Beta CA build for R/C controllers

[justin_le]

I was thinking about a way to do a setup for the CA-shunt and the Castle CapPack along the wire.......viewtopic.php?f=2&t=33268&p=483605#p483605

Have you guys Any suggestion?

That should be pretty straightforwards, just add the shunt inline with the black lead much as Kepler did, and seal the whole shunt+cappack with a piece of heatshrink to keep it looking tidy:

Shunt on Controller.jpg (31.31 KiB) Viewed 1513 times

-Justin

[justin_le]

Presumably, you're going to stock the Thun sensored BBs?

Hoping so, just sent an inquiry to Thun last night. It does mean that people wanting to use this have to drill a small hole in the BB shell of their bike to feed the sensor wire through and then crimp on their own connectors, but given the level of hardware hacking that goes on here that shouldn't pose much of a problem. -Justin

[mrbill]

I dunno about others here, but I've always been dying to see just how many watts my legs were adding to the mix when riding an ebike.

As someone who comes to e-biking from the sport/fitness side, I can say that I've always been interested in seeing how power is allocated between the motor system and human pedaling. For many years I've used a PowerTap rear hub, and I've found it invaluable for measuring my own performance and for standing in as a makeshift dynamo on the bench to measure motor system efficiency.

The sport/fitness world seems to be moving toward the ANT+ standard of interoperability for power meters. Various designs (e.g. rear wheel (PowerTap), bottom bracket (Ergomo, THUN), crank (SRM, Quarg), and pedal axle (Garmin Vector)) are wireless devices (2.4Ghz) and log data wirelessly to a head unit (such as a Garmin GPS device or proprietary head unit) that logs other performance parameters such as speed, cadence, heart rate, altitude, location (using GPS), that can subsequently be correlated and uploaded to a PC or Mac for further analysis. Have you considered adding ANT+ compatibility so that a variety of power-measuring devices could be used?

Those of us using bottom bracket motor drives would have to use crank-based or pedal-based torque sensors to isolate human-only power. Bottom bracket or rear hub torque measuring devices will only show combined power.

[www.recumbents.com]

Justin,

I'm very interested in the Thun hub, please let me know the price when you get it. And of course your new RC CA! I just bought one and will be using it soon, but I may need to buy another when the new one is available!

-Warren.

[nicobie]

This is really good news!

The new add ons will cover every thing I need a bike monitor to do. The pedal torque sensor sounds intriguing, I'm almost afraid to find out just how much I'm contributing.

Will you be stocking a temp probe? Or will the CA be adjustable for any probe?

It looks like I need to start saving for the new model.

[Miles]

Those of us using bottom bracket motor drives would have to use crank-based or pedal-based torque sensors to isolate human-only power. Bottom bracket or rear hub torque measuring devices will only show combined power.

Bill,

The Thun system measures the torque transferred through the axle from the left pedal. With a Cyclone type BB drive, it should only register human power, I think.

[mrbill]

It does mean that people wanting to use this have to drill a small hole in the BB shell of their bike to feed the sensor wire through and then crimp on their own connectors, but given the level of hardware hacking that goes on here that shouldn't pose much of a problem.

Perhaps not, but drilling holes into a frame anywhere will weaken it. A video on Thun's web site shows that in 2012 they will be coming out with an "X-Cell RT 2.0" with wireless transmission. No indication if this will be Thun-proprietary or ANT+ compatible.

http://www.youtube.com/watch?v=EXS1pGaMOeg

[mrbill]

Those of us using bottom bracket motor drives would have to use crank-based or pedal-based torque sensors to isolate human-only power. Bottom bracket or rear hub torque measuring devices will only show combined power.

Bill,

The Thun system measures the torque transferred through the axle from the left pedal. With a Cyclone type BB drive, it should only register human power, I think.

Depends on the attachment. My first Cyclone drive was to the left crank. StokeMonkey also attaches to the left crank. Both systems force the pedals around when motoring, for better or worse. A right-crank or "4th chainring" motor connection would have the Thun (or Ergomo) measuring only left leg torque. Calculating power from one leg and multiplying by 2 increases system error due to uneven leg strength.

There's room for value in this space. Most of these torque-measuring/power devices are in the range of $1000US or higher (although Ergomo is running a sale on their 2007 model, $400US). What is the cost of a Thun BB?

[Miles]

A right-crank or "4th chainring" motor connection would have the Thun (or Ergomo) measuring only left leg torque.

Which is all it can do, anyway.....

There's room for value in this space. Most of these torque-measuring/power devices are in the range of $1000US or higher.....

I was thinking the same. I'm surprised that Thun haven't done it...

[liveforphysics]

In a frame that is well suited towards the loads and abuse an Ebike endures, drilling a small hole on the bottom bracket shell will have no effect. Round holes don't create focused stress riser points, so they do very little to structural integrity.

[panurge]

I was thinking about a way to do a setup for the CA-shunt and the Castle CapPack along the wire.......viewtopic.php?f=2&t=33268&p=483605#p483605

Have you guys Any suggestion?

That should be pretty straightforwards, just add the shunt inline with the black lead much as Kepler did, and seal the whole shunt+cappack with a piece of heatshrink to keep it looking tidy:

Shunt on Controller.jpg

-Justin

Beautiful.....thanks a lot for the picture.....

I think the thermistor is another great feature......as the separated throttle.......

Regards

Jules

[amberwolf]

So...would you do trade-ins of old CAs to get a price break on the new version?

Cuz it would be nice to be able to have a BionX-like system but with *real* power behind it. That's what I originally wanted to do with CrazyBike2, but never got a working pedal-torque sensor going, plus all the circuitry to parallel that with a throttle control on the brushed controller I was using then.


BTW, a quick google for the Thun sensor itself foudn this tidbit:
http://www.aliexpress.com/fm-store/6055 ... ensor.html
whcih appears to be controller, pedalec sensor and Thun sensor all in one package for only $160 for 5 sets. Buuut...given the way things usually work out, it probably does not include anything but the controllers. Anyway, it did find a pic of the sensor:


I'm having trouble finding places that actually sell the Thun sensor itself, or at least, that show a price for them. That probably means they'll be out of budge range for a lot of people (including me).

Hopefully I'll be able to come up with some other sensor input that can do the same type of output range for input that the Thun does, so the new CA can still be used to control things based on it.

[www.recumbents.com]

The Thun only shows torque from the left pedal? Um, maybe it's not such a good thing then, and only good for telling if the rider is pedaling.

[johnrobholmes]

You could "cheat" the thun being used as a pedelec control by jumping up and down on the bike I love it

[justin_le]

You could "cheat" the thun being used as a pedelec control by jumping up and down on the bike I love it

Indeed, in my first playing around with a THUN unit earlier in the year I was only using one of the two pulses from the cadence encoder, and there was literally no way to prevent pedalec mode from kicking in if you are standing on the pedals left foot forwards. Any slight rocking back and forth looks to the system like you are pedaling with massive force. Proper use of quadrature input solve that

It was also neat to see how much negative torque is applied across the spindle in the course of routine pedaling, where the action of the right leg pressing down does the work of lifting up your left leg. I suppose it's more efficient use of leg muscles to push down rather than to lift up? Interestingly you wouldn't see any of this left<->right pedal play with a strain sensor on the chainring. So even though that would properly show the net power of both legs getting to the wheel, it doesn't show the percentage of torque from one leg that is being absorbed by the other leg rather than being transmitted to the chain.

Anyways, I did hear back from THUN and they're pretty interested to see an ebike display and interface compatible with their sensor, and seem keen to help us make that happen. The pricing is reasonable enough that we'll probably stock them by Jan/Feb of next year and be able to offer a THUN + CA bundle in the $330-$350 price range. I suppose it's a bargain compared to existing bicycle power meters (mostly >$1000 as has been mentioned here), though may seem a bit pricey if all you want is to add pedalec functionality and don't care so much about seeing human wattage.

There are 5 different spindle lengths to choose from with the THUN BB, anybody have an idea on the distribution of spindle lengths in ebike usage?? I've never had consider this question before.

BTW, a quick google for the Thun sensor itself foudn this tidbit:
http://www.aliexpress.com/fm-store/6055 ... ensor.html
whcih appears to be controller, pedalec sensor and Thun sensor all in one package for only $160 for 5 sets. Buuut...given the way things usually work out, it probably does not include anything but the controllers.

That's correct, it's a china trader selling controllers only, the use of the THUN image and name is pretty inappropriate. It also looks (judging from the cable pinouts) like it probably is just using the THUN bottom bracket for cadence sensing, rather than for actual pedal torque.

-Justin

[justin_le]

Will you be stocking a temp probe? Or will the CA be adjustable for any probe?

Currently implemented as a 16 step look-up table with linear interpolation to use a familiar and inexpensive 10K NTC thermistor. But the generic use of a look-up table means that it could be setup to map pretty much any sensor's input/output curve. So if there are other things that people would want to use that channel to measure (like barometric pressure, "G" force meter, etc.) they could set up a mapping table and load that into the CA's eeprom and it will scale things accordingly. Well that's the idea at least.

-Justin

[Dingo2024]

There are 5 different spindle lengths to choose from with the THUN BB, anybody have an idea on the distribution of spindle lengths in ebike usage?? I've never had consider this question before.

Hi Justin,

not sure on the distribution of spindle lengths but if looking at cyclone type setups then the axle length is 152mm to allow crank arm clearance with the motor (and it is all biased to the non drive side)

Q: do you know if they are only available in square taper or also ISIS fitment?

hope this helps and good luck with it all,


Ian

[mrbill]

There are 5 different spindle lengths to choose from with the THUN BB, anybody have an idea on the distribution of spindle lengths in ebike usage?? I've never had consider this question before.

Hi Justin,

not sure on the distribution of spindle lengths but if looking at cyclone type setups then the axle length is 152mm to allow crank arm clearance with the motor (and it is all biased to the non drive side)

Q: do you know if they are only available in square taper or also ISIS fitment?

hope this helps and good luck with it all,


Ian

I've used a variety of spindle lengths on my bikes, usually with an offset to one side or the other. One of my bikes uses 145mm spindle with full-right offset. To my knowledge only Phil Wood (who happen to be a near me) offer custom-length and custom-offset bottom brackets. The standard sizes probably work well for bikes with hub motors, but it would be nice if Thun offered a good range of lengths and if the offsets could be adjusted by the user or dealer.

[justin_le]

Hi Justin,
not sure on the distribution of spindle lengths but if looking at cyclone type setups then the axle length is 152mm to allow crank arm clearance with the motor (and it is all biased to the non drive side)

Hmm, didn't realize that the Cyclone system requires an extra long spindle. The options with THUN appear to be 120mm, 128mm, and 136mm, with the 120 and 128 options available with two different offsets (L and K). Dimension sheet is attached if anyone can make too much sense of it. I suppose it makes sense to err on the side of getting the longer lengths?

Q: do you know if they are only available in square taper or also ISIS fitment?

Looks like square taper only at the moment.

-Justin

[Dingo2024]

Hi Justin,
not sure on the distribution of spindle lengths but if looking at cyclone type setups then the axle length is 152mm to allow crank arm clearance with the motor (and it is all biased to the non drive side)

Hmm, didn't realize that the Cyclone system requires an extra long spindle. The options with THUN appear to be 120mm, 128mm, and 136mm, with the 120 and 128 options available with two different offsets (L and K). Dimension sheet is attached if anyone can make too much sense of it. I suppose it makes sense to err on the side of getting the longer lengths?

Q: do you know if they are only available in square taper or also ISIS fitment?

Looks like square taper only at the moment.

-Justin

ok had a quick look at the spec sheet....and it looks to me that the offsets are all on the right side (L2) ie drive side which would not help on a cyclone setup. This I guess is more for chain alignment on regular single, two and three chinring setups. That said I have spoken with the cyclone dealer in the UK and he assures me that their sytem (3 chain wheel) will work with a regular length bottom bracket no problem and has run this on his demo bike for months. (this was an ISIS BB though and could 'push' the cranks out a little further)

EDIT; this is from the cyclone tw site ........so 128mm works for ISIS

http://www.cyclone-e-bikes.tw/product.p ... _show=3235

also in 140mm flavour;

http://www.cyclone-e-bikes.tw/product.p ... 03&count=1

my guess would be the longer sizes of Thun would be fine

Ian

[mrbill]

ok had a quick look at the spec sheet....and it looks to me that the offsets are all on the right side (L2) ie drive side which would not help on a cyclone setup. This I guess is more for chain alignment on regular single, two and three chinring setups. That said I have spoken with the cyclone dealer in the UK and he assures me that their sytem (3 chain wheel) will work with a regular length bottom bracket no problem and has run this on his demo bike for months. (this was an ISIS BB though and could 'push' the cranks out a little further)

EDIT; this is from the cyclone tw site ........so 128mm works for ISIS

http://www.cyclone-e-bikes.tw/product.p ... _show=3235

also in 140mm flavour;

http://www.cyclone-e-bikes.tw/product.p ... 03&count=1

my guess would be the longer sizes of Thun would be fine

Ian

This depends largely on the tread or Q factor of the cranks in question.

http://sheldonbrown.com/cranks.html#tread

Most newer cranks bow out between spindle and pedal, but many older cranks have little or no tread and are essentially straight rods. The cranks included in the Cyclone kits have a high Q factor.

[mrbill]

I dunno about others here, but I've always been dying to see just how many watts my legs were adding to the mix when riding an ebike.

Justin,

If you're going to add a display for human power and torque, I hope you will include a calculation for human energy and average human power.

[mdd0127]

I've finally had a chance to play around more with the CA-LRC and I think it's amazing! I do have some thoughts for future models though.

CA-LRC Ideas

Issue: Due to the minimum speed setting, and also when you release the throttle and coast, there is a small surge/buck as the electric drive catches up to the coasting speed when the throttle is reapplied.

Possible solution:

Integrate two speed sensor inputs into the new upgraded chip CA. One is for the front wheel for overall speed sensing and the other is for the motor or reduction output(so it would need to be scalable). At pedaling startup, as soon as wheel motion is detected, the CA would give the ESC just enough throttle signal to make the motor/drive speed match the wheel speed, giving no drive effect but getting the drive "ready" to engage smoothly once the maximum minimum speed has been reached and the rider is able to engage the throttle. When the throttle is released for coasting, the CA should return to the "match wheel speed throttle output" until the bike is brought to a full stop by the brakes which can easily over-ride the no-load motor current. The speed sensing of the CA would need to be a little more sensitive for this to work well though. Maybe more speed sensor triggers on the wheel...

As a side benefit, the same actual wheel speed/electric drive speed info could be used to implement traction control, reducing damage to the trails, and anti wheelie protection.

I guess another way to do it would be to set the throttle output to always run the motor at a no load current limit when the bike is above the minimum cutoff speed......???

[Grinhill]

Justin,
Sorry if this has already been covered elsewhere in this thread, but is it possible to get this to work on earlier model CAs? (I have another CA which is about two years old, small screen / single button).

Thanks

[rodgah]

I have noticed this buck in bench testing also and it was intermittent, I couldn't get it to do it every time, and thought on the bike it wouldn't be noticeable.

I guess another way to do it would be to set the throttle output to always run the motor at a no load current limit when the bike is above the minimum cutoff speed......???

you could already do this by setting the throttle lower end limit just a little lower than the actual throttle output (cant remember the setting names) though I havnt tried to set the minimum speed, when I did this the drive was in an always on state.

you could change your riding style to not release the throttle completely unless you really have to . I like this because in theory your freewheels dont completely unload so there is no hammering effect when you apply the power again.

Rodger