3D printed in-frame Motor

[Föppel]

This one here was also a bit of a pain to design...

it is the adapter for the freewheel on the side of the 1. reduction (Belt-Side)

"in the old days" i ordered and built 5 LR small block kit (-Bikes) for me and friends, and had some spare parts extra that i wanted to use...
But these were made from aluminum and after some time with 2000-3000 watt, they showed some play, resulting in wobbling Belt-Pulleys... so this one is made from D2 tool-steel, that should work.

I also tried to design it in a way so that it has a maximum possible contact area with the axis. It will be held in place by a 4mm key and 2 C-clips ( i think that is what they are called). Not 100% sure if these clips will hold it in place... Another thing that remains to be seen. I had all the metal parts made by PCBWay.com and they came out great.
To be transparent here, PCBWay is a sponsor of my YouTube channel, just to say...

In the background you can also see a blurry 3D-printed prototype of this part

 

[DanGT86]

Trial and error. That's what I was afraid you were going to say. I always ask and it seems most people who have had success with DIY pulleys had to fight it like that.

I was reading up on the various belt profiles and found some things that didn't play well with my perfectionist and anxious nature. I believe the amount of built in clearance actually changes based on the tooth count so technically its not as easy as just measuring the tooth and wrapping that negative shape around a pulley. I say technically because realistically people like you are out there successfully riding while I am obsessing about tolerances.

I am a bit of a specialist when it comes to letting the perfect be the enemy of the good. Results in more thinking than doing.

 

[DanGT86]

This one here was also a bit of a pain to design...

That's a nice looking part. The D2 will absolutely be better than aluminum long term. The aluminum slowly deforms. As for the retention on the shaft I have yet to find anything that works as well as 2 set screws 90 deg apart. One over the keyway to push down on the key and one 90 from that on the shaft.

Seems like even with a very tight tolerance fit between the shaft and bore, there will still be slight movement that wants to hammer against the key and deform the part. Most slip on fits retained by C-clips are hardened steel and splines.

That's not to say it wont work at the levels of torque you are messing with. Again, trying it beats thinking about it forever.

 

[Föppel]

Trial and error. That's what I was afraid you were going to say. I always ask and it seems most people who have had success with DIY pulleys had to fight it like that.

Thank you for writing that It took me about four full days to get it "right", and I despaired twice a day while working on it... I thought I was doing something wrong, and I didn't like the fact that I had to make adjustments to the measurement data "by hand", which didn't make sense to me in theory and only seemed to make kind of sense from a visual comparison perspective.

I think your theory makes sense, that tolerances could change depending on the size/number of teeth. That would also explain the various adjustments I had to make. Of course, manufacturing also plays a role. My 3D printer is pretty accurate, but it can't produce the part with 100% accuracy... I guess i also had to compensate for that to some extent in the CAD drawing.

"...That's not to say it wont work at the levels of torque you are messing with."
oh i hope so... and dont even know yet what power i will apply. Hoping for 2000w, but also like the idea of using 20 or 40V Powertool-batterys mounted on top of the motor assembly, that way, it would be light and very easy and fast to install/remove the whole thing... but i guess then it would be limited to something about 500w

 

[Nix Liou]

Interesting, have you had any issues with the plastic pulley breaking under peak power? I suppose the large contact area makes up for it being plastic.

You probably don't need to print the pulley entirely out of solid ASA though, it seems that in 3d printing most of the strength comes from the skin, ie wall count, and infill mostly helps to prevent buckling and keep it rigid. I know at least for 3d printed gears making sure the teeth and a bit of where they connect is solid is generally all that is needed.

I have had reasonable success with printing large fairly solid ASA with my P1S, by the high tech solution of wrapping the printer in towels and blankets and letting it sit for 30 minutes with 100C bed temp before printing... and letting it cool down fairly slowly too.

 

[Föppel]

Interesting, have you had any issues with the plastic pulley breaking under peak power? I suppose the large contact area makes up for it being plastic.

Hi Nix, yes the Pulley should be ok, like you mentioned i also guess that it is because the large contact area.
Right, i dont know if it was necessary to print it solid, maybe 5 or even 4 or 3 walls would be enough, but when slicing the model, i could not see that much infill anyway, so i made it solid, just to be sure

by the high tech solution of wrapping the printer in towels

Nice! i heard of this special technique hey when it works, it works!
i did the same in regard of letting it cool down very slowly, i run he print over night, and opened the printer next morning.

 

[Föppel]

i did not write about the bike much, so here is the reason why i wanted it to be a Kona Cowan (again)...

This one i built around... 2015 or so?
at some point i changed the wheels to 24" and it felt amazing. Not sure if it is the geometry (i guess) but i also really like the "kind of BMX"-look of it.

i frame-mounted a LR small block kit but had several problems with it, regarding things that bent and also the chain (not installed on this foto) that runs to the BB was way to long for my likings.

The Bike (frame) itself was/is solid, i rebuilt it several times with BMC and Bafang motors, and it was the perfect size for my Girlfriend, she loved it, but do to reasons, we had to sell it at some point.

Now 10 Years later, i wanted to build the exact same Bike but better for her again, and here we are.

 

[Föppel]

As for the controller, I wanted to try out an ESC and decided on the FS75100 from VESC. It's rated at 100 A... I think that's phase amps and the battery output is around 30 A (?).

In any case, it looks nice and is small enough for my intended use.
i hooked it up to the motor and to my Laptop by USB, run the calibration, and everything seems to work.
I tested it with a 52V Battery from EM3EV. But I'm not sure yet what kind of battery I'll ultimately use. Maybe I'll even try power tool batteries, as they are small and light... 2x 20 volt packs in series... Not sure if that would be enough power for me...
any suggestions?

 

[Nix Liou]

As for the controller, I wanted to try out an ESC and decided on the FS75100 from VESC. It's rated at 100 A... I think that's phase amps and the battery output is around 30 A (?).

In any case, it looks nice and is small enough for my intended use.
i hooked it up to the motor and to my Laptop by USB, run the calibration, and everything seems to work.
I tested it with a 52V Battery from EM3EV. But I'm not sure yet what kind of battery I'll ultimately use. Maybe I'll even try power tool batteries, as they are small and light... 2x 20 volt packs in series... Not sure if that would be enough power for me...
any suggestions?

I don't recommend using power tool batteries. They are usually pretty shit, as they only need to survive short bursts usually. Often they rely on charge balancing only, or sometimes no balancing at all (milwaukee).

Most controllers are rated for phase amps (2.5x or 3x) of the battery amps, so that is probably correct, but you can usually find some reported number for it in a datasheet somewhere. Looks about right for the size though.

Luckily 30A is not that much power relatively, so you can find a few options at least. What's the target voltage here? It would be based on motor kv, the reduction stages, and the target speed. Or however much speed you think the power output of the motor could handle.

If you want something sleek and made to fit the bike, you may want to look into custom builds for the triangle. It's hard to tell the exact dimensions but I think if you don't need a ton of capacity, a suitable pack could be designed to fit in the remaining space around the motor. It might be pretty close though. A 48v 5Ah pack using tabless power cells like the RS50 or EVE 50PL would have just 13 cells, and be somewhat similar to using 2 power tool batteries, but could have actual discharge protection from a BMS.

Of course, 30A from a 5Ah battery will give you around 10 minutes of full throttle power, which may be limiting. A pack like the slim pack offered by affordable ebikes would probably fit well on the under side of the downtube for example, without getting in the way of the front tire, especially with smaller tires on it. That may be the simplest option.

 

[DanGT86]

Looking forward to seeing how this outrunner in a box design works out for you with heat management. I always wonder if the efficiency of a small high RPM motor is lost when the weight of the enclosure is added up. Doesn't take much additional weight to equal something like the lightning-rods LMX style inrunner.

 

[A-DamW]

Cool, did you already built something with the 6368 Motors? i chose this kind of Motor as there are so many of these available from different manufacturers. Even saw some more expensive ones from Flipsky (?) with lower KVs that would maybe even be a better fit.

I have not done anything with them yet, I got them used in new condition for dirt cheap, and how many motors are enough?

N+1 of course!

 

[Föppel]

If you want something sleek and made to fit the bike, you may want to look into custom builds for the triangle. It's hard to tell the exact dimensions but I think if you don't need a ton of capacity, a suitable pack could be designed to fit in the remaining space around the motor. It might be pretty close though. A 48v 5Ah pack using tabless power cells like the RS50 or EVE 50PL would have just 13 cells, and be somewhat similar to using 2 power tool batteries, but could have actual discharge protection from a BMS.

Of course, 30A from a 5Ah battery will give you around 10 minutes of full throttle power, which may be limiting. A pack like the slim pack offered by affordable ebikes would probably fit well on the under side of the downtube for example, without getting in the way of the front tire, especially with smaller tires on it. That may be the simplest option.

Hi Nix, and thanks for your thoughts and Ideas. Much appreciated!
Yes you are right, the Powertool batteries would have been a bad idea. Sleek, but to much of a compromise in regard of power.

I ordered a BMS for 10-14S and plan to re-use some of my Samsung 30Q to make a 10-14S 1P pack. I want it on top of the Motor assembly, together with the Flipsky controller. So that everything can be removed and mounted with only the 4 "Butten-Bolts" (?).

if money would not be a problem, i would go with some high discharge 21700 5Ah. That would be optimal, but i will use what i have for now. For longer rides i can hook up a bigger pack when it is needed, but i want something as small and light as possible for this this to begin with.

 

[Föppel]

As you can see, i already hove some holes at the top of the Motor assembly. They are meant to hold the Controller and Battery... in some way that i have to figure out

 

[Föppel]

so here we go, starting with the battery for this Bike, or should i say for this Motor-Kit… because as i wrote before, i want both, the bike going 1500 - or even 2000 Watt, and the battery beeing as small and lightweight as possible. Here are the components i bought to make a small battery (probably 10s1p) that should fit on the motor kit in the bikes frame.

BMS with bluetooth for 10-14s battery

quite small, and 40A discharge will be way to much for 1p, but that should work.

then i also wanted to have the charger as small and light as possible, so i found this cheap one here:

… that one comes with a 5.2mm charging plug

so i bought some of these female 5.2mm plugs to solder one to the BMS and integrate the plug on the side of the battery.

i started drawing some battery Ideas in CAD, but had not much time to really work on it, e.g. printing some prototypes.
In the upcoming weeks i will work on that and post some pictures of the progress.

It is clear that the motor will not run on more than... lets say 250 watt or so with this kind of battery, but that is okay for its intended use (of this battery, not on the motor kit itself). For the testing with higher power, i will probably just make a XT60 connector on the Controller to hook up a 14S5p battery that will go in my backpack for the fun rides

the reduction of the motor is more set to a 10s pack than to 14s, but i think (hope) that in the highest gear it should be fine

Thanks for the good tips and thoughts Nix

 

[Föppel]

Some progress;

i was finally able to get the ACS Freewheel removed from the old and meshed up Axle-Adapter (made from Aluminum) and install it on the new one that is made from D2 Tool steel and is also way beefier and with smaller tolerances, to get rid of the wobble i had when using the Adapters from GNG and LR.

everything fits well, and is now secured by C-Clips with almost zero play

then i drilled out the holes of the 3D printed parts that hold the motor kit to the frame by these "Nautic push button pins" .
The tolerances on these 3D printed parts were a bit to tight for getting the Pins in and out easy. Not sure if - and what kind of - oil i could use there with the PETG Material used there, if you know this, please let me know.

Now i can start working on the controller mounting, soldering cables and connectors and all that funny stuff

As i am pretty sure that i want to test the Motor Kit with an 52V battery from my backpack, i thought to just mount the controller on top of the Motor-kit, something like this.

Just for interest, does someone know where to get Freewheels like the (old) ACS ones i am using? I think they are not being made anymore... White Enos would be an alternative, but way to expensive. Are there any alternatives that are "okay"?

 

[Föppel]

Hey Folks, i worked the whole day on it and had some progress with the 2. stage reduction (219 chain on the right side)

installed a chain tensioner and a bigger 219 sprocket. i changed from a 73T to a 91T, that should help with the overall reduction, and just looks sick

Chain line is not set yet, but looks promisingly straight. First tried to use a Motorcycle tensioning pulley, that didn fit well, so I had to design and print a new one. The right side and inner core of it are made from PCTG, a Filament that has very good layer adhesion, so it should be plenty strong for this.
The outer Sleve and right side are made from a 70D TPU - quite soft to make it a bit quieter. At least that is what i think/hope for.

With the tensioning mechanism more teeth are engaging at the small sprocket, which i good.

The tension arm is from a E-Scooter (when i remember correctly. you can open it and change the position/tension. Found that on Aliexpress some years ago. Finally i can use it.
The tensions arm is screwed in with a 6mm screw on the Aluminum Part that holds the whole Motor-Kit to the down tube. I had to cut some threads to mount it.

The Aluminum Plate that is holding the sprockets is 2.5mm thick. i hope it will withstand 2-2.5 KW.
I dont have the right screws yet you see there are a lot missing, that is just for checking if everything align as it should. will buy the missing screws for the final assembly on Monday

here you can see a bit better how this tension pulley is made. the two parts (PCTG Core and TPU Seve) are press-fitted.
In the middle are two bearings and the axle is a stainless steel 8mm Bolt

Thats it for today, i wish you all a nice weekend

 

[scianiac]

Ha, just a little while ago I did a similar TPU printed 219 chain tensioner. I just made the whole part from TPU and forced it over a 608 bearing. I think the TPU was harder than 70D, maybe in the 85-90 range but I don't know for sure as it's just some old spool with the label long gone. I also put a ridge down the center so the tensioner contacts both the plates and the rollers with chamfers all around to hopefully help keep it centered and reduce wear, although I don't expect much wear. I think it's a great application for TPU, much better than many of the tensioner adaptations and designs I've seen.

PETG is quite resistance to most oils so you can use about anything, some classic silicone grease works well for lots of plastics though.

 

[Föppel]

Great, so i will use some silicon grease for the PETG parts Thanks for your answer Scianiac, did it all work out with your TPU pulley/tensioner?

Though the whole motor kit feels sturdy, i am a bit nervous as the moment of truth gets near - maybe everything explodes or melts at first try

 

[scianiac]

Great, so i will use some silicon grease for the PETG parts Thanks for your answer Scianiac, did it all work out with your TPU pulley/tensioner?

Though the whole motor kit feels sturdy, i am a bit nervous as the moment of truth gets near - maybe everything explodes or melts at first try

Haven't got around to finishing the bike yet, still needs some wiring, controller and encoder install and some assembly before a test drive with a test battery and then I need to actually build a battery powerful enough to wring the neck of this LR Big Block.

Having used a lot of printed parts for high stress Ebike parts I don't see anything wrong with your design, if anything it's probably built stronger than many parts I've used as I have a habit of trying to push the strength as close as I can. The number of things I've had break is pretty low but I have had things break. Many are due to crashes or freak accidents. I broke a mud guard in a very smaller crash but reprinted it in TPU and I see no reason it would ever break now. I blew up two chain tensioners (ones mounted on to the derailleur mount) but one was a crash and one was I guess you could say was half my fault and half a freak accident since I didn't think to put deflectors around the wheels and a stick got ejected from the wheel and perfectly stuck between the gear and chain causing it to lock up and blow itself up. Both times though I was able to zip tie them back together and ride out.

I broke several experimental flexure 3D printed suspension seatposts but those were always pushing the material properties and some were my fault, the nylon one for instance broke because I forgot to water pack it. I don't normally print PETG as like in one of those seatposts I swear PETG has just terrible crack propagation properties and just likes to give way if any stress riser pushes it too far.

I print most things in ABS because the properties are basically decent at everything and god is it cheap. Then exotic stuff for parts that have very high local stress or need extreme properties. Like my bar end kncukle guards are printed from Lexan so it will actually take the impact of me clipping a tree. Nylons for parts that need to flex a little. If I'm building something for somebody else it's all super over built in carbon fiber nylons and stuff but when I make something for my own projects it's like eh that is probably enough perimeters and I'll just print it in this $8/kg ABS and it'll be fine, I can make a stronger one if it breaks.

I think the part that makes me not worry about printed parts my chainring to cassette adapter that is just one printed part and so far has held up fine in spite of the very low gear ratio and pretty good torque from the LR small block.

 

[Föppel]

Haha that all sounds very interessting, do you have like a build threads of the LR Stuff? I Loved riding the Small block kits, built 5 of them for me and Friends, and very interested to see these builds of yours.

About the Materials; you seem to have a lot of experience in 3D printing and especially in the usecases of 3D printing in E-Bike parts. Mind if i ask you some questions from time to time when i run into problems with upcoming projects?

I started using PETG just because i had enough lying around at the time i started. I Think it will be strong enough to not bend, since in my design, it is pretty wide in the areas where the forces apply, that should be ok. Will see how it behaves when the motor is running hot

Thanks for your reply, wish you a nice sunday

 

[scianiac]

Someday I'll get around to doing a build thread on my other bikes, my small block bike has has the most development and I'm still tweaking and adding things to squeeze out a little extra performance.

I would love to talk 3D printing, hit me up whenever. I was thinking why exactly I use PETG so infrequently when it seems to be the go to material now and weirdly I think the answer is when I started 3D printing it wasn't very common. I haven't been printing since the beginning but it's been a long time. PETG is very strong and with good settings it's layer bonding is really impressive but that impact performance and honestly printability. Man does ABS just print so easy as long as you have a heated chamber. I also like that it's cheap and you can perfectly bond it with ABS cement which makes making very large parts like battery cases easy and allows for more design freedom since you can print something as two parts and make it one without a weak point. It doesn't get gummy and annoying to work with if you cut, drill, sand, tap, etc. And ABS layer bonding is actually pretty good if you print it hot enough and in a chamber.

Yeah I think heat is the only thing I would be worried about but that depends how hard you push the motor. I was going to make some suggestions but upon closer inspection you've already done most of them so the only thing I have to add is if the motor runs hot you can put a centrifugal fan mounted to the back of the motor that will suck air through the vents in the motor bell. Probably want to put a cover on that and the clearance to the chain will be tight but probably can fit something in there.

 

[Föppel]

Someday I'll get around to doing a build thread on my other bikes, my small block bike has has the most development and I'm still tweaking and adding things to squeeze out a little extra performance.

I would love to talk 3D printing, hit me up whenever. I was thinking why exactly I use PETG so infrequently when it seems to be the go to material now and weirdly I think the answer is when I started 3D printing it wasn't very common. I haven't been printing since the beginning but it's been a long time. PETG is very strong and with good settings it's layer bonding is really impressive but that impact performance and honestly printability. Man does ABS just print so easy as long as you have a heated chamber. I also like that it's cheap and you can perfectly bond it with ABS cement which makes making very large parts like battery cases easy and allows for more design freedom since you can print something as two parts and make it one without a weak point. It doesn't get gummy and annoying to work with if you cut, drill, sand, tap, etc. And ABS layer bonding is actually pretty good if you print it hot enough and in a chamber.

Yeah I think heat is the only thing I would be worried about but that depends how hard you push the motor. I was going to make some suggestions but upon closer inspection you've already done most of them so the only thing I have to add is if the motor runs hot you can put a centrifugal fan mounted to the back of the motor that will suck air through the vents in the motor bell. Probably want to put a cover on that and the clearance to the chain will be tight but probably can fit something in there.

Absolutely, everything you said...
I only did the HTD5 pulley from ABS, yes heated chamber is a must. At least for bigger and strong parts. What Printer are you using?


About Motor cooling, what you wrote is exactly what is planned i will start without fan/turbine for some test rides, and design/install them later. Thought of add a fan blade ring to the outside of the motor housing. As this will touch the hot motor, i thought of using PET CF. No Idea how hot it will get ( 100°C + i guess) and there would also be the centrifugal force this part has to withstand... Any suggestion for best Material for that?

The Dyson fan on the rear of the motor will be added last (hardest for me to design this, never draw a spiral fan before), together with the motor plate. These are planed to print from PA6 GF.

 

[scianiac]

I've got 3 old heavily modified printers, Pegasus 10in and 12in bed slingers two from a brand that no longer exists and one an I3 clone from geeeeeeeetech. A lightly modded Kingroon KLP1 that along with the 3 older printers is in a custom build heated chamber. The old printers I run at 70C but only 60C on the KLP1 to keep the toolboard alive although it's now actively cooled so maybe it can take more. Then the printers that get most of the work are two Qidi Plus4s and a snapmaker U1, all lightly modded. More insulation, chamber heater tweaks, etc. U1 has the control board mounted outside the chamber so I can run the inside around 60C when needed. Old printers don't get used much, some having not been on in many months and they have and will continue to be used for weird experiments and such. What you running?

PET-CF will probably be fine for the fan, it may get a little warm but it's also actively cooling itself.

 

[Föppel]

Haha, that is quite a fleet
I use a Qidi Q2 for the more "technical" Filaments. The P2S and a Elegoo CC2 for the lighter Stuff... And there is a A1 Mini that lives in my Campervan Build so i can print things in the woods

 

[scianiac]

Haha, that is quite a fleet
I use a Qidi Q2 for the more "technical" Filaments. The P2S and a Elegoo CC2 for the lighter Stuff... And there is a A1 Mini that lives in my Campervan Build so i can print things in the woods

Oh interesting, I have been really considering getting a Q2. A new version of a product I make requires really long print times so might be tight on throughput and it seems like a good option. My reluctance though is mostly due to the mixed bag of experience with the Plus4s, they work well now but required a lot faff to get them here. The thing I really want to know that I have not seen answered in any reviews is does the new nozzle load cell bed sensor actually work well and more specifically does work when the chamber is hot? I quickly gave up on using the bed probe offset on the Plus4 because it was just a flawed design and didn't work at all when the chamber was hot. It seems like many designs of bed probe have issues with drift or just not working when inside a heated chamber.

One of my Plus4s basically only prints nylon, if you have the right setup carbon fiber nylon is actually pretty easy to print and is a great all around high performance material. I use a lot of the polymaker PA612-CF, well priced and very strong, good all around properties for nylon. Not too flexy like the PA6 can be although you trade impact performance. Super strong with good layer bonding. Go with a PA6-CF if you need some more impact resistance. The real trick with any nylon though is finding a good bed, in the old days I used phenolic garolite beds which work amazing but also a bit too well as you tend to destroy the beds. For modern magnetic sheets though the two best options I've found are G10 (I use one from west3D) or BIQU Glacier beds both with good adhesives. Best adhesive I've found so far is PVP-K90/K30 in Isopropyl. The PVP works ok on PEI needs a lot of pads, brims, etc.