MAC 10T Hub 48v - Not sure if its given up

[BossyGoat]

So basically, the motor started dropping out as I was riding along. Would cut in, cut out and back in again whilst riding. I thought it was a clutch failure on the planetary gears (the motor is getting on for 16years old and has easily covered 15,000km+ and i've never needed to change them) but this is where my issues are now arising. Turns out one of the planetary gears was lets say a little worse for wear and the clutch plate was to put it simply, finished. I replaced those, couldn't see any other issues inside whilst motor was open.

Everything was put back together I still have the same issue, when the bike is upside down the wheel just judders and doesn't spin like its trying to engage but doesn't. Previously whilst riding and it cutting in and out I honestly wouldn't be able to say if it was the same sound or not as I never had to stop the bike to get it to re-engage, just press the throttle sometimes once, sometimes countless times. So I wouldn't know if this was the same noise or not.

One thing I did notice when removing the wheel to take the gears out. I had to unwire it to get it off (there's no plug connector at the wheel, just direct from motor to connector for the controller) whilst removing the electrical tape I noticed a fair amount of green (corrosion) Phase wires seemed fine no indicator of this. Is it possible the hall sensor wires are causing this due to corrosion? It baffles me a little as I can eventually get the thing to work as it should just intermittent rather than just not working. Could it be the corrosion on the wires causing a good connection, then a bad connection? I just want to rule everything out before attempting to find a similar motor which will be my next problem as I don't want to replace it with something that doesn't perform with the high amount of torque these do. Please help lol

 

[amberwolf]

If the wires are corroded, then there's also the possiblitiy of a short between the wires, or a break in the wires, at or near the point where the water intrusion occured.

Do the halls all turn on and off as the wheel is rotated backwards manually, while the controller is connected and powered on but no throttle applied?

If so, then the only issue is the wires, and you can simply replace the cable between motor and controller. In that event, I would recommend adding something with a (waterproof) connector for future serviceability; I use the L1019 (?) on my GMAC and on my Ultramotors. You can get them in various places with different cable types and lengths, like these Male L1019 Cable for All-Axle Motor . I'd get a pair that has long enough wires to run the entire way from inside the motor to inside the controller with no splicing, then install them inside the motor and controller directly to the points the existing wires do (replacing those), one wire at a time so you can make sure the right matching wires go to each spot.

If any halls are not fully switching on or off, or are just plain stuck on or off, even if you move the wires around a little, then you could have a failed hall or a failed controller input pullup resistor / connection. Since you see this intermittently, it's not very likely--wires are almost always the cause of intermittentness.

 

[neptronix]

My money's on hall sensors/connectors. Or controller.
There's not much else that can fail!

15km? damn, you got your money's worth even if the motor turns out to be toast. I'm amazed at that longevity!

 

[hillslayer@]

If it's the hall sensor just swap to a dual mode controller they work without the hall sensors.

 

[neptronix]

But how well does sensorless work with these geared motors? do you know?

I have tried both a Phaserunner and VESC on a DD hub and had to tune around bucking/stuttering at startup. Curing that involved lowering the low to mid end torque massively. So it was a big compromise to use sensorless.

 

[amberwolf]

PR (v6) sensorless is...well, it depends. It can startup alright on a GMAC because it has no clutch, but I'm not sure if it would do so on a regular MAC with a clutch--it might do what I've seen on other sensorless setups, where it hammers trying to figure out the right direction, or just plain spins the wrong way because it detects that the motor *is* spinning...it's just not driving the wheel. (The PR is probably smarter than that last).

But...from experience on my trike, if something like going over a bump allows the (GMAC, sensorless from mechanical damage) motor to spin up, and then it errors out when the motor hits the ground again. That doesn't happen on the exact same situation using a sensored motor (Ultramotor DD) on the same wheel of the same trike with the same PR. Keeps just going perfectly. I don't recall any startup issues or bucking, but I'd have to go back to look at the SB Cruiser thread from that time.

The older Grinfineon with sine-sensored and sensorless (trap) fallback worked ok for all the sensorless things on the MXUS 450x motors, but AFAICR didn't try it with the GMAC. It was weak on loaded startup as sensorless, but it didn't stutter.

The ancient Lishui controller that came with the Fusin "1000w" geared hub did sensorless by using a module that detected the phase feedback and emulated hall signals, like some of the older "infineon" sensorless did, rather than actually doing it inside the MCU like modern controllers do. So it didn't do any of the startup stuff very well, and nearly always spun the motor the wrong way before swapping directions and doing it the right way, sometimes doing a hammering thing by repeating that. Once it was going it was fine.

 

[BossyGoat]

My money's on hall sensors/connectors. Or controller.
There's not much else that can fail!

15km? damn, you got your money's worth even if the motor turns out to be toast. I'm amazed at that longevity!

I striped the wires back a little and there was my issue, one snapped wire hanging on by a thread. Rewired everything, worked perfect. I then taped everything and it went back to do doing the same thing. I've ordered complete new wiring, waterproof connectors. I know the problem is clearly there but working with 16year old cables I can't be bothered troubleshooting now easier to just replace it.

And honestly it's probably done more than 15,000km that was just a guess. It's used daily for work around 7/8km for the last 10years, not so much on the weekends now but I was always out on it before. My dad built this when I was 16/17, I'm now 32, so it's around 15/16years old, it's still going strong. As you can imagine it's also been hammered to death I was a teenage when I got it.. if it ever needed replacing I'd 100% want another MAC motor it's bulletproof.

 

[BossyGoat]

Ok, so update to my troubles. I have replaced all of the wiring, still doesn't work! I don't really understand because once I'd reconnected all of them previous the bike worked as it should, when I taped them up it stopped working again so I'd just assumed the wiring was shot being as it was so old so full replacement was my answer. Turns out not, still does exactly the same thing.

So I'm back for some more much need and appreciated help. When bike is switched on, I tested the voltages going through the hall wires.

Ground + live - 4.6v
Ground + blue - 5v
Ground + green - 5v
Ground + yellow - 0v

Am I right in thinking either the yellow wire has a connection problem or the actual sensor is dead? Seems a little confusing being as the first time I reconnected the wires it worked. I am probably as far away from an electrical expert you could probably get so if I'm barking up the wrong tree just ignore me lol.

I've taken the speed controller apart and all connections from there to the end of the cable is fine so I'm guessing it's motor wire back where my problem lies. I am going to get home tonight, strip the motor back down and test to make sure I've got a full connection on each wire but if there's anything else I could look for whilst doing this it would be greatly appreciated. I think I've got one more attempt in me before it gets thrown in the bin, honestly started to push my buttons now.

 

[fechter]

You want to measure each of the hall signal wires while slowly turning the wheel backward by hand. Each signal should toggle between 0 and 5v (approximately). If any don't toggle, there's your problem.

 

[amberwolf]

By "replaced all of the wiring" does that include the wiring into and out of each device, all the way to the circuit board/etc inside?

Or was it just the cables that run from the connectors on those devices to each other?


Additionally, was every point wired directly from it's original source point to it's original destination point? (meaning are they all wired exactly the same as they were before, with none of them mixed up)


On the hall signal wires, at the controller they will read 5v with no motor connected. At the motor they will read (close to) 0v with no controller connected.

When connected, one of the halls will always read different from the others, but they will all toggle between around 5v and around 0-1v as you slowly rotate the motor by hand in reverse (because it's a clutched geared motor).

 

[BossyGoat]

By "replaced all of the wiring" does that include the wiring into and out of each device, all the way to the circuit board/etc inside?

Or was it just the cables that run from the connectors on those devices to each other?


Additionally, was every point wired directly from it's original source point to it's original destination point? (meaning are they all wired exactly the same as they were before, with none of them mixed up)


On the hall signal wires, at the controller they will read 5v with no motor connected. At the motor they will read (close to) 0v with no controller connected.

When connected, one of the halls will always read different from the others, but they will all toggle between around 5v and around 0-1v as you slowly rotate the motor by hand in reverse (because it's a clutched geared motor).

No wires from the controller unit have been replaced, those were fine. I took that apart (whilst all disconnected) and just checked to make sure each connection was fine and there was no issues with that at all, also no evidence of anything being burnt out etc.

In regards to motor wiring, yes everything was replaced. Motor taken apart, hall wires unsoldered from PCB, motor wires unsoldered and replaced. Full cables all the way back to where it connects to the controller, new plug fitted, a decent waterproof connector also not one of the cheap ones they come with, which I also know was fine because after it didn't work I cut them at a point the plug was still connected and tested again from the controller and circuit was fine.

Yes all connected exactly the same as previous, I took pictures so I knew where they all went.

I will check the hall voltages again with the wheel being turned I didn't turn the wheel to get the voltages provided, they were static with battery on.

You're a god send btw, I'd have binned it by now without yours and others responses on here

 

[amberwolf]

I will check the hall voltages again with the wheel being turned I didn't turn the wheel to get the voltages provided, they were static with battery on.

Well, so far that's ok; one is different from the others. Now as long as they all toggle, and keep that part of the pattern....

At least you don't have to fix this:

The SB Cruiser : Amberwolf's 2WD Heavy Cargo Trike & Dog Carrier

clamping / pinching dropouts both ends, tight and flat to axle both ends both sides, no nuts needed on either side (have one on outside anyway just because there's axle length for it) is hubmotor with solid axles, can't see anything like qr axles or hollow ones? shear was at shoulder of axle...

endless-sphere.com

 

[BossyGoat]

So I checked them, green and yellow both moved between 0v and 5v with them wheel moving. The blue one stayed static at 5v. Am I right in saying this means either I haven't soldered the wire correctly, there's a problem with the wire or the actual hall sensor is dead?

I suppose my next step is to open the motor back up and test it's connection in there.

Yea if I opened that I'd be buying a new one some things I don't mind taking on but that looks beyond my capability

 

[amberwolf]

So I checked them, green and yellow both moved between 0v and 5v with them wheel moving. The blue one stayed static at 5v. Am I right in saying this means either I haven't soldered the wire correctly, there's a problem with the wire or the actual hall sensor is dead?

If it stays at 5v, it means the sensor isn't grounding the signal when a magnet switches it's state. That either means the controller's pullup is connected to the point you're measuring but the motor hall isn't, or that the hall itself is no longer working.

Yea if I opened that I'd be buying a new one some things I don't mind taking on but that looks beyond my capability

I got it cleaned up and regreased, fixed the board damage, and used it successfully. But in a small wheel (20" for my trike) it has to spin really fast to get 20MPH, and that is VERY LOUD even packed with grease. I may well use it as the front wheel of the new trike, in a 29" wheel, or as the core of the middrive for it's rear wheel, where it can spin slower and be quieter, cuz it sure does have some torque in there.

I don't really have the budget to be picky when there's a chance I can fix something without spending more money on it or it's replacement.

 

[fechter]

Right, the blue one has a problem. On the actual legs of the sensor you can check to see if the 5v and ground are good. If either one has a bad connection, the output will stay high (pulled up by the controller).