My RC Mid Drive with Single Stage Reduction

Kepler said:
Make sure you have set the App to ADC and UART. Also make sure you have Baud rate set to 115200.
Thanks! I had the VESC set to ADC. Changed it to ADC and UART, and now I have data. Now I need to go for a ride!
 
c_saayman said:
What Android app are you using? Are you able to connect with the VESC app? It is very easy to use and just works. My Metr App also connects and works, but I haven't really looked at it yet. I've done nothing special.
I should have said "using the Metr app on my Android phone". I'm just planning to collect data, not change parameters on the fly (for now anyway)...
 
The Modes feature on the Metr app I think is one of its best features. I use modes all the time.

In wet or damp weather I use reduced power (to stop slip due to being on a friction drives setup) I also remove the regen in damp weather due to lack of traction.
 

Attachments

  • Screenshot_20190808-132335.png
    Screenshot_20190808-132335.png
    76.3 KB · Views: 4,644
Kepler said:
Hats off to the analogue approach but maybe consider going down the Arduino path. Opens up endless possibilities to how you control the drive. You don't need to be a programmer either with the block programming solutions now available.

Thanks! I do agree that the analogue approach these days are very primitive. I had a look at the Arduino when you first mentioned it on this forum, but first impressions was it's complicated and will slow my build down. Definitely the way to go and when I get time I will start the research. For now I just want to finish and get on the bike. Once I start using it I will get a better understanding of what type of control and power I need. Power cut-out brakes may also be required as you suggested. Will have to wait and see.
 
So I took the bike for its first test run. The best part of the bike is that it still has the feel of my original bike, just a bit heavier.New Rebel chainring.jpg I had a pedal assist, with trottle, Bafang BBS02, Norco MTB and still have a torque assist Bafang mid drive MTB, so I think I can make a fair comparison. The first thing that I need to mention is that you have to be fit, or have some level of fitness, to use this single stage MTB. With the Norco MTB I could easily throttle on tough sections when I get tired and with the torque assist bike I can just level up and it gets very easy.

With this new bike if you want power you need to spin the chain ring fast to get the motor working. It reminded me of the spinning gym classes when I though people were crazy torturing themselves like that. I don't think this is a bad thing. I can only get better.

With my gear ratio not being so good as FZBob (10T/90T - 32T-50T) I need to pedal a bit fast to get the same power. I also have a smaller battery, 40A max BMS, so I am limiting my battery current to 27A, motor to 40A. With the settings FZBob is using he must be smiling all the way. Very good idea to limit the max RPM, thanks Bob. When the motor kicks in and I've gone over the hill I was struggling to keep up with paddling.
I added a reed switch to my analogue CCT to cut out the motor when I pull the front brake in just a little bit. I nearly knock out my knee cap in the computer room while testing the bike. I quickly realised that the flying pedals was dangerous if you don't focus. It also came in handy on my first ride.Reed switch.jpg The green power button and the red cruise button is very close to each other. A couple of times I have locked the red button when I only wanted a quick burst, with the pedals going out of control. Killing the motor with the brake helped. This should get better once I get more familiar with the controls.Push button.jpg I also think maybe I will only need two power levels, not three. There is not enough difference between the three levels. I was cruising in the low level on easy flat bits, just because I am unfit. With the technical steep climbs I was using the max power. The motor definitely helps a lot. I manage to go to the top of a track that I only visit with the torque assist ebike. I have to admit I stopped about five times to catch my breath.

Overall I like the bike. There are lots of options/settings/improvements to play with. It will keep me busy and get me fit.
 
c_saayman said:
So I took the bike for its first test run. The best part of the bike is that it still has the feel of my original bike, just a bit heavier.

With this new bike if you want power you need to spin the chain ring fast to get the motor working. It reminded me of the spinning gym classes when I though people were crazy torturing themselves like that. I don't think this is a bad thing. I can only get better.

With my gear ratio not being so good as FZBob (10T/90T - 32T-50T) I need to pedal a bit fast to get the same power.
Nice looking conversion! How did the ISCG mount work out?

I like your battery holder, very clean integration. Do you have any more pictures, especially of how the batteries fit inside? Details? What's it made of?

Any chance you can experiment with gearing the bike down a bit? Maybe try a 28T chainring on the front, and try a 9T on the motor? (Just for research...) I have re-geared every bike I have owned since the '70's in order to make climbing easier. I know that with a 1x drivetrain, you would loose top speed, but it might be worth the tradeoff.

I do recommend removing the chain when working on the stand with the battery connected. It doesn't take long, and could prevent an accident.
 
I've had the Metr working for a while now. The data is very interesting. I implemented the cruise control switch at the same time.

The first surprise was that I stop and rest more often than I thought, especially when riding with friends. I was sure we stopped three times on our standard ride. Turns out we stop nine times! Lots more re-grouping and chatting than I realized.

I now have a better idea of my power usage. I have the cruise control set to about 135 watts, which is working well. When I am using power, I would estimate that 90-95% of the time I'm using cruise control. I'm using higher power less than 1% of the time. The remaining 4-9% of the time, I'm using less than 135 watts, usually when riding with friends. The flat plateaus on the "motor Current" trace (dark red bottom) are cruise control. With the hand throttle, I cannot hold it steady (3rd chart)

metr 1a.jpg

metr 2a.jpg

metr 3a.jpg

I did notice that at high power, the MOSFET temperature increased quickly. The graph below shows the rate of temperature increase in degrees (F) per second at various power levels. At the cruise control setting of around 135 watts, I saw 0.10 - 0.12 Degree F/Second temperature increase. However, at over 500 watts, I saw between 1 and 2 degrees F/second. This works fine for me, as I only use high power for brief bursts. I would not recommend high power for more than 30 seconds or so.

The controller lives in a "waterproof" pouch, so no cooling air flow. Probably worst case.

I'm using a 4S battery pack. I'm curious if 8S would drop temperatures...

Temp Increase b.png
 
FZBob said:
c_saayman said:
So I took the bike for its first test run. The best part of the bike is that it still has the feel of my original bike, just a bit heavier.

With this new bike if you want power you need to spin the chain ring fast to get the motor working. It reminded me of the spinning gym classes when I though people were crazy torturing themselves like that. I don't think this is a bad thing. I can only get better.

With my gear ratio not being so good as FZBob (10T/90T - 32T-50T) I need to pedal a bit fast to get the same power.
Nice looking conversion! How did the ISCG mount work out?

I like your battery holder, very clean integration. Do you have any more pictures, especially of how the batteries fit inside? Details? What's it made of?

Any chance you can experiment with gearing the bike down a bit? Maybe try a 28T chainring on the front, and try a 9T on the motor? (Just for research...) I have re-geared every bike I have owned since the '70's in order to make climbing easier. I know that with a 1x drivetrain, you would loose top speed, but it might be worth the tradeoff.

I do recommend removing the chain when working on the stand with the battery connected. It doesn't take long, and could prevent an accident.


Thanks - The ISCG mount works really well. No movement on the motor.

The battery was custom made from 18650 Sanyo 3400Ah cells - 4S4P, fitted inside a 30mm x 50mm rectangular tube. Animalector spot welded the cells for me. He has done a great job. The biggest BMS that fitted into the tube was 40A. The battery is a bit heavy, but I wanted it to be as solid as possible.
Battery inside_1.jpgBattery inside_2.jpgBattery inside_3.jpgBattery total weight.jpgBattery outside.jpg

The 12 speed SRAM chainring I'm using is an X Sync 2 and it only goes down to a 32T. I have not seen any other smaller 12 speed chainring yet. Not sure if the chain is anything special. I'm not very impress with the SRAM NX 12 speed drive train. I would like to convert it to a 11 speed Shimano XT drive train with a 11T-46T cassette and maybe 28T or 30T on the front. At this stage it is just too expensive +/- $500

I will still try the 9T on the motor. Just haven't manage to cut the key into it yet.
 
c_saayman said:
The battery was custom made from 18650 Sanyo 3400Ah cells - 4S4P, fitted inside a 30mm x 50mm rectangular tube. Animalector spot welded the cells for me. He has done a great job. The biggest BMS that fitted into the tube was 40A. The battery is a bit heavy, but I wanted it to be as solid as possible.

The 12 speed SRAM chainring I'm using is an X Sync 2 and it only goes down to a 32T. I have not seen any other smaller 12 speed chainring yet. Not sure if the chain is anything special. I'm not very impress with the SRAM NX 12 speed drive train. I would like to convert it to a 11 speed Shimano XT drive train with a 11T-46T cassette and maybe 28T or 30T on the front. At this stage it is just too expensive +/- $500

I will still try the 9T on the motor. Just haven't manage to cut the key into it yet.
That's a nice pack. It really blends into the bike. Looks like I need to learn about 18650 cells as an alternative to RC batteries. It seems like the most difficult part is spot welding tabs to the batteries.

Do you have a direct mount spider on your crank? Would something like this open up some options? https://northshorebillet.com/products/2-x10-104-bcd-spider-for-sram-x0-x9-cranks
2x10 spider a.png
I think this chainring will work on 12 speed SRAM chain, and it's offset to match the chainline. https://www.wolftoothcomponents.com/collections/chainrings/products/64-bcd-chainrings
Wolf Tooth a.png

Filing out keyways is no fun! I use a round file to clear out the "D", and then spend 15 minutes or so with a square jewlers file to make the keyway. Every time I do it, I tell myself to buy a keyway broach...
 
Thanks for the info. I will have a look. I like the Wolftooth products. I am using a 104 BCD spider.SRAM adaptor.JPG The new Rebel sprocket is also a 104BCD, so I would like to stay with this mount. The smallest I've seen on a 104BCD was 30T, with the spider interfering with the chain. My SRAM chainring is having some really funny looking teeth.SRAM 32T on Ebay.JPG

The 10T motor sprocket I'm using is a bit different then yours.10T motor sprocket.JPG I had to drill through about 10mm and then file it out to make the key. The 9T I'm using seems to be from a much harder metal, and the drill was not making a dent. Something worth mentioning is that I've placed a rubber o-ring on the narrow part (like you did) and the motor and chain are very quite. I sometimes don't even know that it is on.
 
c_saayman said:
Thanks for the info. I will have a look. I like the Wolftooth products. I am using a 104 BCD spider. The new Rebel sprocket is also a 104BCD, so I would like to stay with this mount. The smallest I've seen on a 104BCD was 30T, with the spider interfering with the chain. My SRAM chainring is having some really funny looking teeth.

The 10T motor sprocket I'm using is a bit different then yours.10T motor sprocket.JPG I had to drill through about 10mm and then file it out to make the key. The 9T I'm using seems to be from a much harder metal, and the drill was not making a dent. Something worth mentioning is that I've placed a rubber o-ring on the narrow part (like you did) and the motor and chain are very quite. I sometimes don't even know that it is on.
I believe the North Shore spider has both 104 BCD and 64 BCD mounts. This would allow for a 26 or 28 T chainring on the 64 BCD mount, and the Rebel sprocket on the 104 BCD. (Possibly even a 38T and 26T 2x setup and the 90T Rebel sprocket, with an ISCG mount front derailleur, if it all fits...).

Making a keyway in that motor sprocket looks extremely difficult, especially if it's hardened! I'm assuming that shipping costs for the electricscooterparts.com sprockets is not an option. If I had the sprocket in your picture, I might try the set screws. I would file two flats on the motor shaft, about 1mm deep, outboard of the sprocket (the shaft is not hardened...), and use plenty of loctite.
 
Thanks FZBob, you made me spent more money. I've ordered the 30T Wolf Tooth chainring. I don't think changing from 32T to 30T will make much of a difference. The main reason I bought it was that when I'm on the 50T at the back the chain line is very bad. With the 30T I hope to drop to 46T to get a better chain line and then only use the 50T on really hard climbs.
104BCD-30T-Red-02_590x.jpg
Changing the motor sproket from 10T to 9T will have the biggest impact. I'm currently working on that.

I've done now another two rides and getting use to the set-up. The main problem is that when I have max power selected and I use the push button, the ride is very jerky. This will not be the problem if you use a throttle. I've now change the way I ride. On a fairly steady climb with lots of switch backs, I lock the power to 75% (cruise control) and just pedal hard. Any steeper bits on the track I slide the level control to max power and when I get to the top I just drop the switch back to 75% ( cruise control still locked in). At any time the front brake can cut the power if I need to. This seems to work, but it is a bit of a hassle. I'm sure Kepler's Arduino will be able to solve this problem.
 
Polygon finished!

I have fitted the 9T sprocket on the motor and the 30T chainring. It all works very well. The 9T works better, but as mentioned by everyone else, it is more noisy. All that's left now to do is to ENJOY IT!
 

Attachments

  • Final CCT.jpg
    Final CCT.jpg
    161.7 KB · Views: 4,003
  • Polygon final1.jpg
    Polygon final1.jpg
    194.3 KB · Views: 4,003
  • Polygon final2.jpg
    Polygon final2.jpg
    211.5 KB · Views: 4,002
  • Polygon final4.jpg
    Polygon final4.jpg
    234.4 KB · Views: 4,002
  • Wolf Tooth fitted.jpg
    Wolf Tooth fitted.jpg
    202.6 KB · Views: 4,003
  • 9T sprocket.jpg
    9T sprocket.jpg
    78.3 KB · Views: 4,002
c_saayman said:
Polygon finished!

I have fitted the 9T sprocket on the motor and the 30T chainring. It all works very well. The 9T works better, but as mentioned by everyone else, it is more noisy. All that's left now to do is to ENJOY IT!

Fantastic work. Especially love the battery and cable / controller management solution.
 
I've ridden this bike, and it's a credit to Chris how well the bike is built. It looks amazing and is super fun to ride.

Great work Chris. Catch you on our next ride.

Andy
 
That is nicely integrated! I like the overall looks better than the commercial E-Bikes. Lots of cool little details, like the cover over the motor phase wires.

If you made that keyway with a file, you have my respect!

Have fun riding!
 
I'm planning out a fixie build with sick bike parts cranks. [strike]The biggest cog I found for the hub is 22T 3/32 chain[/strike].
Assuming wolf tooth 30T with the 3/32. And lets say 53 rebel and 9T motor sprocket, I'd be going pretty fast on 10s.
26in 2.4 wheels.
1200Rpm at the wheel I think, is this a good idea ? :twisted:
edit:there are bigger fixed cogs and adapters, I could also mount a disc holder for splines hg
 
cheapcookie said:
I'm planning out a fixie build with sick bike parts cranks. [strike]The biggest cog I found for the hub is 22T 3/32 chain[/strike].
Assuming wolf tooth 30T with the 3/32. And lets say 53 rebel and 9T motor sprocket, I'd be going pretty fast on 10s.
26in 2.4 wheels.
1200Rpm at the wheel I think, is this a good idea ? :twisted:
edit: here's how I make bigger cogs http://www.63xc.com/jasom/grannycog.htm, so my new question is what single speed cog do I want
What is your goal? A bit of assist to help out on hills and make pedaling a bit easier, or big serious power so you can go FAST?

This setup, as currently configured, is useful for just a bit of assist. Like around 150-200 watts continuous and bursts of maybe 500 watts for a few seconds. At 500 watts, the controller will shut down after about 30 seconds, or it will melt. (see my post from a few weeks ago). Study this chart - Power to overcome aero drag increases with the square of speed. Looks like to reach your 1200 RPM (93MPH) you would need at least 7000 watts at the back tire for just for aero drag. I suspect more like 15,000 watts going in to the motor to account for rolling resistance, motor efficiency, acceleration, etc.

bike-speed-power-graph.gif

If you do go forward, let us know how it works out. I'm curious...

If you really want to go 93MPH (1200RPM), I suggest you visit your local Kawasaki/Honda/Suzuki/Yamaha dealer... Proper tool for the job, and FAR more fun to ride fast.
 
Sorry I dont want 93mph at all, 1200*60*0.000695km= 50Kmh? going close to 40mph is of interest to me.
I am highly interested in the vesc6 and power dense motor
 
cheapcookie said:
Sorry I dont want 93mph at all, 1200*60*0.000695km= 50Kmh? going close to 40mph is of interest to me.
I am highly interested in the vesc6 and power dense motor
Got it. The VESC 6 does look interesting. Just curious - is this your main transportation or a toy? Are you OK with the time and expense of a development project?

So just to fix the math -
Tire rollout = 26" * 3.14 /12 = 6.8 feet per rev
6.8 feet per rev *1200 RPM = 8164 feet per minute
8164 *60/5280 = 92.77 mph

So for 40 MPH you would want about 517 RPM on the rear wheel.

Strictly looking at Gear RATIOS, 10T and 90T primary and 30T Chainring, 24T rear sprocket MIGHT get you there, but I would not be surprised if a 35T chainring might work better. Some testing and development required. Note that the chainring will be spinning at over 400 RPM.

Power and Chain Strength: Since you will spin your motor about 2.5 or 3 times as fast as mine, you can probably run at 2.5 or 3 times the power I run, with the same chain tension/stress. I run at 150 watts continuous, and 200 is probably OK. So you might be able to run at 600 watts continuous with reliability. On the other hand, I have run over 500 watts for 30 seconds, limited by my VESC. So if the VESC 6 supports it, you might be able to run at 1500 watts for longer. I'm not sure how the little motor drive chain would tolerate that. I suspect it might have a short life, but I would love to hear actual results!

Power required for 40 MPH? Based on the graph, I would guess around 1500 watts. I was able to get my Mtn bike (on knobbies) up to 30MPH using 500 watts AND pedaling as hard as I could. Flat pavement and no wind.

Safety? I have caught a shoelace in my 90T a couple of times, but I can easily stop my motor with my legs. With your setup (Fixie & Freewheel Crank at 400RPM), nothing is going to stop that chainring... If the freewheel siezed, I would bet big money it would chuck you right over the bars. A freewheel in back would add some safety.

Acceleration? A few gears in back and you would get up to speed FAR quicker.

If you are looking for a fun development project/hobby, this might be interesting. I would love to hear about it! Definitely no guarantee of success though. If you need transportation, I would suggest a Bafang or similar.
 
Forgot to pi my diameter...
If I went with 30T+ rear sprocket, the motor rpm would be totally mismatched to the human. If I could cover a wide range of speeds: low speeds high torque with human and high speed with bldc that'd be cool I guess.
 
A month ago I posted some data on VESC heating vs power. It looked like this put a limit on continuous power. That was using a 4s battery.

I was curious to see if changing from 4S to 6S (NO gear ratio change) would improve things. More Voltage and less current... So I ran a quick & dirty experiment. 500+ watts up a short steep hill, same speed, same gear. I collected data from the METR. (numbers are not exact, as I was picking them off a chart)

4S:
Battery Voltage: 15.5 Volts
Battery Current: 34.5 Amps
Duty Cycle: 60%
Motor Current: 68 Amps
VESC Heating Rate: ~2.5 Degrees/Second

6S:
Battery Voltage: 23.25 Volts
Battery Current: 23 Amps
Duty Cycle: 40%
Motor Current: 70 Amps
Vesc Heating Rate: ~2.55 Degrees/Second

Not a big surprise. It looks like just adding battery voltage doesn't make much difference. Battery current goes down, but motor current remains the same. Heat rate apparently goes with motor current. The bike also felt about the same.

It looks like the path to more power is more RPM, while keeping the Motor Current/Torque low. This would also keep the stress on the #25 chain down. Maybe using a freewheel crank for a second stage of reduction, as cheapcookie mentioned, although this will spin too fast to allow for pedal assist at high speed. Might make an interesting commuter though...
 
FZBob said:
A month ago I posted some data on VESC heating vs power. It looked like this put a limit on continuous power. That was using a 4s battery.

I was curious to see if changing from 4S to 6S (NO gear ratio change) would improve things. More Voltage and less current... So I ran a quick & dirty experiment. 500+ watts up a short steep hill, same speed, same gear. I collected data from the METR. (numbers are not exact, as I was picking them off a chart)

4S:
Battery Voltage: 15.5 Volts
Battery Current: 34.5 Amps
Duty Cycle: 60%
Motor Current: 68 Amps
VESC Heating Rate: ~2.5 Degrees/Second

6S:
Battery Voltage: 23.25 Volts
Battery Current: 23 Amps
Duty Cycle: 40%
Motor Current: 70 Amps
Vesc Heating Rate: ~2.55 Degrees/Second

Not a big surprise. It looks like just adding battery voltage doesn't make much difference. Battery current goes down, but motor current remains the same. Heat rate apparently goes with motor current. The bike also felt about the same.

It looks like the path to more power is more RPM, while keeping the Motor Current/Torque low. This would also keep the stress on the #25 chain down. Maybe using a freewheel crank for a second stage of reduction, as cheapcookie mentioned, although this will spin too fast to allow for pedal assist at high speed. Might make an interesting commuter though...

Interesting, do some of these outrunners come with temp sensors? whats the motor heating rate 4svs6s?
 
cheapcookie said:
Interesting, do some of these outrunners come with temp sensors? whats the motor heating rate 4svs6s?
All six of my motors have sensor cables. Five of them have six wires in the sensor cable, which I believe indicates they have temperature sensors. As I run sensorless FOC, my sensor cable is not connected...

My big motors don't get warm to the touch in my normal riding (~150 watts mostly), but I'm not sure about high power. If I get a chance I will plug my cable into the VESC, and see if I get motor temperature readings. Not sure when, as my wire routing is a pain...
 
Back
Top