:

So that explains the advice from the guy from French reseller Urban360 : "while you're at it, use a very fine blade to spread the four parts of the connector further away so as to make better contact between the two parts of the connectors when plugin them together, to reduce resistance from bad contact, and thus reduce heat".

This guy does the maintenance for all wheels sold by them. He knows what he is talking about, it seems.

The "best" way (or at least lowest resistance) would probably be to directly twist and solder the phase wires of the motor and the mainboard together (without any connector in-between), but that would make possible future maintenance/repairs harder, as the wires would either need to be cut or desoldered to detach the motor from the board.

[LuckySXM]

Totally agree.

[kour]

was going to bump up some more posts in here but was informed i have a '1 reputation-a-day' limit; interesting stuff!

[Marty Backe]

1 hour ago, kour said:

was going to bump up some more posts in here but was informed i have a '1 reputation-a-day' limit; interesting stuff!

Hang in there. I'm allowed up to 40 (which I've bumped up against believe it or not) 

[kour]

haha, word - thanks for the encouragement