AVIAN featherweight mid drive

neptronix said:
That is cool as hell. Too bad it's a one-off.
Thanks!
Actually two-off. The one on the bike, with 19:1 ratio, was built for APS 6374 outrunner and the one open with visible rotors, 13:1 ratio, was built for GNG motor AKA small block.
 

Attachments

  • GNG with 13-1 cycloid.JPG
    GNG with 13-1 cycloid.JPG
    32.8 KB · Views: 3,459
Ecyclist said:
I designed and built the hypocycloidal transmission based on three different Sumitomo transmissions that I bought and studied. Sumitomo makes industrial type of transmissions, which are . ...

Can you expand on that "Studied" bit?

What were your thoughts, what, why, how? What measurements were taken, etc. Seriously, but man that is awesome! Bought three, then .. hardcore stuff. amazing dedication and drive. With an actual budget, so not just dreaming and "I wonder" stuff. Actually doing it. Boggle.

Something like this:
1) so there I was, and then I had this idea..
2) So I bought three, and then
3) I tore them apart and measured ..
4) then I said "Screw that! I can do it better with x,y,z and then kill a,b,c because that is stupid.."
 
wow brilliant stuff, do You know what rpm the reduction can take? I am interested because i have laying arround a 8kw outrunner with 530 kv and the plan was to build a light bike ,your transmission would be ideal :p
 
bionicon said:
wow brilliant stuff, do You know what rpm the reduction can take? I am interested because i have laying arround a 8kw outrunner with 530 kv and the plan was to build a light bike ,your transmission would be ideal :p
Thanks!
530Kv x 48V (possibly more)=25440 RPM (possibly more) That spells P R O B L E M S.
I think that getting a motor with much lower Kv is a way to go. Personally, I wouldn't use a motor with Kv over 200. But that's just me.
 
RobertC said:
Ecyclist said:
I designed and built the hypocycloidal transmission based on three different Sumitomo transmissions that I bought and studied. Sumitomo makes industrial type of transmissions, which are . ...

Can you expand on that "Studied" bit?

What were your thoughts, what, why, how? What measurements were taken, etc. Seriously, but man that is awesome! Bought three, then .. hardcore stuff. amazing dedication and drive. With an actual budget, so not just dreaming and "I wonder" stuff. Actually doing it. Boggle.

Something like this:
1) so there I was, and then I had this idea..
2) So I bought three, and then
3) I tore them apart and measured ..
4) then I said "Screw that! I can do it better with x,y,z and then kill a,b,c because that is stupid.."
Sure.
First I would like to give credit to Tangent Dave. From what I know, he was the first one who used cycloidal transmission on e-bike.
I liked the idea but wanted something different, so at first I decided to buy industrial type transmission with a three phase motor.
My plan was to pull the tranny from the motor and hook it up with a brushless inrunner.
Unfortunately, these transmissions are kind of long and I would have to make a lot of changes to set it up the way I wanted.
I bought three different ones because I was hoping that one of them would fit the bill. That didn't work out as planned, but it gave me an opportunity to study differences between them. In one photo you see rotors that look about the same but work totally different.
The other photo is of a much older design cycloid. All made by Sumitomo. For me, it was like being a kid in the candy store.
 

Attachments

  • Sumitomo 2.JPG
    Sumitomo 2.JPG
    50.3 KB · Views: 3,409
  • Sumitomo 5.JPG
    Sumitomo 5.JPG
    55 KB · Views: 3,409
So what was it originally? Some kind of traction motor? It looks sturdy.

Sumitomo is a huge conglomerate, they do most everything.. I did study engineering in Uni, but, wow, this is completely beyond any math or fabrication skills I ever dreamed of having. Statics, dynamics, inertia, ok, cycloidal gears, a machine does that? huh..

Just wow..

The unit is metal, and I am guessing, but maybe 10-20 kw for the size?

Still, I have no idea how one would "evaluate" such a thing. I feel like a monkey "evaluating" a calculator. I do not know the right questions.
 
So here are photos of my newer tranny and other components including weights.
All in pounds. I could do it in kg but in the US we still use pounds.
If you need kg, just divide ounces by 16, add result to pounds and multiply by .4536.
Example: transmission is 1lb 13.2 ounces 1+13.2/16x.4536=.82782kg.
Kit (transmission braces and brackets) without the motor is 2.738lb.
Kit with motor is 4.63lb. or 2.1 kg. Like I said, featherweight. :D

@RobertC That ugly looking cycloidal transmission above is most likely from industrial fluid transfer pump.
 

Attachments

  • Cycloid 19-1 1.JPG
    Cycloid 19-1 1.JPG
    61.1 KB · Views: 3,330
  • Cycloid 19-1 4.JPG
    Cycloid 19-1 4.JPG
    35.5 KB · Views: 3,330
  • Cycloid 19-1 3.JPG
    Cycloid 19-1 3.JPG
    35.4 KB · Views: 3,328
  • Spider 3.JPG
    Spider 3.JPG
    44 KB · Views: 3,330
Brass gear/steel pin/ALU housing? Everything I wanted is in your excellent job.
How about the acoustic/vibration? I'm dying to see the actual working video.
 
The curtain's been lifted, the wizard's exposed! Fantastic job! Yes, please comment on the sound/noise level or even better some video please. This gives us new hope for big reductions.
 
Thank you guys!
Tranny doesn't produce vibrations and sound is on par with the motor. I'm happy with that, but not 100% satisfied. From what I read and what I experienced with Sumitomo cycloids, this type of transmission should be silent. However, in order to make it very very quiet I would have to beef up housing, and I don't want to add extra weight to it. I'll post video sometimes in the future. I'm still going to make a couple of minor changes, but, at this point, I'm going to focus more on the case for VESC and another one for batteries, and so on.
@Leeleeducati I read about your HPC Revolution XXX build, and I'm very impressed. You put tons of love into that project.
https://endless-sphere.com/forums/viewtopic.php?f=28&t=71051&start=325
It would be great to have someone like you next door, so I could bounce ideas back and forth. In the meantime, I will chisel away with help from the forum.
BTW: Can you guys help me with set-up and hooking up of a dead man switch to VESC? That is something I would like to have, but I'm not sure if I can do it with VESC. Keep in mind that I have a hall sensor and a throttle already connected, and they both work perfectly.
I would really appreciate it.
Thanks and have fun.
 
Hi guys,
I'm trying to figure out the best way to encase VESC and heatsink it at the same time. The problem I'm having is with capacitors. They are in my way. Would it be electrically and electronically sound to move capacitors a little away from the board and put the power switch between the board and capacitors? This way I could build a compact case for the VESC and house capacitors inside the battery case. Also, with that, I wouldn't have to worry about the pre-charge switch because capacitors would be connected to the battery pack in a semi-permanent way. :idea: What do you think :?:
 
Ecyclist said:
bionicon said:
wow brilliant stuff, do You know what rpm the reduction can take? I am interested because i have laying arround a 8kw outrunner with 530 kv and the plan was to build a light bike ,your transmission would be ideal :p
Thanks!
530Kv x 48V (possibly more)=25440 RPM (possibly more) That spells P R O B L E M S.
I think that getting a motor with much lower Kv is a way to go. Personally, I wouldn't use a motor with Kv over 200. But that's just me.

Bionicon, Take Ecyclist's advice. Think of it like this: Higher KV's = Higher wear on the transmission/drivetrain. You're using a transmission to lower the motors rpm so why not start off with a motor with the lowest KV possible for your application.
 
Ecyclist said:
Hi guys,
I'm trying to figure out the best way to encase VESC and heatsink it at the same time. The problem I'm having is with capacitors. They are in my way. Would it be electrically and electronically sound to move capacitors a little away from the board and put the power switch between the board and capacitors? This way I could build a compact case for the VESC and house capacitors inside the battery case. Also, with that, I wouldn't have to worry about the pre-charge switch because capacitors would be connected to the battery pack in a semi-permanent way. :idea: What do you think :?:
capacitors should be as close as possible (in vesc 6 its around 1-2mm from pcb with fets)
 
artteth said:
Ecyclist said:
Hi guys,
I'm trying to figure out the best way to encase VESC and heatsink it at the same time. The problem I'm having is with capacitors. They are in my way. Would it be electrically and electronically sound to move capacitors a little away from the board and put the power switch between the board and capacitors? This way I could build a compact case for the VESC and house capacitors inside the battery case. Also, with that, I wouldn't have to worry about the pre-charge switch because capacitors would be connected to the battery pack in a semi-permanent way. :idea: What do you think :?:
capacitors should be as close as possible (in vesc 6 its around 1-2mm from pcb with fets)
Yes, I kind of knew that they needed to be close. But why?
Anyways, in that case, every VESC controller v.14.2 is junk because capacitors on these VESCs are 2.5 to 3 inches away from the board connection.
I'm surprised they still operate. Maybe it is not as critical as some say.
I would like a second opinion please. :lol:
 
I don't like the idea of fitting the battery and controller together. I don't know your operation temps, but for all of the bikes that I have built, both controller and battery have periods of heat. Enclosing them together would only make it worse, unless the enclosure has an effective cooling system.

My controllers are always in full air flow, and batteries never insulated from the enclosure face that receives air flow. I know that some are enclosing them together without any problems, but they are building low power bikes with battery chemistries that are not likely to overheat.
 
MadRhino said:
I don't like the idea of fitting the battery and controller together. I don't know your operation temps, but for all of the bikes that I have built, both controller and battery have periods of heat. Enclosing them together would only make it worse, unless the enclosure has an effective cooling system.

My controllers are always in full air flow, and batteries never insulated from the enclosure face that receives air flow. I know that some are enclosing them together without any problems, but they are building low power bikes with battery chemistries that are not likely to overheat.
It looks like you misunderstood me. I'm planning on having the controller in a separate, heatsinked case. The other case, ventilated, will contain capacitors, battery and other good stuff.
Let's put it in a different way. Can I hook up capacitors to the battery so they will be always connected and charged? Will that affect the life of the battery or capacitors? Did anyone try that? Is there a thread about that?
Thank you for your help.
 
Capacitors should be as close possible to the Fets and they make no sense on battery side. Without caps evening the ripple current you probably will burn the fets on first loaded switching cycle.
 
Caps are used to stabilize voltage transients and should be close to the controller. I seem to recall that due to such transients destroying older VESC's (driver IC) the newer versions moved the caps closer. Someone, please correct me if I'm wrong.
 
Ecyclist said:
artteth said:
Ecyclist said:
Hi guys,
I'm trying to figure out the best way to encase VESC and heatsink it at the same time. The problem I'm having is with capacitors. They are in my way. Would it be electrically and electronically sound to move capacitors a little away from the board and put the power switch between the board and capacitors? This way I could build a compact case for the VESC and house capacitors inside the battery case. Also, with that, I wouldn't have to worry about the pre-charge switch because capacitors would be connected to the battery pack in a semi-permanent way. :idea: What do you think :?:
capacitors should be as close as possible (in vesc 6 its around 1-2mm from pcb with fets)
Yes, I kind of knew that they needed to be close. But why?
Anyways, in that case, every VESC controller v.14.2 is junk because capacitors on these VESCs are 2.5 to 3 inches away from the board connection.
I'm surprised they still operate. Maybe it is not as critical as some say.
I would like a second opinion please. :lol:

ok
try to remove caps and connect battery to vesc and your vesc will not work correct in best case or will die immediately or at calibration stage (when i try to power up without caps my vesc die immediately(i don't remember 100% but how i remember drv chip die )
if you want facts and number welcome to this video https://www.youtube.com/watch?v=x6lPdI9OVQg&t=1026s
 
Guys, thank you for your input. Again my idea was not about removing capacitors from the circuit but to move them closer to the battery and put the power switch between capacitors and controller. I have a good reason for that. But the idea might be just a brain fart. :D
Obviously, Vesc-6 was designed so the capacitors are very close to the board, but is that absolutely necessary?
For the last two days I read everything I could find about capacitors and I'm getting some ideas, but nothing solid yet.
Is anyone out there who actually did some testing? What about placing three large (or two larger) capacitors a little away from circuit and adding one more smaller capacitor directly to the board?
So far I see only NAYS.
Any YEA?
 
The farther the caps are from the FETs on the board, the more inductance and resistance there is between them, and the less the caps can do their job (which is to minimize the ripple and spikes at the FETs, because those can cause damage or failure, or at the least unexpected and unwanted behavior).

The less capacitance there is, and/or the higher the ESR of that capacitance, and/or how far away that capacitance is from the FETs, the worse the problem at the FETs is.

If you have an oscilloscope you can see it if you set up the probes in the right places. I'm not sure where those would be in your controller, but Arlo's controller threads, and probably some of the VESC threads, show various controller problems and scope displays and may explain where they were probing at and why.


You can always experiment, just be aware of the possible consequences. :)
 
Ecyclist said:
So far I see only NAYS.
Any YEA?
There is also a tread, where best EE brains of ES explain, why cables from the battery to a controller should be as short as possible. The problem is the inductance created in them and the longer they get, more capacitance is needed on Fets side to tame the induced spikes.
So YEA, theoretically, if your cables between battery and Fets are super short, twisted, battery has a very low resistance, and good controller topology to minimize inductance you may not need caps at all. I may be wrong though.
 
Thank you guys for your help.
Special thanks goes to amberwolf who put me on the right track and pointed my attention to Arlo's controller thread. If you are into controllers, you should definitely check it out.
I think that I have enough info so I can get going with my project again.
I ordered an extra VESC and a bunch of capacitors for testing. Eventually I will play with that.
In the meantime, with a board I have, I will move capacitors closer to FETs and put them at 90 deg. to the board. I will design a case with heatsink around that. It is not going to be pretty, but it should make it better than the original and the box will be shorter. The caps will be exposed similarly to Castle controllers.
I will go with conventional pre-charge circuit and try something different later. Maybe in a next build.
 
Back
Top