Mark_A_W's DH Team/MAC Shanghai/64v Headway build

Disaster. I'll let Mark tell the story when he's done reassembling the motor with the fresh set of metal gears, here are the pics:

Yeah, the plastic gears lasted 3 days.


And I was running less current, and going easy...ok, easier.


Interestingly, they failed with a CRACK, but on a tiny little bump, barely under power, at walking pace.


What's concerning is the little steel key, which sits in the keyway on the axle, and stops the freewheel assy rotating, was completely flogged out. Unfortunately the keyway on the axle is a bit rough too - the axle is hardened, but not that hard.

The freewheel/gear assy should sit on a SPLINE, not a single KEY.

I made a key out of some bar stock, and the freewheel is no longer rocking back and forth ~15 degrees.


I have 3 steel gears from Full-throttle to get me home, then I'll take them out to give back to him.


Nylon gears are no good. Steel gears are no good.



Edit: Gears 3.jpg above shows my steel gear, after 200km's, running greased. It is showing serious wear. It is not being reused.

great

but...they've failed in the hands of an excellent Mechanical Engineer :wink: perhaps a real solution will come out of this? We need to figure out how to make the motors handle 1.5-2kW peak loads, even if it's beyond the design limit, otherwise they're not that useful to a *power commuter*.

Mark_A_W said:

Nylon gears are no good. Steel gears are no good.

Click to expand...

Damn Mark :-( Whats your next course of action? I spoke with Matt.P yesterday and
he is keen to get one of the MAC motors so he will be keenly
watching your progress no doubt, i just emailed him with this recent news.
Time to hunt down a machinist to crank out some higher quality gears from some more
durable material is it?

KiM

A glass reinforced plastic - not sure what plastic. Could even be an Aramid fibre reinforced plastic.


Or cloth reinforced bakelite (phenolic).



The gears worry me less than the keyway slot on the axle. I can see myself having to completely disassemble the axle to get another keyway machined on the other side. Then that will fail too...

It's a cyclic shock load. It needs a spline. But then you couldn't get the rotor off. I see why they did it the way they did.

The BMC motor has the same keyway/freewheel design.

The key is the little steel block in the pics above.

I dunno how to fix the keyway properly, not without designing a whole motor from scratch.

Bugger

Aren't some of the BMCs in the US shipping out with 50a controllers ? If the internals (bar the gear composition) are similar and they're using the same keyway and freewheel design why aren't they failing ? Cheaper, softer metal in the macs ??

Hyena said:

Aren't some of the BMCs in the US shipping out with 50a controllers ? If the internals (bar the gear composition) are similar and they're using the same keyway and freewheel design why aren't they failing ? Cheaper, softer metal in the macs ??

Click to expand...

Good question. Should go in conjunction with: Why BMC V2 is four times the price?

Edit: I might as well point out that Bafang BPM uses a spline instead of key hole and while the gears are nylon they are double in width. I have no problems with mine so far after 4,000km @ 700W cont / 1,400W peak. Oh, and its lighter, cheaper and well sealed. Too bad the axle does not fit Mark's DH 150mm dropout spacing.

Mark,

sorry to but in on your build thread but I just wanted to say a little about the issues you've had with the Mac steel gear. I've forwarded your comments about the Mac motor to the company owner and he's looking into it now. Hopefully he'll either post on here or forward something to me and I'll relay it on.

I've personally been running a Mac 500W for a few months and have been using the steel gear motor the last couple of weeks. I'm gonna strip it down soon and see how the steel gear is fairing but I have taken onboard what you have said about the gears and general construction. I'm gonna be supplying the motors I have with the nylon gears for now unless otherwise requested and will supply a spare steel gear FOC. I've also requested spares prices and hope to have these parts available soon. If you need me to sort these out for you just give me a shout.

Also I'd just like to say that I didn't choose to buy Mac motors because they were the cheapest, I just liked to do business with them and I think they are a good honest company that genuinely care about their products. They want to improve their products and are always keen to hear feedback. However they do not generally receive much in the way of negative comments about the motor performance, because in fact it is under rated at only 500W, it performs very well at what it was designed to do, and is very reasonably priced. I haven't been able to devote as much time as I would have liked to these motors and to put together a kit as I've still got shedloads to do with the cells so haven't tested the new batch of motors as thoroughly as I would have liked. I'm not a mechanical Engineer but I do have an Engineering background, granted more Electrical/Electronics but I would hope I have a reasonable grasp of Mechanical Engineering.

Anyway if there is anything I can do, or if you want to arrange a group buy or something, give me a shout and I'll see what I can do.

Cheers
Paul

If the sun gear didn't wear, and the ring gear didn't wear, only the planets wore, then harden the tooth surface of the planets.

If you find the right metal treatment place, they should be able to double or tripple the surface hardness (depending on the alloy of course), and then apply a DLC coating to the gear.

We sometimes get tester camshaft grinds that show 100,000miles of wear in a single 2hr dyno session. Get the same cam inductively hardened, then DLC coated, and after a season of abuse it looks like new.

Since you've all ready got a set of wrecked gears, poke a rockwell tester into the tip of a tooth, then the side of a tooth, then the middle of the side of the gear. If you see roughly the same hardness result in all places, it means the gear was left as-cut on the hobber. They would never hob a pre-hardened gear blank of course, and likely just figured being a step-up by maybe 100x over the nylon gears that it just wasnt needed.


Trying to make something with gears reliable at 4x the design power isn't generally too easy or cheap. Making Honda tranny's with 200hp intended design loading handle 800hp for an extended period of time is not an easy or cheap game to play. This game looks like a much easier fix if ONLY the planets wear and the sun/ring are in good shape still. Perticularly if they are still in good shape after having meshed with worn planet tooth profiles for a while, which would show they are pleanty tough for the job.

Thanks guys.


Yes, just to be clear, I cannot complain about this motor at all.


I am putting three to four times the power through it than it is designed for. It is my choice. If I ran 48v 25A it would be fine, with great speed for that power level.


Yes, I can get the steel gears hardened (and yes they are straight off the hobber). But you shouldn't run hardened steel gears without proper lubrication - they need to be immersed in oil....and that causes too much drag for the rotor.

And they are noisy, even in oil, but new gears are quieter than worn ones.


For feedback to MAC Shanghai, the weak points (at 1500-2000 watts) are:

- Gears

- Keyway on the axle

- Key itself

- Posts that hold the gears to the freewheel


If the Freewheel/gear assembly was on a spline (the Bafang BPM uses four splines), then it would be fine.

It appears the BMC composite gears are adequate, nobody has stripped them as far as I am aware.


I'm going to inspect my motor this weekend and check out the key that I made to replace the original. If I can fix that, then I may get some stronger gears hobbed (bakelite for instance).

Oh, Mr Physics...

Any idea if I could braze the (softer) key to the damaged (harder) keyway on the axle?


I don't think welding would work.

But if I file/machine the damaged load face flat, so the key makes good contact, then fill the gap on the other side with braze solder, it won't be able to rock back and forth.


It is the rocking that kills it.


I'll try and get a picture, if the new key is working loose already, as the keyway is too wide now.


Or I could get the keyway machined a touch wider, and make a key with a step in it. 5mm wide for the freewheel, 6mm wide for the axle. Then braze that in so it doesn't wiggle...

May save a bundle of money getting a set hardened over getting custom gears made.

The DLC coating is to act as a wet lube bath replacement on the hardened gear. It has some pretty amazing anti-wear properities, both on it's own surface, and the surfaces it meshes with.

Maybe a hardened tooth profile, micro-polished finish, then DLC coated. Here is a white paper about DLC as a replacement for lube on high pressure hardened surfaces here:


http://www.thefreelibrary.com/Scuffing+resistance+of+DLC-coated+gears+lubricated+with+ecological...-a0216178973


I would bet under $150 for the hardening, micro-polish, and DLC coating all combined. Maybe under $80 if you can find a single shop to do it all in-house.

For the sloppy keyway problem, perhaps assembly with a little metalized liquid epoxy? Something like the devcon for steel? It can do wonders for getting some more life out of worn splined connections, never seen it used with a keyway, but the multi-sized metal particles in it lets it tend to stay in place under compression and pack-in rather than walking out like normal expoxies. On a fresh motor with a fresh tight fitting keyway, perhaps just some de-greased surfaces and a bit of loc-tite could be enough to stop the wiggle, which stops the wear. This type of situation doesn't seem like the keyway is inadquate for holding the torque, but rather just that the play/wiggle lets it wear. Loc-tite may be a $0.02 solution. For disassembly, just heat-gun the loc-tited area to about 250f and it softens up and slides apart.

Mark_A_W said:

Oh, Mr Physics...

Any idea if I could braze the (softer) key to the damaged (harder) keyway on the axle?


I don't think welding would work.

But if I file/machine the damaged load face flat, so the key makes good contact, then fill the gap on the other side with braze solder, it won't be able to rock back and forth.


It is the rocking that kills it.


I'll try and get a picture, if the new key is working loose already, as the keyway is too wide now.


Or I could get the keyway machined a touch wider, and make a key with a step in it. 5mm wide for the freewheel, 6mm wide for the axle. Then braze that in so it doesn't wiggle...

Click to expand...

If the axle is able to be removed to the point that nothing in the HAZ will mind seeing 800-900deg, brazing the key in the slot sounds like it would be simple and a solid fix.

Time to upgrade phase wires....


Yeah, I think brazing would be right, as it would flow into the gaps around the key.

Welding would just make a big mess.


And it would be painful, but I could get a brazed key out again, by repeated application of curse words.

liveforphysics said:

May save a bundle of money getting a set hardened over getting custom gears made.

The DLC coating is to act as a wet lube bath replacement on the hardened gear. It has some pretty amazing anti-wear properities, both on it's own surface, and the surfaces it meshes with.

Maybe a hardened tooth profile, micro-polished finish, then DLC coated. Here is a white paper about DLC as a replacement for lube on high pressure hardened surfaces here:


http://www.thefreelibrary.com/Scuffing+resistance+of+DLC-coated+gears+lubricated+with+ecological...-a0216178973


I would bet under $150 for the hardening, micro-polish, and DLC coating all combined. Maybe under $80 if you can find a single shop to do it all in-house.

For the sloppy keyway problem, perhaps assembly with a little metalized liquid epoxy? Something like the devcon for steel? It can do wonders for getting some more life out of worn splined connections, never seen it used with a keyway, but the multi-sized metal particles in it lets it tend to stay in place under compression and pack-in rather than walking out like normal expoxies. On a fresh motor with a fresh tight fitting keyway, perhaps just some de-greased surfaces and a bit of loc-tite could be enough to stop the wiggle, which stops the wear. This type of situation doesn't seem like the keyway is inadquate for holding the torque, but rather just that the play/wiggle lets it wear. Loc-tite may be a $0.02 solution. For disassembly, just heat-gun the loc-tited area to about 250f and it softens up and slides apart.

Click to expand...

Oh, missed this, sorry.

Yep, I might as well try it.

Edit: I can get a small tube of metal putty (epoxy) for $15 on ebay. I'll look at the hardware store tomorrow, but if they don't have any, I'll get it from ebay.

Thanks Luke and Mark for your work. If there is a reasonably cost effective method to improve the strength of the gears I'd definitely be very interested in getting them implemented somehow. Hopefully I will have some more time soon to focus a bit more on the motors and controllers.

Luke, the coating that you described, when I read through the link it also used oil and also all the gears were coated. Would the dry gears of the motor really be suitable for this treatment and I think it would get rather involved if you needed all 3 planet gears and the sun gear coated?

Oh, and apologies Mark for crashing in on your build thread.

No, it's ok, the motor is part of the build.


If we can get a $160 motor reliable at 1500w+ it would be awesome.

Well I personally don't think that around 1500W but not too much over that figure is a problem for the standard motor. I mean you look at the 1000W BMC and they advise not to run it over 1500W and that costs how many time more than the price of the MAC? I've had some thoughts on this and I was thinking maybe using the 350W motor in your build with the higher volts you are using may work out ok. It's a slower winding motor and obviously it's not gonna pull the current the 500W does, but at the voltage you are running, the low resistance of the 500W windings is not necessary IMO and means you are likely hitting current limit much of the time. I'm sure that at say a 30A current limit, the 350W motor will still pull all those amps. Anyway attached is some graphs of the 350W motor at 36 and 48V.

View attachment Hub Motor 36V350W Performance.pdf

View attachment Hub Motor 48V800W Performance.pdf

Yes, this is a 365 rpm at 36 volt motor.

I would prefer a 300 rpm at 36v motor.


I'm not on the current limiter as much as the ebikes.ca simulator predicts.

From the graphs 300rpm at 36V just so happens to be what the 350W is. They are also a few dollars cheaper and for you, in the interests of research, I'm sure I could sort 1 out for you at a very special price and arrange some spare's to go along with it. I haven't personally tried 1 yet but do have a few sat here. It's just such a PITA finding the right length spokes otherwise I'd get some put into rims...

I think you are likely to lose some top speed but it might not be so much. I think it's worth a try and if it gives the gears a bit less of a hard time they should stand up a bit better. I find the 500W with it's very low winding resistance can have a bit of a kick off the line. With the very high volts you are running I wonder if that contributing to putting the gears under undue stress. I know the BMC controllers implement a soft start. Like any gears it's the shock loading that will kill them more than outright power.

cell_man said:

Luke, the coating that you described, when I read through the link it also used oil and also all the gears were coated. Would the dry gears of the motor really be suitable for this treatment and I think it would get rather involved if you needed all 3 planet gears and the sun gear coated?

Click to expand...

The way I see gearing after years of sending every type of transmission and rear-end gear set made by Honda/Toyota/Nissan spinning in fragments down the street is simple. Make as few of changes as possible to get best results. If only 1 set is wearing, only alter the wearing set, then run it and see how it goes. Altering more things at a time tends to open a big can of worms that may not need to be opened. For example, the heat from coating the ring gear may cause aligment or other metalurgy problems, etc etc. 1-step at a time, test, if it doesn't solve it, look for the issue and try 1 more thing, test, etc.


It would be good to check if it's not the gear's fault at all, but the mounting of the gear flexing that causes the problem. This is an issue with many car rearends and trannys. The gears are great and pleanty strong, but the case flexes so the gears get tip-loaded, and of course abnormally wear and fail.

If you want to test this, apply a bluing compound to the teeth of the planets, put it together, run around for a few days with very low power, then disassemble and photograph the bluing compound wear marks. Then re-blue it, and run on high power for a bit, disassemble and take a peek.

If you notice the shiny marks where the bluing is worn away are further out on the tip in the high-power run than the low-power run, then you have gear center location problems rather than gear problems, and no amount of stronger gears will be much more than a band-aid fix until you can fix the gear center location.

Thanks Luke,

maybe I should start a new thread or something as I don't want to clog up Marks build thread. Please Mark or anyone else just say the word.

I do hear what you're saying about gearboxes flexing, as although I've never taken a transmission apart I was and still am very interested in High Performance Japanese Cars and the Skyline GTR in particular. Reinforced mounting plates was a popular mod in their gearboxes when you started pushing big power. However, with my limited knowledge of gears, I would have thought that the epicyclic gears is quite a different situation to the transmission in a car. With the epicyclic gears you have the input and output to the planets gears at 180degs and in the same plane, so I cannot understand how it would cause the mountings to be damaged.

How about this scenario. The single steel gear will not have the give that the other 2 nylon gears have and so it will take all of the strain when things are being pushed hard. This would result in uneven pressure on the gears and maybe cause some misalignment. I would have thought that it's very possible that the steel gear may have caused some of the other issues noted by Mark. The steel gear may have already damaged the gear assembly to such a point that when it was replaced with a nylon gear, it was already in bad shape and that contributed to the early demise of the nylon gears. I'd be much happier if the whole gear assembly was replaced to make the tests a bit more scientific. If it's just patched up and then fails again, we can't really be sure of the cause.

I also believe that some of the things that are very hard on geared hub motors could be addressed in Software in the controller. A soft start as well as some form of traction control that would prevent the wheel spinning up when it becomes unloaded would likely reduce the shock loading on the gears enough to improve reliability. It would just take some sort of algorithm in the controller that didn't allow acceleration beyond a preset point to avoid what is likely the biggest killer of these gears. The Mac is a 32 pole motor and with the right control you could very tightly control the max acceleration of the motor.

My freewheel/gear carrier doesn't have any wiggle in it now - I punched the pins, they are rock solid.


However when I ran one steel gear it wore badly, but the plastic gears didn't. So they were flexing.


3 steel gears, even with oil, is still too noisy for me.


Cell-man. I will PM you about a 300 rpm motor and some spares.

And LFP is right, for high power, the first thing that needs to be done is the key needs to be locktite'd in place.


I bought some metal putty to try and shore up my keyway. Just need to find some time to strip the motor....again..


Thanks guys

Mark

Well just give me a shout and I'll get 1 sorted for you. Don't have spares right now but could maybe take some out of a complete motor if I can't get some fairly quickly. If you need a new axle for your motor I should be able to arrange that too.

What you mentioned about the pins for the gears flexing could still be due to the fact a single steel gear was in there. With all 3 gears of the same material the forces on all 3 gears will be matched and shouldn't cause the gear assembly to be stressed in the same way. The single steel gear was likely taking all of the strain and likely resulted in a very different set of circumstances than if the gears were all matched. At lower rpm and power the steel gear maybe wouldn't cause such issues??

If there's ways the motor can be improved easily I'm happy to do the mods or request they are done at the factory although I do feel the standard motor with Nylon gears is more than up to the job of running a 25A controller at 36 or 48V. Thanks for taking the time to post your feedback in a nice clear way so that it can be acted upon and I'll do my best to make them as ES proof as possible

I've got the 500W motor apart that I've been using for about 8 months on and off at 25A, both 36 and 48V. The keyway looks fine. The additional shocks resulting from having that steel gear could have contributed to the keyway issue in your motor IMO. I'll have to strip down the single steel gear motor I've been using and see how that looks. I really think that having a motor with no damaged parts would be a good place to start just to take out the possibility of a damaged part causing other issues. I think the power levels should be reduced a bit too. After the motor has been hit with 2500W plus, it could still fail when running at a lower power at a later date.

Hi, I am Jack, engineer of MAC Shanghai. Thank you guys for posting your feedback on our motor. We will consider to use better material for the gears so that people could use our motor for higher power. We will also pay attention on other items you mentioned here. We hope we could supply a "perfect" motor to the market in near future. I will keep you updated.

Mark, we could send you one set of clutch and gear assembly for the replacement. Please contact me at jack@macmc.com.cn.

Thanks again.