meepmeepmayer

[one more GOTWAY WARNING] ACM died on a hill (it was bad cabling + high stress, final update pg 16)

379 posts in this topic

1 hour ago, Jason McNeil said:

I haven't been able to follow too much of this thread, but would it be fair to summarize the conclusions of meepmeepmayers's ACM autopsy that GW need to run 14AWG wire all the way down to the stator?

So far, this is the best conclusion. Whether 14AWG (whatever that means) is enough, I don't know. There have been some calculations here on how thick the wires should be given the max currents, and they should be bigger, and naturally all the way to where they end.

One thing to note, the cables melted in a spot where they are bent the most. I've come to believe that maybe they were damaged in some way during assembly there. This would explain why the cables failed, instead of the usual weak spot which are/were the motor connectors. This is my personal theory.

I have not heard Ian's professional opinion yet (delays are all one me, don't blame him!), maybe he'll also want to look at the wheel in person. Since you're a professional too, you can look at the objective information - the photos, you can find them on page 5. If you would like me to take any specific more detailed photos to help you,  I will do so!

Very happy to hear you giving feedback to GW (will mention that to Ian too, but skipped it for now to not make my email even longer). Just maybe wait until the picture is more clear.

I agree on your other comments, though I find it hard to believe you could end up in a cycle of upgrading each component in turn and never finish. For such a simple (?) electrical device, you should be able to design a clear weakest spot, and continuously monitor how endangered it is. But that's my amateur view.

3

Share this post


Link to post
Share on other sites
On 3/19/2017 at 3:17 PM, meepmeepmayer said:

Thinking about it... how did it even melt the silicon? It's a few cm away from the melted cables, and the black motor wire right as it comes out of the hub goes between the two, and it has no visible heat damage. Induction via the magnet?

On my V3 820Wh, that is not silicon, but rather Hot Glue.  They use Hot Glue all over the place.  I think that's a bad decision as it fails prematurely.  

On 3/19/2017 at 4:27 PM, Hunka Hunka Burning Love said:

If I were Gotway I'd be specifying that their motor supplier use heavier gauge wiring coming out of their motors and insulating them from each other and the hall sensor wiring.  They seem to be using a thicker wire already from the control board supplier side.

I just don't know if the diameter of the hole out of the motor might define a wire size gauge limit.  It's looks to be purely a case of older components not designed for the larger battery packs sending more current out with the weak spot being the thinner motor wiring.

You can sort of see why @EUC Extreme  changed out the wiring completely into the motor itself all the way to the copper wiring coils.

EUC Extreme did indeed create a larger opening in the axle and nut to allow larger wires.

On 3/19/2017 at 4:53 PM, Marty Backe said:

There's nothing preventing (is there?) Gotway from feeding cables thru both sides of the wheel.

That's a good idea.  I don't think the other side is pre-drilled, but it could be designed into future wheels to accommodate larger wires.

Also, on another note, on my V3 820Wh, the wires from the motor and the wires from the board are Both 16AWG.  Just the insulator is larger on the Board side.

3

Share this post


Link to post
Share on other sites
5 minutes ago, SuperSport said:

On my V3 820Wh, that is not silicon, but rather Hot Glue.  They use Hot Glue all over the place.  I think that's a bad decision as it fails prematurely.  

<snip>

I've seen hot glue in my other wheels too. I think the Chinese love hot glue, not realizing that it's a horrible adhesive for many applications.

3

Share this post


Link to post
Share on other sites

Posted (edited)

1 hour ago, meepmeepmayer said:

to note, the cables melted in a spot where they are bent the most. I've come to believe that maybe they were damaged in some way during assembly there. This would explain why the cables failed, instead of the usual weak sp

To be honest...i have seen some of the GW production videos now...and i dont see why that cables would have been damaged!

This is 2 things: Lousy quality control, as the heat isolation around the cables seen on all other wheels is missing totally in your wheel, perhaps that would just have saved it at all?

second, after all those disconnects because of solder melted, even connectors melted on other pictures, smaller wire on motor side, which is just asking for trouble/heat, a very clear indikator -for me- that the cables are just to thin!

And as your wheel looks inside...and all circumstances together ...for me it is clear:

a long time we have seen the connector/solder as the reason for all...now this at all makes more sense, that they were just the weakest point to get blown first....

Time will tell: GW stated that now on Msupers the connectors are crimped since January, and even on all ACM 1300++ from beginning...

We will see -if- from now on on the newer models the "melting" probs will appear instead of the "disconnects" on very high torque settings...as the ACM 1300++ are relatively new much things will come later.... :-(

Edited by KingSong69
3

Share this post


Link to post
Share on other sites

Posted (edited)

2 hours ago, meepmeepmayer said:

One thing to note, the cables melted in a spot where they are bent the most. I've come to believe that maybe they were damaged in some way during assembly there. This would explain why the cables failed, instead of the usual weak spot which are/were the motor connectors. This is my personal theory.

This might be somewhat far-fetched, but was the "bent" part of the cable(s) spiraled/coiled? I've read of cases where welding machine or other cables with high currents start to melt due to being spooled on a floor, as the inductance of the coil heats up the cable during use...

Edited by esaj
Typos
0

Share this post


Link to post
Share on other sites

Posted (edited)

@esaj You mean the cables form a spiral which heats itself up additionally? I can't really say it looks like that, but the general area is very crowded so maybe induction was a factor (you can see the most in the 3rd to last picture). The most striking thing is simply that they are bent the most there. Maybe it was just a random weak point in the cable, compromised isolation, who knows.

Edited by meepmeepmayer
1

Share this post


Link to post
Share on other sites
4 minutes ago, meepmeepmayer said:

@esaj You mean the cables form a spiral which heats itself up additionally? I can't really say it looks like that, but the general area is very crowded so maybe induction was a factor (you can see the most in the 3rd to last picture). The most striking thing is simply that they are bent the most there. Maybe it was just a random weak point in the cable, compromised isolation, who knows.

Yeah, looking at the pictures, I don't think that was the case, would probably need multiple loops before there's any meaningful inductance/reactance there. Carry on ;)

0

Share this post


Link to post
Share on other sites

Posted (edited)

7 hours ago, KingSong69 said:

Time will tell: GW stated that now on Msupers the connectors are crimped since January, and even on all ACM 1300++ from beginning...

On my ACM 1600 the connectors are both crimped and soldered. Interestingly during my teardown the insulation on the 3 motor wires had shrunken back behind the electrical connector exposing bare wire strands. I believe this was due to the excessive heat being emitted during the soldering process as the soldier had wicked its' way beyond the connector and traveled up the wire all the way to beneath the insulation, melting a portion of it. The protective insulative sleeves over my 3 motor wires was the only thing that provided an insulating factor.

If @meepmeepmayer's ACM did not have the additional insulative, protective sleeves (I saw no signs of a sleeve in his photos) and the insulation had retracted back enough during the soldering process, the wires could have became exposed enough to cause a short circuit. 

Speculation runs rampant in my family:)

Edited by Rehab1
2

Share this post


Link to post
Share on other sites
7 minutes ago, Rehab1 said:

On my ACM 1600 the connectors are both crimped and soldered. Interestingly during my teardown the insulation on the 3 motor wires had shrunken back behind the electrical connector exposing bare wire strands. I believe this was due to the excessive heat being emitted during the soldering process as the soldier had wicked its' way beyond the connector and traveled up the wire all the way to beneath the insulation, melting a portion of it. The protective insulative sleeves over my 3 motor wires was the only thing that provided an insulating factor.

If @meepmeepmayer's ACM did not have the additional insulative, protective sleeves (I saw no signs of a sleeve in his photos) and the insulation had retracted back enough during the soldering process, the wires could have became exposed enough to cause a short circuit. 

Speculation runs rampant in my family:)

The bare wire was exposed on all three of my Monster connectors too, and I think on my ACM too. At the time (when I was applying sealant over everything) I thought that was really poor workmanship and/or manufacturing design.

1

Share this post


Link to post
Share on other sites
11 hours ago, Duf said:

Oh wow I had no idea your Monster got that warm.  If it gets that hot with two fans, damn.  I have never seen the temp on my Msuper out of the 40's but flat Florida helps me there.

Oddly enough my Monster hit 70C today at 8C outside ambient.  My phone said I hit 66kph (Runtastic cycle) but I don't believe it.  I was riding pretty hard through so steep hills and hard acceleration with stop starts (I was racing the road bicycles) along a 15km commute through Vancouver.  I had better pay attention in the summer. Merritt gets upwards of 40C then.

1

Share this post


Link to post
Share on other sites
1 hour ago, Merrittous MonoWheel said:

Oddly enough my Monster hit 70C today at 8C outside ambient.  My phone said I hit 66kph (Runtastic cycle) but I don't believe it.  I was riding pretty hard through so steep hills and hard acceleration with stop starts (I was racing the road bicycles) along a 15km commute through Vancouver.  I had better pay attention in the summer. Merritt gets upwards of 40C then.

Yes, I think we're all going to discover if the Monster overheats in the next few months, but the signs are not looking good.

I find this very strange. Since the ACM V2 was released Gotway has gained a reputation for creating wheels that rarely overheat. Then they create the Monster which appears to be on the verge of overheating all the time.

0

Share this post


Link to post
Share on other sites

Posted (edited)

On 20.3.2017 at 7:15 PM, Rehab1 said:

2 photos depicting the upper and lower mosfet banks attached to the 'so called' heat shield. In my opinion the heat sink on the ACM 1600 is nothing more than a piece of extruded aluminum. There are no fins to dissipate the heat.

Another upgrade added to my discrepancy list.

 

Be carefull while fastening the heatshield to the Mosfets again - you see the chip of metal at the third hole from the bottom?

Plastic insulation pieces for the screws are used and still intact to prevent the screws from touching the heatshield? Also the nut on the other side?

Edited by Chriull
3

Share this post


Link to post
Share on other sites

I just wanna add something to this topic, the other day I went to my local hobby shop to have new battery connectors installed because I fried one while trying to plug it back together while the guy was soldering the xt60 which is what my msuper v3 820 came with and I think my msuper v3 1600 has the same connectors, he said that where the two batteries are connected I need xt90 because the xt60 are too small. Am not sure what the monster comes with but if it's the xt60's I would definitely swap them out. 🤷🏿‍♂️Not sure what gotway is thinking but alot of these accidents are avoidable

2

Share this post


Link to post
Share on other sites
16 minutes ago, Hunka Hunka Burning Love said:

I wonder whether these XT90 with spark arrestor are a good choice:

We've used them on the Satiator Charger, but not on the wheel itself.  My only concern, but I don't know the answer yet is, what happens if the Anti-Spark circuit is shorted or overloaded.  Does it cut the power?  Or, is the power still routed through, just without the Anti-Spark feature?  IF it cuts out, then that's not a good scenario.

1

Share this post


Link to post
Share on other sites

Posted (edited)

The anti-spark is just a resistor that connects when you partially plug things together, but once fully connected it doesn't come into play any more since it is bypassed.  The concern is if you only partially connect it and don't push the connectors fully together or that's what I gathered from this thread:

https://endless-sphere.com/forums/viewtopic.php?f=35&t=63210

So as long as you don't dwell too long on partially connecting the two parts it should be okay?

Edited by Hunka Hunka Burning Love
1

Share this post


Link to post
Share on other sites

Posted (edited)

@Jason McNeil Great work talking to GW! But their remedies seem to relate to the motor connectors failing, not the wires themselves?

If it's really the wires (looks like it), fixing the connectors just moves the weak point somewhere else. While I appreciate GW being receptive to feedback to some degree, sometimes fixes can't help when a proper (re)design is needed (or at least some kind of theoretical calculation whether every electrical component is good enough for maximum theoretical load). Also there's no proper cable pathes/cable management solution (a pre-designed location for every wire instead of the cables basically being crammed in there somewhere) and the motor cable is too long, the bend to the left only exists so the additional cable length fits in the shell. You could easily have a combined board/cable management design so everything has its controlled place. KS is miles ahead in that regard.

Edited by meepmeepmayer
2

Share this post


Link to post
Share on other sites
13 minutes ago, esaj said:

How about replacing the connectors and even some wires with bus bars instead? I don't know if the connectors would hold securely enough with the vibration even with locking washers/thread lock glue or whatever, but other than that, at least the connection resistance should be pretty minimal? ;) Of course the shells / compartments would have to have mounting points for the bars themselves...

85858731-SS4M10-1000-1.jpg

insulated_conductor_busbar.png

EDIT: Or copper braids:

1384334783_Power%20Shunts%20-%20Copper%2

 

 

 

Old school. I like it. :D

 

1

Share this post


Link to post
Share on other sites
8 hours ago, Chriull said:

Be carefull while fastening the heatshield to the Mosfets again - you see the chip of metal at the third hole from the bottom?

Plastic insulation pieces for the screws are used and still intact to prevent the screws from touching the heatshield? Also the nut on the other side?

Thanks! Good eye but I think the hole in insulation strip was buckled up a bit on the edge giving the illusion of a metal shaving. I did find some other small filings floating around. Yes the 6 bottom mosfet anchoring screws have small plastic insulating ferrules but the top row of Mosfets do not! Wondering why now!

1

Share this post


Link to post
Share on other sites
3 minutes ago, Rehab1 said:

Thanks! Good eye but I think the hole in insulation strip was buckled up a bit on the edge giving the illusion of a metal shaving. I did find some other small filings floating around. Yes the 6 bottom mosfet anchoring screws have small plastic insulating ferrules but the top row of Mosfets do not! Wondering why now!

Typical speculation, but if the top 6 mosfets are the high-side mosfets, it doesn't matter if their drains (the backside/middle leg) are connected, as they're all connected to the same place anyway. If it would be the low-side mosfets, then you'd have a bunch of dead mosfets in no time  ;) 

_scaled_mosfet_symbol.jpg

Typical N-channel mosfet, which are used in the wheels.

 

Simplified block diagram of a 3-phase motor drive

Q1-Q3 (this is a 6-mosfet setup, for 12 mosfets, just add second ones in parallel with each separate in the above picture) are the high-side, their drains are connected to same (battery) potential, so even if the backsides of the components are in direct contact to the heatsink, it wouldn't matter (other than that the heatsink will be in the same potential as the battery). Q4-Q6 are the low-side, their drains are connected to separate phases of the motor (as well as the sources of the high-side mosfets). If the low-side mosfet backsides would be electrically connected to the heatsink, all the motor phases would be shorted together (the motor would be "stuck") and it couldn't be driven. If also the high side drains (backsides of the components) would be in direct contact with the heatsink, along with the low-sides (basically "bypassing" the high-side mosfets, ie. all the phases are at battery potential), a low-side mosfet starting to conduct would short the battery to the ground over itself, destroying the mosfet probably pretty much instantly.

 

3

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now