Kepler eboost. The journey and evolution

[adrian_sm]

Price sounds reasonable, when you consider the feature list is a lot longer than a Cycle Analyst.

 

[Kepler]

Hey,

Just wondering. If you have any slip detection built into the interface. This would help us tune the maximum amount of engagement the drives have. It is easy enough for us to tune the geometry to ensure there is no slip, but if you have too much engagement you are just wasting power to the additional rolling resistance. Since you already match motor speed to wheel speed on start up, you have most if the maths already worked out.

If you got really excited you could have traction control. But I am more interested in using it to optimism the tyre engagement limits during developement, so that we can simplify the installation instructions.

Just a thought.

Adrian

We have considered designing slip detection into interface but it requires us to measure actual motor rpm for us to do the comparison. For speed matching, we have a calculation that just estimates the correct motor speed to tire speed ratio. This took a bit of work to get the balance right but this calculation method gets it close enough for comfortable engagement. 5S uses a different calculation to 6S but this all happens behind the scenes when you select the battery type.

I agree that slip detection would be a nice feature but havent included it at this stage as we wanted to avoid the extra complication of an RPM sensor. Slip really isn't much of an issue with current limiting and the speed matching is accurate enough with the current method used. We might re visit this though in the future.

 

[Kepler]

I am really concentrating on getting the interface finished off. Unfortunately, we keep on coming up with too many cool ideas to make it even better

I hope I don't sound pushy, but...you really need an affordable base model. Then, save a few desirable features for the upgrade V2.0 (example: Cycle-anylist..."big-screen" cycle-analyst)

Keep in mind, this an RC interface that has a display screen as part of the package. We were going to offer the unit without the display but decided this didn't make sense as every unit needs to be setup to suit the bike installation and such needed some sort of display to do this simply. We thought about using beep programming like an RC ESC or to use a GUI and laptop. Again we decided against this as beep programming sucks and programming through a a laptop needs a USB interface device to connect to the unit that added another $25 anyway.

If we can think of a way of offering the interface in a cut down format, we will, but as it stands I think this device will be great value for money.

 

[Kepler]

The first of the new prototype interfaces are now being assembled. Below is a shot of the main board and the display board with a test cable between the two. The production version uses a multi core with DIN plugs to connect the 2.

Note the chip top middle with the 8 legs and pins on each side. This is the current sensing chip. the battery side of the ESC cable sits over the top of this and is pierced by the pins to pick up voltage. Cable ties through the board are then used to hold the cable securely in place. The six pins out the side are only for testing purposes and wont be there on the production units however the 6 pins sticking up are for the future addition of a a Bluetooth. The board also has previsions for a photo sensor, tail light, and second current sensing input if we decide to go to phase current sensing also. Board measures 39mm x 45mm.

The LCD is also used as a junction box to plug in the throttle button, wheel speed sensor, and brake sensor. There is also a plug for a high powered 12W 24V head light.

Display has been kept quite basic for now. The main screen displays speed in the top section and has a bar graph fuel gauge below it. As soon as you stop, the display automatically displays distance travelled. Pressing the throttle button while stationary displays mah's used.

Throttle package has 3 buttons. One large main button as the throttle and the other for functions like programming to suit the bike and region and control of head light and tail light.

We plan to fit the LCD display into an enclosure however the interface will be supplied without an enclosure but of cause with plugs ready to go. The reason for this that the interface is designed to fit directly into my drives and as such doesn’t need a separate enclosure. I can't see this being too much of an issue for most people here, but if it is, let me know and I will see what we can come up with.

So there you have it. Getting closer. Don’t ask me for release date as I don’t want to disappoint. However we are doing all we can to get then out into the field.

 

[AussieJester]

Great to see ongoing advancements with the eboost setup Kepler
when this new 'gadgetry' is released and combined with current offerings
it will be pretty much a complete setup for those wishing to run lightweight friction setups!

All the best with continued development and production buddy

KiM

 

[41south]

Great work, I'm looking forward to being able to place an order - no pressure of course 8) This solution is right up my alley

 

[spinningmagnets]

There are people who have never bought a single thing from ebikes.ca...except the cycleanalyst. I think this gadget will do well, best of luck my friend!

 

[Kepler]

Thanks guys,

AJ, I really think this is the missing link to a reliable low cost friction drive setup. The fact that I have never blown a motor or ESC proves the concept of current limiting mapped against wheel speed works well even when measuring battery current as apposed to phase current. Thanks for the support mate.

41south, cheers mate. Hopefully I can supply in the very near future.

Spinner, funny you mention the CA, I am one of those who have only bought CA's from ebikes.ca and nothing else (not that I have anything against their other stuff, just no need for it)
Not sure how far reaching the market is for the interface as a stand alone unit. To be honest, its not a real concern as the main reason for the interface is so my RC drive could be safely sold to the general public. Interesting times ahead. Just hope I dont run out of money before I get there. Thanks also for the support. Much appreciated.

 

[Kepler]

Latest drives are now being put together. Trying some different anodising combinations. Most are still anodised black but I thought it would be worth experimenting with some different colours and textures.

Also working on a new configuration that mounts underneath a standard Topeak seat post mount rack. This configuration will use most of the drive's standard parts so it will be just a matter changing out a few low cost parts to re configure the drive. Should have the prototype finished in a few days.

 

[41south]

Loving the bling :lol:

 

[Kepler]

Another shot of a blinged up drive.



This shot shows the drive with a 6S 5000mah pack piggybacked on. This works well for short hauls of around 10 to 20 kms depending on the amount of assist used. The idea being that the entire drive package is removed form the bike in seconds.

 

[amberwolf]

I've gotta say, it still amazes me that the whole thing fits in such a small, neat, and interesting-looking package.

 

[maydaverave]

with the new 8ah hk packs this would be a super compact option with decent range. 8)

 

[Kepler]

Latest batch of drives are nearly ready for sale. Unfortunately the interface is still a not quite ready but I should see them in a 2 or 3 weeks. Been testing the drive with a simple servo tester, ramp control, hall throttle setup. This works fine for the person willing to act as the current limiter and who understands how to ride around the RC friction drive idiosyncrasies.

Mid mount option is also available upon request.

 

[recumpence]

Hah! I love it! Man, I recognize that "Table full of custom assemblies for resale".

At the peak of my RC helicopter manufacturing, my oldest two kids would be separating screws and spacers, folding instruction sheets, and heat-sealing bags full of parts for me. I have a nearly 3 foot high stack of Paypal forms from completed orders that were all filled by hand.

I am so glad I am onto making fewer, but more costly, assemblies for people.

Anyway, I love seeing cottage industries spring up and flourish.

Matt

 

[Kepler]

Just when I think I am getting close to a finished product something else comes up.

Lots of testing was carried out with the launch ramp design. Although it worked quite well, I wasn't happy with how fiddly it was to set up. As such, the concept has been ditched for a more reliable method of engagement. Ideally my aim is for the drive to operate as seamlessly as possible without any tire contact when not in use. This has proven to be more difficult then first anticipated with flex in the seat post and tires not being perfectly round further compounding the problem. The reality is that the motor needs to clear the wheel by a few mm to ensure no contact with the tire.

My latest solution is to employ a small lightweight RC servo with a cam arrangement on the output to give the swing arm a push upon startup. Proof of concept was carried out essentially using the launch servo on a Y harness with the ESC. More throttle meant more servo output. This worked quite well however there were circumstances where the swing arm would push back on the servo output and cause the gears to strip. Solution for this was to fit a dedicated servo output from the interface so the servo could be controlled separately from the ESC output. A routine was programmed so that when the throttle button is pushed, the servo pulses on then retracts back safely out of harms way. Once the motor comes in contact with the tire, the normal progressive contact pressure takes over. Pulsing the throttle button to increase speed has no impact on the launch servo operation with the servo staying in the parked position until the throttle button is completely released and the motor stopped.

Below is a short video of the servo launch operating when still driven via a Y harness and ESC input signal. The pulse method is more difficult to demonstrate on the bike stand as some load is required to keep the motor in contact with the tire. On the road, engagement is very consistant and smooth. Although motor to tire clearance still needs to be reasonably accurate, its no where near as critical as it use to be.

Time to get some kms under my belt with this setup but I think this method will be a keeper


[youtube]f4buU4pUN1w[/youtube]

 

[Whiplash]

Very nice! I must say I love your design! I think you will sell a ton of them! Would you be interested in a US distributor?? My site is just about finished and should be up soon with pricing and what not... Just asking..

 

[Kepler]

Thanks Whiplash. Much appreciated.

Cost wise as it stands, the mechanics set costs $350 USD. Then you need to add around another $350 for ESC, motor, interface, and a single 6S 5800 mah pack. At $700 it isn't exactly cheap but in small quantities, that's simply what it needs to sell for to make it worth my while. My target sell price is around $550 for a complete unit which I think would open up a huge market. It just needs to be refined as possible to then invest in volume manufacture which would in turn allow the sell price to come down to where it needs to be.

I still see another 6 months hard work to get it to a stage that I would then be happy to take it to the open market in reasonably high volume. At that stage I will certainly be looking distributors.

 

[Whiplash]

Well, I think you will get there for sure, it is a very quality unit and I truly think it will have a place with the competitive bikers out there that want to ride to work assisted so they are not worn out and sweaty when they get there and then do their workout on the way home! Very quality looking piece and the electronics sound great as well! Keep me on your list for sure when you are ready for sales/sellers!

 

[deVries]

Sun May 15, 2011 Latest batch of drives are nearly ready for sale. Unfortunately the interface is still a not quite ready but I should see them in a 2 or 3 weeks.

It is a mystery to me as to why you are having such long delays in the development of this interface? Are you working with the same "top secret" ES member, or have you had to start over with other developers or new ideas, or is this just much more difficult than expected, or

The curious cat has got my attention about all the trials & tribulations & the back-story for all this interface development. Can you cut us in on the inside story to this intriguing missing link for FD?

Edit: No intent to be rude or crass here. (I did razz you about this at the end of last year though. ) I am just curious & want to know what all the troublesome issues are in layman's terms. You know I've been following this "throttlizer idea" for more than a year now... before you decided to offer your product.

 

[adrian_sm]

That is a bit harsh.

There is a big difference between hacking something together that works, and releasing your own custom hardware/software in a large scale to paying customers. The feature set that Kepler has highlighted, and the complexity of achieving this takes time. No one wins if it is half baked when released.

Keep up the good work Kepler.

- Adrian

 

[deVries]

That is a bit harsh.

There is a big difference between hacking something together that works, and releasing your own custom hardware/software in a large scale to paying customers. The feature set that Kepler has highlighted, and the complexity of achieving this takes time. No one wins if it is half baked when released.

Keep up the good work Kepler.

- Adrian

I'm really just wanting an update to what obviously has become a much longer-term project than "we" all imagined. Last year Kepler was going to adopt or adapt the "throttlizer" but that fell through, and towards the end of 2010 Kepler took on this ESC-ramp project with some ES member helping too.

I recall another post about this ESC throttle device:

Mon Dec 20, 2010 2:43 pm
Wow ultimatums. Thats a bit rough.

I understand keeping things a bit more under wraps when it is a commercial developement. I would prefer John spends his time getting it right and on sale, rather than keeping us updated.

Keep up the good work.

====================

I don't think I've been really "harsh" about my expectations "when" this product will materialize, but it would be nice to know a real ETA that can meet that expectation given by the developer himself. I mean, back in December of last year it seemed it was only weeks or at most a month or two away from final testing or near completion.

I understand delays & development issues, etc. Just want to have a legit update about it all...

If an ETA is just not possible within X time-frame, then simply say that instead of offering an ETA.

 

[Kepler]

It disappoints me not having the interface out in market already especially after reporting on a number of occasions that the interface will be ready in a few weeks. I apologise for this and are as keen as anyone to get it out in the market place as l really would like to start to recoup some of my investment.

This being said, I really want it to be right and this has meant that just when I think it can be released, we find something else that needs to be sorted out or end up adding another feature. I think I am best off not trying to make any promises as I know this can be frustrating for people. Bottom line is that it will be released soon in small numbers. I have a few beta units that are being fitted to drives locally and expect another 5 to be completed in the next few weeks. This is what l expect but cant make any firm promises.

 

[deVries]

It disappoints me not having the interface out in market already especially after reporting on a number of occasions that the interface will be ready in a few weeks. I apologise for this and are as keen as anyone to get it out in the market place as l really would like to start to recoup some of my investment.

This being said, I really want it to be right and this has meant that just when I think it can be released, we find something else that needs to be sorted out or end up adding another feature. I think I am best off not trying to make any promises as I know this can be frustrating for people. Bottom line is that it will be released soon in small numbers. I have a few beta units that are being fitted to drives locally and expect another 5 to be completed in the next few weeks. This is what l expect but cant make any firm promises.

Thanks for the update. I understand & relate to the idea of being overly optimistic about when some product/project will be completed. It's not possible to account for unforeseen delays that "appear" & weren't even thought of or known before. It's good you are finally down to the point of beta testing a few units. Are some of these beta testers ES members too? :wink:

One point you have made in the past is that these units would be made available & be usable for end users that have their own FD or a design that is different than your method. For example, in my case, a FD that might be based more on EVTodd's original design with a sliding mechanism using a roller. Btw, I think it is EVTodd's original design that inspired Kepler & later Adrian to make their own FD designs.

I'm not certain that Adrian's throttle control that he is developing is meant to be free standing too, so it could be successfully used by other FD designs too? But Kepler has maintained that his throttle-ramp control would be available & should work for other FDs too. Is this still your intention Kepler, & will it still likely work for other FDs?

Hey Adrian, maybe you should make your throttle control available for other FDs too! I would encourage you to do that, if possible, since frying ESCs is such a universal problem including a need for better sync issues. Also, two offerings are better than one, since having other product solutions may improve the end product(s) for everyone.

And thanks Adrian for what you are trying to do too! I think you are certainly contributing to FD development with your product too! Why not offer your throttle control for use in other FDs too?

Bottom line, we needed an ES solution more than a year ago, so if you guys don't provide something soon I hope someone will. Even Adrian & Kepler recognize the need for a solution, so any solution is in their best interest too.

 

[Kepler]

Thought I would post up some pictures of a customer's bike with the latest drive and interface fitted. The bike is riden as a normal bike most of the time with the drive only used as a low power assist when tackling hills. 6S 5Ahr is plenty for this sort of work and with the battery piggy backed on the actual drive, the entire drive package (except for the speedo and throttle) can be removed in less then 10 seconds.

Interfaces are finally sorted and debugged so its now time to get serious.