GNG, 1000W 48V BB-drive, $400
- Thread starter spinningmagnets
- Start date
kfong
100 kW
tri-lobe said:
http://www.endless-sphere.com/forums/viewtopic.php?f=28&t=44258&p=644365&hilit=backyard+millyard...midmount#p644365
christerljung
100 W
Report from todays riding:
Secondary idler has been further tested(see above). No problem so far. Belt seem to withstand heavy torque now. 72v and great accelleration. Back chain (the standard bicycle chain) is jumping coggs under heavy load and high speed/low gears. Maybe because the chain is new and cassette is a bit worn.
Anyway im so happy that the belt actually seem to handle the 72 volt stress with this setup. At least for a 30minute trip.
But how will it stand in the long haul?
Secondary idler has been further tested(see above). No problem so far. Belt seem to withstand heavy torque now. 72v and great accelleration. Back chain (the standard bicycle chain) is jumping coggs under heavy load and high speed/low gears. Maybe because the chain is new and cassette is a bit worn.
Anyway im so happy that the belt actually seem to handle the 72 volt stress with this setup. At least for a 30minute trip.
But how will it stand in the long haul?
full-throttle
1 MW
Did you remove the spring from the primary idler?christerljung said:Did some testing with a second idler today.
christerljung
100 W
There was no spring in my kit. I think they have shipped three diferrent versions from GNG.
crossbreak
1 MW
Miles said:
Thanks for the correction miles, assuming no amp limit, you are absolutely right. It's sad that the stock belt drive isn't strong enough to lift this treasure. The 1/1.4 was something i used to calc with, but it does not make much sense i understand now. It would be also nice to know the best reduction ratio for efficiency. But for this we would need to know the efficiency map of the motor. Gaining the best possible efficiency is very important when using battery power IMO. If i'd know someone who owns a dyno i would try to make an efficiency map.
christerljung
100 W
Did more 2:nd idler testing today. 30minutes@72v mixed offroad, hilly and flat, no belt problem, check vid:
http://www.youtube.com/watch?v=m2pxP2FJ7EE&feature=youtu.be
Closeup video on 2:nd idler:
http://www.youtube.com/watch?v=_oUJSAn1oGA
Im thinking about doing this idler setup more serious. Right now im using a worn out wheelbearing from a pw80. It´s 12mm wide and the belt is 15mm.
I´m considering this one, 16mm wide
http://www.ebay.co.uk/itm/4200-2RS-Series-Rubber-Sealed-Double-Row-Deep-Groove-Ball-Bearings-Choose-Size-/220953699049?pt=LH_DefaultDomain_3&var=&hash=item3371dde2e9
Or maybe no RS? rolls easier?
or maybe a roller one:
http://www.ebay.co.uk/itm/SKF-4202-ATN9-ROLLER-BEARING-/180778585160?pt=LH_DefaultDomain_3&hash=item2a173e4c48
I´m really not a mechanics expert so what do you guys think?
Here is an overwiev of the current setup, maybe move the mounting point somewhere for best performance?:
http://www.youtube.com/watch?v=m2pxP2FJ7EE&feature=youtu.be
Closeup video on 2:nd idler:
http://www.youtube.com/watch?v=_oUJSAn1oGA
Im thinking about doing this idler setup more serious. Right now im using a worn out wheelbearing from a pw80. It´s 12mm wide and the belt is 15mm.
I´m considering this one, 16mm wide
http://www.ebay.co.uk/itm/4200-2RS-Series-Rubber-Sealed-Double-Row-Deep-Groove-Ball-Bearings-Choose-Size-/220953699049?pt=LH_DefaultDomain_3&var=&hash=item3371dde2e9
Or maybe no RS? rolls easier?
or maybe a roller one:
http://www.ebay.co.uk/itm/SKF-4202-ATN9-ROLLER-BEARING-/180778585160?pt=LH_DefaultDomain_3&hash=item2a173e4c48
I´m really not a mechanics expert so what do you guys think?
Here is an overwiev of the current setup, maybe move the mounting point somewhere for best performance?:
Whiplash
1 MW
- Joined
- May 10, 2010
- Messages
- 2,906
I agree, the Idler on the tension side would need to be quite strong, but by doing it that way, you MAY eliminate some drag compared to having the belt make a radical wrap on the slack side..... What kind of top speed on the flat are you seeing with the 72V? How many watts under load?
christerljung
100 W
It´s 50kph top speed with or without the second idler. I havent bought a cycleanalyst yet, have no idea about wattage used so far, sorry.
I tested mine today at 36v and 25 amps using a KU93 9FET controller. It had massive torque - about equivalent to a geared hub motor at 40 amps or more. I also tried it at 15 amps with a 6 FET controller and it gave much less toorque. With 15 ampps it went up our 15% test hill slowly in first gear, but with 25 amps the motoor didn;t even slow down as i changed up thhrough the gears to fourth.
Does anyone knoow the amp rating of the motor and standard controller? I know it's 450w, but that would be 9.5 amps, which seems a bit low.
Does anyone knoow the amp rating of the motor and standard controller? I know it's 450w, but that would be 9.5 amps, which seems a bit low.
evrunna
10 µW
The standard controller is 22 amps well thats what printed on my controller but i think Fullthrottle stated the same when he received his one.
Vince
Vince
spinningmagnets
100 TW
A good method of raising the amp limit on the controller was posted a while back by someone else. The common method of just adding solder to the shunt to raise the amp limit is rough and inaccurate, and its also easy to raise the amps too much, leading to too much heat.
I recall in the post, someone wrapped the shunt at one end with bare copper wire for just a couple wraps, then soldered only the wire wraps. Then he ran the controller to see how many more amps it put out under load. This told him how many more amps he would get per wire wrap. He monitored temps on his commute, and kept adding to the shunt until the temperatures were as high as he was willing to risk.
I recall in the post, someone wrapped the shunt at one end with bare copper wire for just a couple wraps, then soldered only the wire wraps. Then he ran the controller to see how many more amps it put out under load. This told him how many more amps he would get per wire wrap. He monitored temps on his commute, and kept adding to the shunt until the temperatures were as high as he was willing to risk.
wavezz2k
100 mW
christerljung - Your posts are great, thanks for sharing the helmet mount. gives a very good idea of the standard performance.
Awesome. Idler geometry- my day job is mechanical and structural systems in aerospace. Although unconventional, if the two idler were symmetric and tied at the top of the post over the belt 'cross strap' (furthest away from the motor) then the separation thrust would be taken out in the strap & self canceling except for drag loads, and thus would not be trying to bend the bolt studs. This would maximize the teeth on the drive pulley, and provide higher tensions that would otherwise with these rather small diameter studs, which now take the kick load in bending back to the motor plate. Drive wear would improve with a reduced angle. Just a thought, I'll try to get a sketch uploaded, and no, I'm not going to CAD it, sorry. Great thread, and really appreciate everyone's posts. Super video, again thanks.
Awesome. Idler geometry- my day job is mechanical and structural systems in aerospace. Although unconventional, if the two idler were symmetric and tied at the top of the post over the belt 'cross strap' (furthest away from the motor) then the separation thrust would be taken out in the strap & self canceling except for drag loads, and thus would not be trying to bend the bolt studs. This would maximize the teeth on the drive pulley, and provide higher tensions that would otherwise with these rather small diameter studs, which now take the kick load in bending back to the motor plate. Drive wear would improve with a reduced angle. Just a thought, I'll try to get a sketch uploaded, and no, I'm not going to CAD it, sorry. Great thread, and really appreciate everyone's posts. Super video, again thanks.
The idlers seem to be such a waste of energy, though. I think it might be better to just use the one on the tensioned side as a snubber.
wavezz2k
100 mW
Idler concepts -
Top figure left, belt tension resultant 'double arrow head' resolves into the idler thrust (b) and
2nd component (a). Don't care about (a) load it self cancels both sides of the idler. (b) Load
can be quite high to get adequate belt tension. 'Strap concept', top right, shows with two symmetric
idlers connected with a 'strap' would react the (b) thrust load, allows high tension without over stress the idler mount bolt.
'Strap, (#2)' would actually be a revised pulley cover plate water jet cut, plus tensioner hardware (below).
Middle page - current idler has two small bearings with single strut, bolted to motor case and pulley cover plate (2).
Better approach is single, larger roller bearing with two struts, so the center of the belt doesn't crease.
Bench test has had small bits of belt wearing off. It would also lower the rolling friction.
The bottom right view shows elevation view of two idler bearings with top cross strap and obround slot holes,
drive bracket, and tension screw.
Using the two symmetric idlers, it would be easy with standard shop tool and steel tube stock,
to make a mount which also has includes a high force driver tensioner. This drawing needs minor tweaks, but hopefully,
conveys the concept.
(rev, edit)
Tension bolt (4) rotates free in (3), pulls second idler axis on right closed (draws plate 2 from right to left), decreasing the spacing of
two idlers. This is one concept, certainly could be better approaches.
My GNG motor, currently, is down. LiFePO batteries arrived yesterday on the plus side. Likely a hall sensor issue.
I am going to use a HK Turnigy amp meter not a CA ($20). Don't need it, CA is certainly cool though.
Top figure left, belt tension resultant 'double arrow head' resolves into the idler thrust (b) and
2nd component (a). Don't care about (a) load it self cancels both sides of the idler. (b) Load
can be quite high to get adequate belt tension. 'Strap concept', top right, shows with two symmetric
idlers connected with a 'strap' would react the (b) thrust load, allows high tension without over stress the idler mount bolt.
'Strap, (#2)' would actually be a revised pulley cover plate water jet cut, plus tensioner hardware (below).
Middle page - current idler has two small bearings with single strut, bolted to motor case and pulley cover plate (2).
Better approach is single, larger roller bearing with two struts, so the center of the belt doesn't crease.
Bench test has had small bits of belt wearing off. It would also lower the rolling friction.
The bottom right view shows elevation view of two idler bearings with top cross strap and obround slot holes,
drive bracket, and tension screw.
Using the two symmetric idlers, it would be easy with standard shop tool and steel tube stock,
to make a mount which also has includes a high force driver tensioner. This drawing needs minor tweaks, but hopefully,
conveys the concept.
(rev, edit)
Tension bolt (4) rotates free in (3), pulls second idler axis on right closed (draws plate 2 from right to left), decreasing the spacing of
two idlers. This is one concept, certainly could be better approaches.
My GNG motor, currently, is down. LiFePO batteries arrived yesterday on the plus side. Likely a hall sensor issue.
I am going to use a HK Turnigy amp meter not a CA ($20). Don't need it, CA is certainly cool though.
Attachments
spinningmagnets
100 TW
To anyone here with experience using sensorless controllers, since we have high RPMs with this motor..."should" it work well with sensorless?
adrian_sm
1 MW
wavezz2k said:
Load (b) will be different for each side of the belt, when the drive is applying torque. So your strap (2) will try to move.
wavezz2k
100 mW
"'So your strap (2) will try to move."" This is true, if it were just a strap.
Implementation must use the current pulley cover plate, where the 'strap' is simply incorporated to the water jet template. As such, the 'strap' is actually grounded to the motor face with bolts, as in the current setup for the cover plate. Idler's are kept from walking, by the cover plate. 'Strap' is just to convey the idea. Location (1) on Strap, partially cut off in the phone capture image, shows the bolt to 'ground'. As aside, the 'obround' should only be on the left side, not at both idler bolts - as shown, screw tensioner would not work... my assertion is it's a non trivial exercise to determine the difference in belt tensions on one side versus the other at various power and system drag and different idler tensions, so it might take a design iteration. Dependent on how hard you ride, may not be needed at all. The 72v rides are looking pretty cool though, and this belt 'skip' appears to be the weak link. I think the efficiency gets hurt by deep included angle, generates belt heat, but not as much from drive tension, that's just a hunch. Roller bearings handle high thrust with low drag routinely.
Regarding Dyno,we could get a good idea of efficiency by using CA (or amps) with a Hub Tap, like this guru, whom is nearby silicon valley - might be talked into run the curves for this drive; ref: http://mrbill.homeip.net/hybridBike.php#efficiencyCurves
Implementation must use the current pulley cover plate, where the 'strap' is simply incorporated to the water jet template. As such, the 'strap' is actually grounded to the motor face with bolts, as in the current setup for the cover plate. Idler's are kept from walking, by the cover plate. 'Strap' is just to convey the idea. Location (1) on Strap, partially cut off in the phone capture image, shows the bolt to 'ground'. As aside, the 'obround' should only be on the left side, not at both idler bolts - as shown, screw tensioner would not work... my assertion is it's a non trivial exercise to determine the difference in belt tensions on one side versus the other at various power and system drag and different idler tensions, so it might take a design iteration. Dependent on how hard you ride, may not be needed at all. The 72v rides are looking pretty cool though, and this belt 'skip' appears to be the weak link. I think the efficiency gets hurt by deep included angle, generates belt heat, but not as much from drive tension, that's just a hunch. Roller bearings handle high thrust with low drag routinely.
Regarding Dyno,we could get a good idea of efficiency by using CA (or amps) with a Hub Tap, like this guru, whom is nearby silicon valley - might be talked into run the curves for this drive; ref: http://mrbill.homeip.net/hybridBike.php#efficiencyCurves
adrian_sm
1 MW
wavezz2k said:
The difference in belt tension is simply a function of the torque.
eramachinery
10 µW
- Joined
- Oct 3, 2012
- Messages
- 5
Need Help from you guyz ^^?
i wanna to ask is the crystalyte controller 48v 45A controller with sensor can be use to the gng motor brushless motor ?
how about 72v 45 A controller ?
thanks
i wanna to ask is the crystalyte controller 48v 45A controller with sensor can be use to the gng motor brushless motor ?
how about 72v 45 A controller ?
thanks
wavezz2k
100 mW
D8VEH-
"25 amps using a KU93 9FET controller ..massive torque .. didn't even slow down as i changed up through the gears to fourth" . Could you share more data on your ride, sounds great. Were you using CA or ammeter, or relying on controller rating, do you have a auto fuse inline, did you check temps? How long was test climb, sufficient to get to steady state temperature, did controller or motor seem hot? Thanks for your post.
"25 amps using a KU93 9FET controller ..massive torque .. didn't even slow down as i changed up through the gears to fourth" . Could you share more data on your ride, sounds great. Were you using CA or ammeter, or relying on controller rating, do you have a auto fuse inline, did you check temps? How long was test climb, sufficient to get to steady state temperature, did controller or motor seem hot? Thanks for your post.
I finally got mine installed on the bike and double check Chinese QA. I had adjust the belt cover for more clearance to get rid of the metal to metal scraping noise. I also had to take out some shimmies from the tension gear with the spring on it, for it to line up with the crank. I also added a washer and loosen up the screw, so the tension gear actually moves along as it should. No pics yet, but will post when the the rest of the bike is complete. I'll install the CA with the new external shunt on it and see what it does.
I'm waiting on disk brakes to arrive before I hurt myself.
BTW, mine is just bare metal and arrived within 5 business days.
I'm waiting on disk brakes to arrive before I hurt myself.
BTW, mine is just bare metal and arrived within 5 business days.
Similar threads
