While the CC controllers seem to be undoubtedly better than cheaper ESCs for startup, I think the startup syncing problem is a limitation of sensorless brushless controllers. The controller needs to know which phase to apply power to and the only way it can determine that is from the back-EMF from the motor, which isn't being produced below a certain RPM. With a sensored motor, the controller knows where the motor is at all times, even at a dead stop, so there's no guessing involved at startup.
Grinhill's Medium-power RC-Motor Hardtail build
- Thread starter Grinhill
- Start date
And, ahhhhhh, here comes a person with actual personal experience with a CC controller.
Yes, CC controllers do start hard, there's no doubt about that. But, I experienced a 1/2 second delay where it stutters in guessing which way the motor goes from an absolute stop, so it's not as instantaneous as a sensored set-up but when it finds it, it definitely does lay down the power. If your bike is going like 1 mph, it has absolutely no problem syncing with the motor and bursting into action, so just push the bike forward a little and you shouldn't experience any problems with starting.
Yes, CC controllers do start hard, there's no doubt about that. But, I experienced a 1/2 second delay where it stutters in guessing which way the motor goes from an absolute stop, so it's not as instantaneous as a sensored set-up but when it finds it, it definitely does lay down the power. If your bike is going like 1 mph, it has absolutely no problem syncing with the motor and bursting into action, so just push the bike forward a little and you shouldn't experience any problems with starting.
Grinhill
10 kW
Joepostal said:Hey Grin
Where did you get the fan for your motor? I have an old broken drill I was thinking I would tear apart.
MMM HOCKEY
Joe
Yep, that's the idea. Mine was off a cheapo jig-saw. Some drilling and dremelling to modify it. It's hard to say how much of a difference it makes.
Grinhill
10 kW
matt_in_mtl said:
Great idea, thanks for the info. There has been a few thoughts along those lines from the high-power crew on other threads.
Gwhy, here's a description of the sort of sync issues I have:
Light loading (wheel off ground) - never lose sync, no matter what I do.
Standing start level ground - I think this can be done, but I prefer to avoid those huge starting currents.
Pedal start, level gound or uphill - very gradual throttle increase works fine, rapid throttle increase loses sync.
Running at medium speed (say 30 km/h), then accelerate to high speed - often loses sync, worse at higher speeds and higher loads. Back off throttle immediately and increase throttel at a slower pace will get me up to top speed.
gwhy!
100 kW
swbluto said:
For my application moving the bike is not a option it needs to come to life with a twist of the throttle and nothing else.
this is more the use of what these esc were designed for
I think this can also be done but will depend on how the good the firmware works in the esc (hardstart setting like in the CC controllers) and yes if sensored contollers/motors were used the initial jitter can be over come but does that jitter really matter as long as its consistent as it will only happen from a dead stop and may only last a fraction of a second
I think this may be a firmware problem just the fact that the changing loads are far greater that would be seen in a model ( why I think a boat or car controller may be a better option and should be much better at starting up under a higher load ) , Richards limiter may fix this because of a couple of reasons not just the current limiting . I have just made one of richards limiters to play with on the cheapo controllers that i have I will let you know what the out come is.
matt_in_mtl
100 mW
- Joined
- Aug 13, 2008
- Messages
- 49
This has likely been outlined somewhere else, but:
A sensorless BLDC driver depends on BEMF to detect which phases of the motor to energize. The BEMF is a result of the magnet moving in relation to the motor coils, causing a changing magnetic field. This means that the motor must be turning above a certain threshold before the controller will work properly.
Below this threshold, the controller has no feedback on true motor position, instead it energizes the phases in sequence, with a higher current than should be required to turn the motor, and slowly brings the motor up to speed. This is the same method used to control a stepper motor. The open loop torque will never be as high as what would be available running closed loop due to a number of factors. This process is very inefficient, but should not be required for very long. On an RC airplane, the load on the motor is smaller at lower rpm, so the controller can be conservative and use a fairly low current during this open-loop period. On your bike, in order to get that large mass moving, the torque required would be much higher. It may be possible to increase the open-loop current high enough to get reliable starts, although it will likely heat up the motor/controller and be very inefficient. So, as far as I understand, the problem will never go away completely, but it may be possible to make it manageable.
There is a method of detecting motor phase position without sensors, it has been referred to as the 'ping' method. The idea is that a coil's inductance will be higher when it is in a stronger magnetic field, so by energizing a phase, and analyzing how it reacts, it is possible detect where it is within 60deg electrical. I have not heard of any hobby esc's that use this method. I am not even sure if an RC esc has the required hardware.
I hope that was sufficiently clear. Personally I am interested in modifying an RC esc to provide current limiting, analog throttle, and integrate hall sensors into an RC motor. This would provide high power, full power starts, without toasting the controller/motor if the motor stalls.
-Matt
A sensorless BLDC driver depends on BEMF to detect which phases of the motor to energize. The BEMF is a result of the magnet moving in relation to the motor coils, causing a changing magnetic field. This means that the motor must be turning above a certain threshold before the controller will work properly.
Below this threshold, the controller has no feedback on true motor position, instead it energizes the phases in sequence, with a higher current than should be required to turn the motor, and slowly brings the motor up to speed. This is the same method used to control a stepper motor. The open loop torque will never be as high as what would be available running closed loop due to a number of factors. This process is very inefficient, but should not be required for very long. On an RC airplane, the load on the motor is smaller at lower rpm, so the controller can be conservative and use a fairly low current during this open-loop period. On your bike, in order to get that large mass moving, the torque required would be much higher. It may be possible to increase the open-loop current high enough to get reliable starts, although it will likely heat up the motor/controller and be very inefficient. So, as far as I understand, the problem will never go away completely, but it may be possible to make it manageable.
There is a method of detecting motor phase position without sensors, it has been referred to as the 'ping' method. The idea is that a coil's inductance will be higher when it is in a stronger magnetic field, so by energizing a phase, and analyzing how it reacts, it is possible detect where it is within 60deg electrical. I have not heard of any hobby esc's that use this method. I am not even sure if an RC esc has the required hardware.
I hope that was sufficiently clear. Personally I am interested in modifying an RC esc to provide current limiting, analog throttle, and integrate hall sensors into an RC motor. This would provide high power, full power starts, without toasting the controller/motor if the motor stalls.
-Matt
See, also, here: http://www.endless-sphere.com/forums/viewtopic.php?f=1&t=9735matt_in_mtl said:
matt_in_mtl
100 mW
- Joined
- Aug 13, 2008
- Messages
- 49
Miles said:
Ah.. that is the one company that I had heard of using the method... and they are located just down the road from me(Montreal) in Ontario! Cool Video, I am surprised more people are not talking about this method.
-Matt
gwhy!
100 kW
Thanks Matt for this detailed description and it makes a lot more sense than the way I try to explain things 
gwhy!
100 kW
I have done much testing now with the cheapo esc on a bike and all I can say is they seem to be reliable but terrible at starting up, once they are running I cant fault them, they do sometimes loose sync when running and I have concluded that this is due to the freewheel/drivetrain slack causing slight irregular loading on the motor when excelerating, I tried a fixed freewheel ( so to speak ) and it was perfect but defo not no good for my purpose because of the startup. I dont think I will be spending £120+ for a CC controller on the off chance that it will start up any better. I have ordered a cheapo boat specific esc to try. If the boat esc still suffers from startup problems I will be going down the route of making it a sensored motor with a sensored controller... :?
Luke:
if you read this please can you tell me if you had any startup problems with your hobbycity 100a hv controllers before you blew them up. cheers.
Matt:
How do the CC startup from a standing start? for example if you we stood on you ripper stationary balanced and give it a handfull would it be 100% at spitting you off the back of the bike ( so to speak ) or Would the CC be able to a walking pace wheelie from a standing start.
Luke:
if you read this please can you tell me if you had any startup problems with your hobbycity 100a hv controllers before you blew them up. cheers.
Matt:
How do the CC startup from a standing start? for example if you we stood on you ripper stationary balanced and give it a handfull would it be 100% at spitting you off the back of the bike ( so to speak ) or Would the CC be able to a walking pace wheelie from a standing start.
MitchJi
10 MW
Hi,
http://endless-sphere.com/forums/viewtopic.php?f=28&t=8160&p=140186&hilit=future#p140186
gwhy! said:
http://endless-sphere.com/forums/viewtopic.php?f=28&t=8160&p=140186&hilit=future#p140186
fechter said:
gwhy!
100 kW
Thanks MitchJ for the heads up. I have read all of that thread at some point but must have forgot this comment from Richard. I will have to wait and see what the outcome is with this cheapo boat controller that I have ordered and make a decision on that outcome, its just a very big price difference, the controller that is currently on my prototype bike worked out at just £8.00 ( £19.00 delivered ) and works and would be ok for a general purpose e-bike with the few mods that I have made to it but not good enough for my purpose because of the torque limiting startup that can not be disabled on this £8.00 esc
Grinhill
10 kW
Hi all
I thought I would give you an update with some performance figures after a few weeks of commuting.
I'm only riding to work two or three days a week, but that's way better than none.
I am finding that I use less Ah than my predictions from swbluto's simulator, and I think the main reason for this is that I'm probably pedalling a bit more than estimated, and the bike seems to be a little quicker going uphill than I expected, which keeps the motor efficiency higher.
For the journey to work, I use a little over 2 Ah, and about 3.5 Ah for the return journey (5 Ah capacity, recharge at work).
I spend most of my time between 30 and 40 km/h (19 to 25 mph) mostly using light pedal assist, heavier assist uphill to keep speed up.
Some sample power figures from the Cycle Analyst:
Flat ground, pedalling, 40 km/h (25 mph) - 360W ...uses 9 Wh/km (14.5 Wh/mi)
Flat ground, no pedalling, 40 km/h (25 mph) - 550W ...uses 13.8 Wh/km (22.1 Wh/mi)
5% grade, pedalling, 35 km/h (22 mph) - 900W ...uses 26 Wh/km (41 Wh/mi)
11% grade, pedalling, 30 km/h (19 mph) - 1100W ...uses 37 Wh/km (59 Wh/mi)
I thought I would give you an update with some performance figures after a few weeks of commuting.
I'm only riding to work two or three days a week, but that's way better than none.
I am finding that I use less Ah than my predictions from swbluto's simulator, and I think the main reason for this is that I'm probably pedalling a bit more than estimated, and the bike seems to be a little quicker going uphill than I expected, which keeps the motor efficiency higher.
For the journey to work, I use a little over 2 Ah, and about 3.5 Ah for the return journey (5 Ah capacity, recharge at work).
I spend most of my time between 30 and 40 km/h (19 to 25 mph) mostly using light pedal assist, heavier assist uphill to keep speed up.
Some sample power figures from the Cycle Analyst:
Flat ground, pedalling, 40 km/h (25 mph) - 360W ...uses 9 Wh/km (14.5 Wh/mi)
Flat ground, no pedalling, 40 km/h (25 mph) - 550W ...uses 13.8 Wh/km (22.1 Wh/mi)
5% grade, pedalling, 35 km/h (22 mph) - 900W ...uses 26 Wh/km (41 Wh/mi)
11% grade, pedalling, 30 km/h (19 mph) - 1100W ...uses 37 Wh/km (59 Wh/mi)
Grinhill,
I keep meaning to tell you how much I like your build. Actually, it's one of my favorites on here right now. I love it's simplicity. If I could figure out a good way to hide the belt and pulley I would probably go the same route.
Have you had any problems with the magnets on your motor yet? Because of all the other posts on here I assumed everyone with a HobbyCity outrunner would have issues with them.
I keep meaning to tell you how much I like your build. Actually, it's one of my favorites on here right now. I love it's simplicity. If I could figure out a good way to hide the belt and pulley I would probably go the same route.
Have you had any problems with the magnets on your motor yet? Because of all the other posts on here I assumed everyone with a HobbyCity outrunner would have issues with them.
RWP
1 kW
Hi Grinhill,
The schematic diagram showing the contactor circuit you asked about for the hard tail Cortina can be found here:
http://endless-sphere.com/forums/viewtopic.php?f=28&t=10428&start=30
The schematic diagram showing the contactor circuit you asked about for the hard tail Cortina can be found here:
http://endless-sphere.com/forums/viewtopic.php?f=28&t=10428&start=30
Grinhill
10 kW
Todd:
I had two magnets come loose early on, luckily no destruction. I used epoxy to repair. The problem I noticed was that they were originally only glued on the edges with a black flexible adhesive, by running the adhesive down the gap in between two magnets. Therefore there was no glue on the back of the magnets. I used glue on the back in my repair. Was it done that way for cost saving, ease of manufacture, or to allow for thermal expansion????
Roy:
Thanks very much for that, I had already spotted it. Very beefy contactor for the $$$.
I had two magnets come loose early on, luckily no destruction. I used epoxy to repair. The problem I noticed was that they were originally only glued on the edges with a black flexible adhesive, by running the adhesive down the gap in between two magnets. Therefore there was no glue on the back of the magnets. I used glue on the back in my repair. Was it done that way for cost saving, ease of manufacture, or to allow for thermal expansion????
Roy:
Thanks very much for that, I had already spotted it. Very beefy contactor for the $$$.
Grinhill
10 kW
Motor Cooling:
The other day when I checked the motor temp after a hard ride uphill, I noticed that the aluminium angle was even hotter than the motor casing. This suggests that it was doing a great job of removing heat, however it had nowhere to go after that, since it is bolted to two 3mm thick plastic spacers instead of directly to the mounting bracket.
Today I replaced one of the spacers with a heatsink which I salvaged from an old power supply, comprising two thin pieces of aluminium sandwiched together with a fin arrangement formed by folding.
I am sure this will be much more effective than the fan. Should have done it ages ago. Should have listened to liveforphysics, guru of cooling.
The other day when I checked the motor temp after a hard ride uphill, I noticed that the aluminium angle was even hotter than the motor casing. This suggests that it was doing a great job of removing heat, however it had nowhere to go after that, since it is bolted to two 3mm thick plastic spacers instead of directly to the mounting bracket.
Today I replaced one of the spacers with a heatsink which I salvaged from an old power supply, comprising two thin pieces of aluminium sandwiched together with a fin arrangement formed by folding.
I am sure this will be much more effective than the fan. Should have done it ages ago. Should have listened to liveforphysics, guru of cooling.
Hi Grinhill,
Thanks for the great build thread. It's good to see how others are going about building their bikes.
I'm planning a build araound the same motor that you have used on your bike. I'm wondering if I could get the bolt pattern dimensions for mounting the motor from you? I want to build a bracket out of heavy gauge sheet metal that will fit inside the frame triangle. I hoping I can design it before I go out and start buying components.
Cheers.
Thanks for the great build thread. It's good to see how others are going about building their bikes.
I'm planning a build araound the same motor that you have used on your bike. I'm wondering if I could get the bolt pattern dimensions for mounting the motor from you? I want to build a bracket out of heavy gauge sheet metal that will fit inside the frame triangle. I hoping I can design it before I go out and start buying components.
Cheers.
Dear Grinhill, thanks for the ongoing performance updates !
An inrunner? for a motor shell heat-sink...of course!
My memory is bad, does Grinhill use an outrunner? are the $400 Astro motors inrunners? I am still adapting my bike, and haven't purchased the electric drive parts yet. I made my MTB a semi-recumbent, and I'm looking at making it a long(er) tail with saddle-bags. I'm still not certain how I'll electrify it yet, but the Grinhill set-up is in the top 3 for consideration.
My memory is bad, does Grinhill use an outrunner? are the $400 Astro motors inrunners? I am still adapting my bike, and haven't purchased the electric drive parts yet. I made my MTB a semi-recumbent, and I'm looking at making it a long(er) tail with saddle-bags. I'm still not certain how I'll electrify it yet, but the Grinhill set-up is in the top 3 for consideration.
Grinhill
10 kW
mclovin said:
There are 4 mounting bolts size M4 on a 30mm PCD (pitch circle diameter). This works out as 4.2mm diameter holes in a 21.2mm x 21.2mm square pattern. Allow a centre hole for the shaft of approx 15mm. Some short countersunk M4 screws are supplied with the motor.
Your bracket should allow for a 50mm diameter motor plus a few extra mm clearance. However, if you ever want to go to a larger motor in the future (e.g. 63mm diameter) it would be a good idea to allow clearance for the larger motor if possible.
