Wiring/plug rating question

[Planemo]

I was looking at swopping out the main XT60 battery connector in my 84v MSX as 1. it's not difficult to do, 2. it would give peace of mind and 3. the clue is in the name - XT60 is rated to 60A continuous.

I wondered if it would be worth it though - I don't hear of people having issues yet surely we must be reaching this many amps on a fairly regular basis?

This led me on to looking at the wire gauge - the main feed from the XT60 to the board is 14AWG. This is even more concerning as 14AWG is only rated to around 45A continuous. As a result, I would prefer to see at least 12AWG fitted. Indeed, even my RC model lipos (2200mah/11.1v) come fitted with 12AWG as standard!

The wiring/plugs on the wheel just all seem a bit weedy to me. Am I worrying too much though?

And of course I am not forgetting the next bottleneck - the motor wires which appear even weedier but that said I don't know the rating and I suspect they may be solid core which helps a fair bit.

[EUC Custom Power-Pads]

But you know that Gotway measures the current differently than King Song or Inmotion? As far as I understand it, Gotway shows the motor current and not the battery current, like King Song / Inmotion. (Read: Battery current vs. Motor current )

The XT60 certainly does not reach its limits.

But I think Aneta or chrisjunlee can explain this better and more precisely. 

[Planemo]

Cheers yeah I'm aware of the readings not being accurate but I just felt that any 84v wheel under hard acceleration from say 20 to 30mph would pull more than 60A but maybe I am miles out on that one.

And that fact my 'tiny' 11.1v lipos come fitted with 12AWG wiring which is beefier than the wiring in the wheel just seemed odd!

[EUC Custom Power-Pads]

26 minutes ago, Planemo said:

And that fact my 'tiny' 11.1v lipos come fitted with 12AWG wiring which is beefier than the wiring in the wheel just seemed odd!

Yes, I thought the same thing when I first opened the 16X and compared it with my 12S Lipo from the helicopter. 

[Planemo]

Well 12S I can kinda understand - the burst amps on that must be mental. Mine are only 3S but even so, the manufacturers see fit to rig them with wire which can tolerate far higher amps than the requirements of an 84v MSX!

[Aneta]

My view is that EUCs should be overbuilt, not underbuilt. I don't think that anything lesser than two AWG12 wires instead of each one AWG14 wire, and two XT90 connectors, instead of one XT60 connector for the battery, and two MT60 for the phase wires, instead of one, should be used throughout. 2 instead of 1 is for a) backup; and b) for reducing heat on each wire or connector by a factor of 4x, since the current is reduced by 2x and heat is I^2. That's my plan for radical rewiring of my Rockwheel GT16.

Edited January 3, 2020 by Aneta

[Planemo]

Thats kinda how I feel about electrics too, but theres over-engineered and theres over-over-engineered. Ignoring the benefits of redundancy for a minute (which I do agree with) lets focus on what sort of currents our single (1x positive/1x negative) power feeds are actually handling. If its a regular max of say 50A then I am quite happy to stick with a single XT60.

Intetestingly, I note that one website shows 50A at 84v as being good for 4200w. Thats not bad.

I still feel that a single power line (unlike the dual lines on the Z10 which go all the way to the board) should really be at least 12awg on an XT90 for the sort of wheels in question, especially given the minimal increase in cost. That said, theres no denying the fact that the wiring on current wheels (msx etc) hasnt been going up in flames or even showing signs of overheating so why over-engineer when we might not even be close to edge...

[Aneta]

There are several examples of burnt XT60 and MT60 connectors here on the forum. It seems that the failure is rather abrupt - they look normal for a long time, and then suddenly they burn. It probably has to do with gradually failing solder - if there's an expanding crack, then initially it won't heat up much, but as it's expanding and the area the current passes through dwindles, the heat will reach the melting point. So, over-engineering and redundancy here is akin to climbers using ropes rated something like 10-20 times their weight, not just 1.5-2x, and using a belay rope as well for backup, because like in climbing, our life literally is hanging on these wires/ropes.

[Aneta]

By the way, there was a recent case (2-3 months ago) of a burnt MT60 battery connector on 16X (for some unknown reason, KS uses MT60, not XT60 or XT90, on 16X?), which serves as a perfect example of need for redundancy: since there are two battery halves, one burnt connector didn't result in a crash, the rider only noticed that something was wrong because of the reduced range. What if the phase connector failed? That would be a 100% faceplant. But if there were two of them in parallel, failure of one would be uneventful.

[Ben Kim]

continuous vs peak. Continuous is defined as the maximum amperage the connector (or wiring) can take before there is significant heat build-up. 

45A continuous on 14AWG silicone is more than enough for the purpose of an EUC as the use case for someone to ride the battery completely dead sustaining over 45A with no stopping is virtually non-existent. 

I do agree overengineering the device is preferred, but its not really any sort of doomsday scenario given Gotway/KS design choices. 

[mrelwood]

 

18 hours ago, Planemo said:

And of course I am not forgetting the next bottleneck - the motor wires

Actually, they are not the next. They are the first. If you overly stress the MSX for a minute, unless there is an underlying issue in other components, the motor wire shields will melt and the wires will short. Happened to EUC Guy while pushing a car, and to me pushing myself up to a ridiculous hill.

In my inderstanding the PWM currents get much higher than the battery current, and increasingly so at slow speeds.

If you are even the slightest bit concerned about any of these, the first thing to do is to take the motor wires out from the wire clamp and separate them as much as you sensibly can. They will go side by side inside the axle, but making the joined trip as short as possible will already buy you a good bit more headroom. The axle probably even cools them down for that section.

For the ”underlying issues”, to me the soldering job on the XT60 connectors and main capacitor legs was slightly worrysome on my mid-2017 84V MSX. Proper soldering with a hot soldering tip and a sufficient amount of solder should lessen the risk. Just make sure the soldering tab on the connector gets hot enough to melt and ”suck” the solder. It’s best to have a dummy connector attached to the one you are soldering so the pins will stay in place.

I haven’t replaced the XT60s on mine, since I’m not worried about them, and besides the stock I have seems to be of lesser quality than the originals. So I wouldn’t replace them for ”just in case”.

But separating the wires. The easiest mod in the world, and may make a world of difference.

[Planemo]

9 hours ago, mrelwood said:

But separating the wires. The easiest mod in the world, and may make a world of difference.

Do you mean cutting back the spring wire section that wraps around the wires as they exit the hub? Does seem like quite an easy thing to do.

As for the soldering in general, it all looks pretty good on my wheel. Maybe GW upped their game since earlier wheels. The capacitors, power plugs and charging port soldering all looks good and sturdy and I dont feel I could improve on it tbh.

I do hear you Ben, our wheels are not subjected to long periods of high current and indeed they do have a rather odd power demand in that its constantly and rapidly fluctuating. That said, I think that they should be built to a level where they *could* sustain top speed from full to empty, to give a benchmark reliability marker. Just as you would expect a car to be able to be driven flat out throughout an entire tank of fuel without failing due to overheating.

If a wheel could achieve the above, it is very unlikely that a component would fail due to overheating during normal riding. This is the kind of headroom I would like personally.

I guess my original question regarding how much current is readily and often seen at the board/plug during 'normal' riding has sorta been answered though. If you think that it is around/less than 45A then I'm happy as the plugs/wiring currently fitted would cope with that. I dont do sustained/steep hills nor push cars so all should be well.

Aneta - I hear you also but I feel that redundancy is a completely separate matter as a wheel should, heat-wise, be able to cope with matters without it. Redundancy should be considered only as part of a total and unforseen failure, such as a plug separating (not properly connected for example) or a solder failure (again if not done properly in the first place).

[Aneta]

Yes, exactly, redundancy should be considered only as part of a total and unforeseen failure, so it should be as standard on our hoverwheels as a reserve parachute for skydivers and a belay rope for climbers. Once it becomes standard, having no backup in life-critical operation will be perceived as crazy and frowned upon.

[mrelwood]

7 hours ago, Planemo said:

Do you mean cutting back the spring wire section that wraps around the wires as they exit the hub? Does seem like quite an easy thing to do.

They use a spring in there now? I don’t have one on mine. Maybe the cables are now shorter as well. On mine there were almost half a metre of redundant cable, which was folded on top of itself and stuffed tightly under a plastic clamp at the rear side of the axle, together with the battery cables. So all heat sources neatly together pressing against eachother, the worst possible situation.

If your motor cables exit the axle directly towards the control board and not folded over, then my rant was fruitless... I’m not sure if cutting a protecting spring for a few inches of separation would be worth it TBH.

[Aneta]

What type of connector is used for Hall wires on MSX? Are wires soldered to the pins or crimped? If crimped, it's hard to crimp thin wires reliably 100% of the time. How long are the pigtails in the crimps? On my GT16 hoverwheel, they were only 1mm long. ONE F-ING MILLIMETER!!! When I removed the hot glue which held the connectors together, a couple of wires just fell off of the pins by themselves. I replaced it with this 5-pin connector:

But I need to add another one for redundancy as a "reserve parachute". As skydivers and climbers and other risky sports athletes say, "Shit happens!"

Edited January 4, 2020 by Aneta

[mrelwood]

12 minutes ago, Aneta said:

What type of connector is used for Hall wires on MSX?

One of those 5 in a row, push to lock PC connectors. The kind that would sure make you cringe... 

[Planemo]

On 1/4/2020 at 8:03 PM, mrelwood said:

They use a spring in there now? I don’t have one on mine. Maybe the cables are now shorter as well. On mine there were almost half a metre of redundant cable, which was folded on top of itself and stuffed tightly under a plastic clamp at the rear side of the axle, together with the battery cables. So all heat sources neatly together pressing against eachother, the worst possible situation.

If your motor cables exit the axle directly towards the control board and not folded over, then my rant was fruitless... I’m not sure if cutting a protecting spring for a few inches of separation would be worth it TBH.

Deffo no excess wire. Have a look at the pic below:

Edit: Pic removed - having a cull of my uploads to retrieve space.

 

Edited March 4, 2022 by Planemo

[mrelwood]

9 hours ago, Planemo said:

Deffo no excess wire. Have a look at the pic below:

Wow, they really upped their game regarding this!

I don’t think I’d recommend doing anything to those wires after all. Seems it was just an early edition issue.

[Planemo]

Yeah I think the motor wiring is about as minimal as it can be tbh. The wiring for the extra 259wh pack is a bit of a spagetti junction though, with several added XT type connectors to hook it into the main packs. If I do add another 259wh pack I may tidy it all up somewhat.

Aneta - mrelwood is correct - the hall sensors are plugged to the board using (I think) JST type connectors. The same type as the empty 3 pin female connector seen at the middle right of the board on the pic above.

I would assume that the hall wires are crimped to the pins on the male JST plug rather than soldered. Not ideal as you say but tbh there is a lot of silicone around the plug and wires and having inspected it I am happy that neither the wires or plug are going anywhere. In fact I am of the opinion that I dont really want to disturb it all just to check the integrity of the crimping which may well be fine. Indeed, I have heard that the military prefer crimping (done well) over soldering as it removes a potential failure point should things get hot enough to melt solder which is about 190 deg C for 60/40. Way above what we would want to see on our wheels but still..

[Chriull]

On 1/3/2020 at 9:08 AM, Planemo said:

And of course I am not forgetting the next bottleneck - the motor wires which appear even weedier but that said I don't know the rating and I suspect they may be solid core which helps a fair bit.

is a nice example of appropriate motor wirings

Our used BLDC motors have 3 phases - so every fourth coil on the circumference gets the same "signal/wiring"(1). Imho similar to the battery cells they can and are connected in some series and parallel connection? In all the motor teardowns can be nicely seen (as in this picture here) that the main phase wires split up in some parallel strains for the coils.

As already written above the burden on the battery wires is much lower than on the motor wires - in average! By the PWMing one has during the duty cycle the full motor current and while 1-duty cycle (100%-duty cycle) zero amps flowing. By the parasitic inductances of the wires and the big capacitors at the motor drivers this current should be (hopefully) averaged out more or less....

With the wirings there are two main aspects - heat and voltage sag. They have to stand the heat from the power dissipation from the current and the voltage sag, so the solder points won't desolder and the insulation won't melt. The caused voltage sag by thin wires (beside this beforementioned heating) causes power loss. At each burden spike (acceleration) the voltage is missing for acceleration. So especially for low voltage RC models thick wires are crucial for performance - or even the controller could brown out.

Perfectly designed wirings in our EUCs would bring much more umpf/zippiness for driving and no more molten wire insulations. But this would fry the insufficiently cooled mosfets quite instantly - the heatsinks for them could be improved too (a bit) but still would need much better firmware protection!

This is imho nicely seen with the GW evolution over time - a borderline walk between fried mofets and molten wires...

Edit:

(1) not 100% sure about this while rereading - the coils for the 3 phases are in a 120° pattern, so this "ever 4th coil" could/should result from this?

Edited January 8, 2020 by Chriull