dontsendbubbamail
10 kW
- Joined
- May 28, 2008
- Messages
- 718
swbluto said:
The shaft can be tapped out of the bell after you remove the two set screws.
Bubba
My belt broke - motor's timing pulley isn't secure
- Thread starter swbluto
- Start date
The shaft can be tapped out of the bell after you remove the two set screws.
Bubba
Ok, so I received the drill from Mcmaster-Carr and I tried to fill the bore of the timing pulley with the JB cold weld epoxy and then I "baked it" at about 160 degrees fahrenheit. After about 30 minutes, I pulled it out and then drilled it.
After drilling it, I put it on the shaft and then put blue lock tight on the set screw. As the blue-loctite went onto the shaft, it started to come into contact with the JB cold weld and that area came off, oddly enough. It might've been the set screw and the little ditch I made for it, who knows.
Anyways, as it fitted onto the shaft, it didn't bounce from side to side! It was pretty darn secure. However, as I started to rotate the motor, it was clear the timing pulley was still wobbling, though less exaggerated than before when the bearings failed. Also, it seems like the motor has an "inner-clickiness", though it seems like something you'd only hear if your ear was within 3 feet of the motor. Are normal bearings supposed to sound like that?
So, okay, I either have a bent shaft, an off-center hole and/or a bore that's not perfectly parallel to the timing pulley teeth. I thought that my drill press's stand was inherently perpendicular to the drill bit as that's integrated into the drill press. I think the machinist vice was laying perfectly flat to the stand, and the timing pulley was laying flat to the machinist vice. (I'm having doubts though, because I just removed a flat area of epoxy on the timing pulley that was getting a bit close to the motor mount and I wonder if resting the timing pulley on that side "flat" would've likely not been flat since the epoxy was uneven. Wait, if I didn't, then the other side which is definitely not flat would've tilted it anyways. Ok, so the bore is definitely not parallel to the timing pulley teeth.)
I've now deduced the bore isn't likely parallel to the timing pulley teeth. Well, let's examine the other possibilities. I measured the shaft from a known fixed point as I rotated the motor. The distance between the shaft and that fixed point didn't not significantly vary (it varied by less than .1 mm), so it seems highly unlikely the shaft is bent.
I then measured the distance between the outside and inside bore of the pulley on two opposite sides. My digital caliper reported 20 mm and 21 mm on the other side (It's actually less than that since my caliper has a fixed off-set that it apparently doesn't have the ability to zero out.). So, does anyone know how to get a fool-proof center hole? I tried by "aiming for the center" but apparently that resulted in in the below difference.
After drilling it, I put it on the shaft and then put blue lock tight on the set screw. As the blue-loctite went onto the shaft, it started to come into contact with the JB cold weld and that area came off, oddly enough. It might've been the set screw and the little ditch I made for it, who knows.
Anyways, as it fitted onto the shaft, it didn't bounce from side to side! It was pretty darn secure. However, as I started to rotate the motor, it was clear the timing pulley was still wobbling, though less exaggerated than before when the bearings failed. Also, it seems like the motor has an "inner-clickiness", though it seems like something you'd only hear if your ear was within 3 feet of the motor. Are normal bearings supposed to sound like that?
So, okay, I either have a bent shaft, an off-center hole and/or a bore that's not perfectly parallel to the timing pulley teeth. I thought that my drill press's stand was inherently perpendicular to the drill bit as that's integrated into the drill press. I think the machinist vice was laying perfectly flat to the stand, and the timing pulley was laying flat to the machinist vice. (I'm having doubts though, because I just removed a flat area of epoxy on the timing pulley that was getting a bit close to the motor mount and I wonder if resting the timing pulley on that side "flat" would've likely not been flat since the epoxy was uneven. Wait, if I didn't, then the other side which is definitely not flat would've tilted it anyways. Ok, so the bore is definitely not parallel to the timing pulley teeth.)
I've now deduced the bore isn't likely parallel to the timing pulley teeth. Well, let's examine the other possibilities. I measured the shaft from a known fixed point as I rotated the motor. The distance between the shaft and that fixed point didn't not significantly vary (it varied by less than .1 mm), so it seems highly unlikely the shaft is bent.
I then measured the distance between the outside and inside bore of the pulley on two opposite sides. My digital caliper reported 20 mm and 21 mm on the other side (It's actually less than that since my caliper has a fixed off-set that it apparently doesn't have the ability to zero out.). So, does anyone know how to get a fool-proof center hole? I tried by "aiming for the center" but apparently that resulted in in the below difference.
Attachments
TylerDurden
100 GW
The drillpress platform may not be level... many can be tilted on one axis. You can check yours with a machinist's square, or similar.
To get a centered hole, hit your mark with a centerpunch; perhaps also drill a small diameter pilot hole that the larger drill will have the tendency to align to.
To get a centered hole, hit your mark with a centerpunch; perhaps also drill a small diameter pilot hole that the larger drill will have the tendency to align to.
Okay, I knew how to use a center punch, but what I didn't know was how to find out where precisely to punch - which point is exactly the center? The "looks good enough" modus operandi was accurate only within .5 mm.
Anyways, I re-epoxied, re-baked, re-drilled and found out that putting over the empty-bottom part of the machine vice (So the drill could go cleanly through the pulley instead of hitting the iron bottom) was the reason why it didn't work the first time - the timing pulley inevitably wasn't level. So I restricted it to the flat parts this time and, oh dude, it was tight this time! I had to physically press the timing pulley onto the shaft in order to get it secured. It definitely wasn't wobbling. Initial hand testing looked good and the wobbling was drastically reduced, so all looked good.
Now that I've run in under load (by scooting with it), though, it seems like I hear some rubbing after when it seemed like there was a pretty healthy gap between the timing pulley and the motor amount. I'll have to investigate more indepth later... it seems like the belt tension significantly increased, so maybe it's "belt strain" noises? Or possibly ball bearings. I'll have to test it out later. I'm more curious to see if the epoxy is holding up.
For now, it seems like I'm having problems with the motor and ESC. I'll save that for another thread.
Anyways, I re-epoxied, re-baked, re-drilled and found out that putting over the empty-bottom part of the machine vice (So the drill could go cleanly through the pulley instead of hitting the iron bottom) was the reason why it didn't work the first time - the timing pulley inevitably wasn't level. So I restricted it to the flat parts this time and, oh dude, it was tight this time! I had to physically press the timing pulley onto the shaft in order to get it secured. It definitely wasn't wobbling. Initial hand testing looked good and the wobbling was drastically reduced, so all looked good.
Now that I've run in under load (by scooting with it), though, it seems like I hear some rubbing after when it seemed like there was a pretty healthy gap between the timing pulley and the motor amount. I'll have to investigate more indepth later... it seems like the belt tension significantly increased, so maybe it's "belt strain" noises? Or possibly ball bearings. I'll have to test it out later. I'm more curious to see if the epoxy is holding up.
For now, it seems like I'm having problems with the motor and ESC. I'll save that for another thread.
12p3phPMDC
1 kW
- Joined
- Mar 16, 2009
- Messages
- 462
Dude, you need lathe. You've got the ambition....but need better tools.
With a lathe, you can chuck the part up and bore the center out in two minutes.
Perfectly centered... Either that or take it to someone who has one already.
You can also dress up the face of you pulley and make it perfectly square to the pulley teeth.
$300 and up.
IT's about as cheap as it gets...you also need tooling....which adds up fast.
http://www.harborfreight.com/cpi/ctaf/displayitem.taf?Itemnumber=95012
You might find a mini lathe used for cheap with some tooling.
There's probably a community college that has a machining class..
Isn't the Seattle area like a machinists mecca? at least is used to be...
You can buy more scrap aluminum plate from Boeing than I could ever dream of using
at rediculous prices. THere's gotta be a hungry machinist there with a lathe...
try craigslist services ...
you can lead a horse to water... :|
With a lathe, you can chuck the part up and bore the center out in two minutes.
Perfectly centered... Either that or take it to someone who has one already.
You can also dress up the face of you pulley and make it perfectly square to the pulley teeth.
$300 and up.
IT's about as cheap as it gets...you also need tooling....which adds up fast.
http://www.harborfreight.com/cpi/ctaf/displayitem.taf?Itemnumber=95012
You might find a mini lathe used for cheap with some tooling.
There's probably a community college that has a machining class..
Isn't the Seattle area like a machinists mecca? at least is used to be...
You can buy more scrap aluminum plate from Boeing than I could ever dream of using
at rediculous prices. THere's gotta be a hungry machinist there with a lathe...
try craigslist services ...
you can lead a horse to water... :|
Don't worry. Just the advice I should find access to a lathe would be good enough - I didn't really know that lathes were used for centering, but it makes sense in retrospect. It appears that my university has a mechanical play-room of sorts, so I'll be investigating that when school starts. I know at least it has a mill (It's used for milling out circuit boards, funny enough) and I'm sure there's a lathe hanging around somewhere...
Anyways, I readjusted the belt tension so it's not so tight (but still fairly so), and took it out. It was practically silent! Well, it's not completely silent, but it makes a pretty adorable gurgling sound as it speeds up for some reason. Don't ask me. I'm pretty sure pedestrians won't notice "the sound" any more, except if they're fairly close, so I'm happy. Anyways, good news, the JB cold weld epoxy is still holding up, the timing pulley is pretty secured to the shaft and the wobbling is pretty minimal. There's still some wobbling for not being "perfectly center", but it seems perfectly acceptable (or at least much more acceptable than before).
For some reason, now that the rubbing has gone away, it doesn't "sound fast" and I noticed it doesn't feel fast. But yet I know that it at least must be as fast or faster as the current with the drive-train attached as the only load is lower meaning less power is wasted on drive-train losses. I think my drive-train is consuming around 2 amps of 30 volts, so it's wasting 60 watts or so. :| I'm not sure if that's typical. (The total input power is about 1200 watts, so I guess that's an efficiency loss of 5%? Does that sound typical?)
Miles said:
I had that suspicion. Okay, well, it peaks at about 1050 watts so that's about an efficiency loss around 6%. It seems like as it approaches the no-load speed, the output power starts to approach 0 watts but yet the drive-train still consumes ~2 amps, so that's about a drive-train loss of 100%.
(Or maybe 30% if you consider the torque to overcome the motor's cogging torque as part of the output torque and thus output power).
liveforphysics
100 TW
swbluto said:
With out having a center, you are wasting your time. Not checking machine setup is also wasting your time, and our time. Finding a center in a circle is very simple. The classic draw 2 intersecting lines with a 90/45deg square trick has been working well for the last 4,000 years or so.
liveforphysics
100 TW
swbluto said:Miles said:
I had that suspicion. Okay, well, it peaks at about 1050 watts so that's about an efficiency loss around 6%. It seems like as it approaches the no-load speed, the output power starts to approach 0 watts but yet the drive-train still consumes ~2 amps, so that's about a drive-train loss of 100%.
Unloaded drivetrain losses are useful only for knowing unloaded drivetrain losses... The fluid drag of the lubrication in the bearings and the wheel/chain/sprocket moving through air is about the only loss that doesn't vary directly with load.
liveforphysics said:
*Yawn*. I'm tired, so it's hard to think of tricks I've never used, despite their simplicity in retrospect. And, I did check my machine setup before my latest attempt which has had results that I've been liking so far, so I've achieved the goals of this thread. If reading of such a success story is wasting your time, feel free to choose to waste it.
liveforphysics said:
Sounds intuitive. Any idea what the magnitude of drive-train losses are under load? I did notice belt tension seems to heavily affect the unloaded drive-train losses (As confirmed by the greater torque as evidenced by how much more quickly it slows and the greater amount of force required to accelerate the wheel), which I'm sure makes up a significant part of loaded drivetrain losses, so I think it should at least give a general idea. However, I do see your implication that that the top speed can't be directly extrapolated from unloaded drive-train loss measurement.
liveforphysics
100 TW
The belt tension becomes a less signifigant loss as drive system torque loading increases. This becomes even more true in situations where mounting is less than rigid. Even 1mm of flex under load can take away all that increase in belt tension drag you measure when unloaded.
To give you an idea of drive system loss, it's generally close to a fixed percentage of the torque you are trying to transfer. The percentage depends greatly on bearings, alignment, vibration, belt materials, belt tooth to gear interaction and other things.
For a quick example with a Honda, an engine made 288bhp, then when put in the vehicle, with the trans/axle/bearing/tire losses, it layed 249whp. About ~13% drivetrain loss. Same engine gets boosted, and is making 575bhp, and when put back into the car, layed 506whp. ~12% drivetrain loss. These are at the same RPM range/speed, and unloaded drivetrain loss between them is identical, as it's the same drivetrain, the only difference is now the engine is twisting that drivetrain harder, which is why the drivetrain went from eating 39hp to eating 69hp.
To give you an idea of drive system loss, it's generally close to a fixed percentage of the torque you are trying to transfer. The percentage depends greatly on bearings, alignment, vibration, belt materials, belt tooth to gear interaction and other things.
For a quick example with a Honda, an engine made 288bhp, then when put in the vehicle, with the trans/axle/bearing/tire losses, it layed 249whp. About ~13% drivetrain loss. Same engine gets boosted, and is making 575bhp, and when put back into the car, layed 506whp. ~12% drivetrain loss. These are at the same RPM range/speed, and unloaded drivetrain loss between them is identical, as it's the same drivetrain, the only difference is now the engine is twisting that drivetrain harder, which is why the drivetrain went from eating 39hp to eating 69hp.
TylerDurden
100 GW
The v-shaped attachment of a machinist's square is for finding centers. The HF cheapie has SAE & metric scalesswbluto said:
http://www.harborfreight.com/cpi/ctaf/displayitem.taf?Itemnumber=92471
Robomaniac
100 µW
swbluto said:
I might be a little bit late on this one, but this is future reference for other builders.
This is how I shim mine.
My original post on my website http://www.jeromedemers.com/blog/2009/07/making-a-shim-for-the-motor-shaft/
The shim does work but are not the best. They wear out.
Not sure if that would of work in your case...
Can't wait to see a video of your scooter in action!
Thanks for sharing, Jerome!
This is an update post on my original method of filling the pulley with epoxy and then drilling it out precisely with a 10mm drill. It's gone about 100 or so miles, some of which were through rain storms, and I've removed the pulley from the shaft a couple of times (2 or 3 times) for various things, and it seems that the epoxy where it attaches to the shaft has worn down and it's now rubbing against the motor mount and going back to it's old behavior.
This suggests that either...
1) JB Cold Weld isn't the best epoxy for this purpose to use. Maybe I should look at some "metal" epoxies.
or
2) The properties of my shaft is too rough for epoxy, like how my drilled spot has a relatively "sharp" rim and that likely shaves a little bit of the epoxy each time the pulley is removed.
I think I'll look at "metal" epoxies in the near-future.
edit:
I searched amazon and didn't find many "metal" epoxies that were also rated. So, I looked at the jb weld offerings and it appeared the marine and normal epoxy were identical, while the "water weld" epoxy isn't and it also has a tensile strength of 6000 psi as compared to the normal epoxy's 3960 psi. I'm not sure if I'm concerned about tensile strength as I think the epoxy was being shaved off when the pulley was pulled off it, but being that it's also water proof, I don't think it would hurt to try. It also appears it has a "hardness" index and I think that might be what I'm looking for - it's rated "65 to 75" psi in hardness.
I looked for the hardness of steel, found a Rockwell C hardness of 40 for low-end carbon steel, I inputted that into this calculator at http://www.unified-eng.com/scitech/hardness/hardness.html, and it yielded 182,000 psi. Now, they say that corresponds to "tensile strength" (Is that the same thing as hardness?), but I think I can safely assume that pretty much all the relevant ratings of steel is much, much higher than epoxy and so epoxy might not be a permanent solution.
I'm thinking... for a permanent solution, I need to replace the timing pulley.
In doing so, I would need to enlarge the bore. And I think it would be nice to replace the shaft (or maybe necessary because the hub of the replacement doesn't seem large enough for a set screw and I couldn't use the existing hole for a roll pin as it's not perfectly perpendicular to the surface.), though it seems like I'll need to cut the shaft, make two flats for the set screws in the back, and make a groove for a C clip. I don't doubt the feasibility of the two flats (Though I haven't done that), but I do wonder if I could accurately make the groove for the c clip.
This is an update post on my original method of filling the pulley with epoxy and then drilling it out precisely with a 10mm drill. It's gone about 100 or so miles, some of which were through rain storms, and I've removed the pulley from the shaft a couple of times (2 or 3 times) for various things, and it seems that the epoxy where it attaches to the shaft has worn down and it's now rubbing against the motor mount and going back to it's old behavior.
This suggests that either...
1) JB Cold Weld isn't the best epoxy for this purpose to use. Maybe I should look at some "metal" epoxies.
or
2) The properties of my shaft is too rough for epoxy, like how my drilled spot has a relatively "sharp" rim and that likely shaves a little bit of the epoxy each time the pulley is removed.
I think I'll look at "metal" epoxies in the near-future.
edit:
I searched amazon and didn't find many "metal" epoxies that were also rated. So, I looked at the jb weld offerings and it appeared the marine and normal epoxy were identical, while the "water weld" epoxy isn't and it also has a tensile strength of 6000 psi as compared to the normal epoxy's 3960 psi. I'm not sure if I'm concerned about tensile strength as I think the epoxy was being shaved off when the pulley was pulled off it, but being that it's also water proof, I don't think it would hurt to try. It also appears it has a "hardness" index and I think that might be what I'm looking for - it's rated "65 to 75" psi in hardness.
I looked for the hardness of steel, found a Rockwell C hardness of 40 for low-end carbon steel, I inputted that into this calculator at http://www.unified-eng.com/scitech/hardness/hardness.html, and it yielded 182,000 psi. Now, they say that corresponds to "tensile strength" (Is that the same thing as hardness?), but I think I can safely assume that pretty much all the relevant ratings of steel is much, much higher than epoxy and so epoxy might not be a permanent solution.
I'm thinking... for a permanent solution, I need to replace the timing pulley.
In doing so, I would need to enlarge the bore. And I think it would be nice to replace the shaft (or maybe necessary because the hub of the replacement doesn't seem large enough for a set screw and I couldn't use the existing hole for a roll pin as it's not perfectly perpendicular to the surface.), though it seems like I'll need to cut the shaft, make two flats for the set screws in the back, and make a groove for a C clip. I don't doubt the feasibility of the two flats (Though I haven't done that), but I do wonder if I could accurately make the groove for the c clip.
gwhy!
100 kW
swbluto said:
I have cut C clip groves with a junior hacksaw, spin the shaft slowly and just use the hacksaw as if you were cutting the shaft. If you have a means of getting oversized hole exactly in the in the center of your pulley then you can use a hub configuration to re-attach to the shaft.
