High power RC motor and drive unit production

[recumpence]

I spoke with Dave about increasing the jackshaft distances and he said no problem. So, I guess it would not be hard to offer a drive that is slightly longer to accomodate a ver low second stage reduction (maybe up to 40 or even 50 to 1). The cost difference for the drive itself would be nominal. However, there would be more complexity involved in setting up a two sprocket outdrive system (dual freewheeling sprockets for pedal side drive).

Matt

 

[deecanio]

Amazing

i can't believe that is even possible.
one thing i did find with even just my puma was that you will accelerate wayyyyy quicker than you can change through the gears, even just using the three sprockets up front i was through them plenty quick so im not sure about having 27 :lol:
that said it does seem a waste to have the drivetrain there and not use it, optibike does the same thing but the motor is pretty tame in comparison to the rc motors we have been looking at so far, also there may be a major issue with chain/sprocket wear with that sort of power.
We also have the problem of noise, most cyclones i have heard sound really noisy - maybe we're best just having left side drive only - which begs another question im afraid, if the drive is a sprocket on the disk side how does freewheeling work? is just a one way on the output shaft?

Cheers Matt,


D

 

[etard]

I was wondering how a one-way bearing works and found this site most helpful.

http://www.rctek.com/info_bearings_one_way.html

Matt,
Are you using the Align T-Rex 500 bearings (H50020)? I ordered a motor with a 10mm shaft, these seem to fit my application, but they are ABEC-3 I think. Is there a better quality bearing? Will these withstand the forces of a bike? You're not lieing, 1 1/4" billet sounds massively ridgid.

So, you are mounting your setup down low behind the bottom bracket, looking at the pic of your bike:

Where? I always envisioned a top of the tire mount (ugly, and not stealth). Now, if you make a U-shaped box that wraps AROUND the bottom bracket, that would truly be a thing of beauty! 8) The issue I see would be the need for a spring loaded tensioner to keep the proper chain tension through the entire motion of the rear suspension.

 

[recumpence]

I will be mounting the drive in front of the bottom bracket (in front of the cranks). It will be mounted to the swing arm pivot bolts as well as a support strut at the top. Kind of hard to explain.

Anyway, there will be minimal chain slack alteration through the suspension travel. But, it will be an issue. Therefore, I will be running a spring loaded tensioner.

As for driving the right side; there will be no more noise than the left side. The complexity goes way up, though. But, you eliminate the disc brake issues.

I will be using a oneway bearing pressed into my outdrive sprocket for freewheeling. I am not sure how much torque it will take, though. The only experience I have with oneways is in RC helis. We used to run a 4mm ID (very tiny) oneway in the main gear. That drive system was 400 watts reduced 11 to 1 driving that tiny 4mm oneway. So, if you scale that up, I think the 1/2 inch ID oneway should be good. Worse case scinerio would be to mount two oneways side-by-side to double teh torque hanlding. That would cost more. But, not hugely so.

The point to this drive is to make the absolute best drive I can with the best effort at cost savings I can muster.

Matt

 

[GGoodrum]

Can't wait to see some pics of the shiney bits.

 

[recumpence]

It should be soon. Dave is running some parts for other customers right now. I am up next. This week should show some serious progress. Heck, he should love me. I actually pay up front. I may even give him a check tomorrow as teaster to get him going. He needs a slit saw for my project anyway.

Matt

 

[boostjuice]

I will be using a oneway bearing pressed into my outdrive sprocket for freewheeling. I am not sure how much torque it will take, though.

I cant wait to see some photos of this box complete!

I thought i might be able to share some experience on this question.

I started designing from scratch, a cam-clutch based freewheeling crank setup to replace the ratchet-pawl style used in the Cyclone freewheel cranks -Known for their fast bearing failure rate. I wanted to improve the durability/lifespan of the system without the noise of a ratchet-pawl one-way which would be constantly ticking away unless the rider was pedal assisting the motor. This was to be an improvement on the durable White-Industries ENO 22t freewheel that ES members often build onto their freewheel cranks to replace the Cyclone supplied unit.
My research indicated 3 options for different style one-way freewheels. Sprag clutches, cam-clutches, and needle-spring style freewheels.
Unfortunately none can compete with the small size/torque ratio of standard ratchet-pawl style used in bicycles.

You can rule out sprag's immediately because of their large size required for high torque, and the fact that they are designed for high speed overrunning based on the cetrifugal force used to disengage.
Needle-spring style often have no inner ring and so are designed to be built onto the shaft. The shaft then has to be hardened steel, any wear that occurs requires the shaft to be replaced. Not ideal for easy repair/replacement.

Cam-clutches looked the best choice. To keep the bearing rings narrow, I found you have to select quite large ID/OD versions to get close to the torque rating of a standard BMX/singlespeed freewheel (>200Nm?). (35mm ID/72mm OD/22mm WIDTH gives 173Nm).
BTW i was looking only at versions with built in dust seals and ball bearing support so parallel support bearings would not be needed.
Selecting some with wider rings can bring down the ID and OD significantly, but is difficult to build these into the narrow wall of a gear casing or sprocket without protruding past the outer ring support surface.
Anyway i ended up giving up trying to find one small enough that had the torque capability. I ended up just building a WI ENO style freewheel crank.
I hope your case dimensions don't limit you the same as me because the silence and increased efficiency offered by cam clutches would be a valuable addition to the gearunit.

Tsubaki make the most competitively priced cam clutches i could find. Japanese made. Keep in mind they are still VERY expensive compared with regular bearings and mass produced bicycle freewheels.

The BB series is what you'd probably want. They come in both keyed and press fit in the sizes i think you' d be looking at. you'll notice the size/torque ratio is much poorer than you might imagine.

http://tsubakimoto.com/tem/pdf/CAM_CLUTCH.pdf

Also another link to a company that make cam clutches integrated with chain sprockets.
http://www.daytonsuperiorproducts.com/pdfs/OneWaySprockClutch.pdf

 

[MitchJi]

Hi,

After I posted the question about the differences between the HV and SHV Controllers I decided I should check their Web Site ( ) and ask Castle instead. This is part of what I sent (I also mentioned that Matt recommends additional caps with the HV110):

What are the differences between the SHV series and the HV series other than the input voltage?

On the web I noticed two differences (other than price ). The HV series have estimated Current Sensing. The SHV have estimated Current Sensing with Measured Sensing optional. The SHV Series has a heatsink for cooling.

It looks like the SHV Series with the optional measured current sensing supports Hall Sensors. Is that correct? How much does that option cost?

Will it handle heat better in low airflow environments due to the heatsink?

Do you have a pdf that delineates the differences?

For a system running at 50 volts or under and roughly 2,000 to 3,000 watts max which do you recommend?

Is the availability of the entire SHV series 3 to 4 month's?

Do you ever plan to maintain them in stock or are they always going to be an advance order item?

This is their reply:

Input voltage, size, switch relays, etc . Totally different controller. It would handle heat better due to the heatsink, and the fact that it's not being pushed anywhere near it limits like the HV110's are. Options, costs, etc the bosses have not figured out yet. An HV110 will handle 2-3kw fine, but you would have to add extra caps as recommended. The ESC is meant to start an airplane motor...not a person on a bike. No estimate on release for the SHV. Until they're made I don't know whether they'll be in stock or on request

I was under the impression that the additional caps were required due to the 7kw Matt is pushing through the unit but I gather its due to the use on an Ebike.

Also a major reliability issue with the Crystalyte is moisture and the Hall Sensors. If David can provide a motor with Hall Sensors that don't have reliability issues with moisture I think that would be a desirable feature for use with the SHV Controllers.

Actually, I may go to one of the production drives for my recumbent as well. The production drive is a hugely rigid design.

Now thats exciting A drive that is as good as or better than your recumbent is something that seemed out of reach a short time ago.

EDIT: When I woke up this morning I realized that Matt is mounting the motor using the swingarm pivot point bolt not on the swingarm itself so I removed that. Also decided to remove suggestion for D.

 

[MitchJi]

Hi Matt,

One foundational aspect of my drive system is a 1 and 1/8 inch OD bearing. That is not alterable (this will become apparent when you see the drive design). There is no 5/8 inch ID bearing that is 1 and 1/8 inch OD. Of course, a 1/2 inch to 5/8 inch sleeve can be used so a freewheel can be used if needed...

The only question mark in my mind is the oneway for the output sprocket. I am not 100% sure a 1/2 inch ID oneway will tolerate 5kw at 16 to 1. We shall see!

I found a bearing that might be suitable with 1 1/16" od with 5/8" ID if you think that might help. This is the link (the SSRI-538EE):
http://www.alpinebearing.com/torque-tube-radial.asp

Two versions, industrial ($9.50) and Stainless ($34). The industrial model has Teflon Seals that do a better job sealing but have a little more drag. The Stainless model uses a metal seal with no drag but there is a small crack so the sealing isn't as good.

 

[deecanio]

Hi All,

Mitch, i think Matt is mounting his reduction box just in front of the cranks on the downtube? and is using the swingarm bolts just to help mount the box, not mounting to the swingarm itself as such.
Of course if you mount the drive to the swingarm there are no issues with chain tension however the box will take a pounding from the suspension, i'll be using a tesioner myself and mounting on the seatube.
Personally i'm set now on left side drive, i will use a standard dh rear hub (10mm through axle) with disk and have a standard cluster on the rear, even though i know i wont use the rear cluster as my gear change it will be nice to select any number of rear cogs 11-34 and then just leave it there and still use the front three for my gear changes.
I think everyone's setup will vary - all we need is some pics from matt so we can all do our working out for our particular setup, one thing for sure there will be some awesome bikes out soon that will take ebikes on to where they should be
one thing i'd like to ask - probably Matt more so - what sort of size would my rear drive sprocket be on the disk side?
i know it depends on all the gearing etc but just a rough estimation would do so i can research hubs better? when i say size i mean diameter in mm? im just trying to work out how much space i'd need and whether the sprocket will be smaller than my disk, im planning to put the disk rotor back to 8" ? do we have any links to the sprockets? are we using #25 chain?
BTW im planning to run the hv110 at 5kw so i'll let you know how that goes, i think it will be fine and given the availability and price of the svh i think best to try the hv110 first.
Matt, can you post up the drawings Dave is machining from just so we have an idea what we will be working with????
pretty please?


Cheers,

D

 

[recumpence]

I hve run the HV110 up at 4kw with my AXI and had no issues. But, at 4kw on my Pletti, I had issues after 200 miles. I wondered about that and asked Castle about it. They said there were a number of issues that contributed to my controller failures. When I asked about extra caps, they basically told me I would be masking a problem by adding caps. Hmm, if that is the case, why do these ESCs have caps to begin with? "Oh, good point, yes I guess extra caps would help...." So, since running extra caps, I have had no problems. I also experimented with my slipper clutch to keep from overamping the controller. However, I still blew one controller just limiting my amp pull to under 3kw max. The caps made all the difference. Maybe I will become a dealer and stock a few cap modded HV110s for EV use.

I will try to draw a couple pics for you guys to see what the drive will look like. It is hard to explain because it is very weird in overall construction and tough to represent in two dimention. When I have time later, I will try to draw it in a decent way so you can get an idea of what it looks like.

Basically, my recumbent drive system is way overly complicated and very time comsuming to make. However, I had never built an E-bike, so I had no idea how to do it. Now that I have experience with this type of drive, I was able to eliminate what I did not like and make the drive totally dedicated to space conservation, modularity, and adaptability to various frame layouts.

The drive systems foundation are the bearing tubes for the jackshafts. They look very similar to bottom bracket tubes on a bike frame, aluminum tubes 2.5 inches long with a bearing pressed into the end of each tube. These bearing tubes are held to each other and the motor plate by split clamps like fork tripple clamps. The secondary chain tensioner will use a split clamp as well for easy swinging to adjust the tensioner. There will be basic mounting plates included to be bolted or welded to the bike.

This drive looks like a patentable design considering its very unique layout. But, when all is said and done, I will have roughly $2,500 spent on the first few of these drives to get the production going. So, looking into any kind of system patent will be out of the question.

With the economy looking poor, E-bikes will only get more popular. I have no interest in high numbers of cheapo crap. But, I am very interested in getting involved with the high-end of the market if it goes well.

Matt

 

[deecanio]

Hi Matt,

If you are finding that the cost is getting very high to fund the initial run, i for one would be happy to give you a deposit or something to help lessen the cost burden?
When you added the extra caps to the hv110 what sort of power did you pull through it? i guess blowing a hv110 is a lot less painful than blowing any svh model financialy?
Really looking forward to seeing the drive appreciate time constraints may not allow the drawing so don't break your neck trying, its a nice to have, not a necessity
do you have any idea of the rear sprocket size yet? we using #25 pitch for the final drive? (i say we as i will mimic your build cog for cog) ;p
rock on Mr Shumaker 8)


Cheers,

D

 

[CNCAddict]

Hmmm, I haven't done much research on installing sensors in my motor. A sensorless controller should work fine if it's designed properly, but if you guys really need sensors I can get a cheap ebike controller and start experimenting with the best way to mount them.

The problem with sensored controllers is that they are usually fairly primitive. Motors need some fancy software to run them the best at all speeds/torques, and the sensorless controllers are very advanced in this area.

 

[recumpence]

John in CR already sent me a deposit. I do not want to take any more, though, because I want to get products built, shown, and ready to ship before I take payment. I figure it is not a big deal because as soon as these are available, I will probably get a couple grand worth of orders anyway. That will get my money back quickly. What I really want to do is just get everything rolling. CNCAddict is sending me a 40 series motor and a one-off 60 series to test. What I would like is to maximize two separate reduction units using the same framework with different drive components. One for the 40 series with a 2kw peak, 1.5kw max sustained output and one for the 60 series and Plettenberg motors with a 5 to 6kw peak and a 3kw sustained output. The main difference between the two is the primary belt width and pitch.

Anyway, all is looking good. I am just frustrated at the wait. I have gone through this many times with my heli products (waiting for outsourcing and R&D). But, no matter how many times I go through it, it is always annoying. :wink:

Oh, I figured a simple way around the 5/8 inch ID freewheel adaptor and my 1/2 inch shaft. I will order some 1/2 inch ID, 5/8 inch OD steel tube as a spacer. I can, then, MIG weld the spacer to the end of the 1/2 inch shaft for those who want it. If I weld it at the outside face of the two, no shaft warping will occur and it should be plenty strong.

Matt

 

[recumpence]

I just spoke with Dave on the phone. He will begin running my parts tomorrow morning.

I may not have a whole lot to show you guys because this structure is comprised of 5 CNC machined parts and Dave is machining part #1 in quantity, then part #2 in quantity and so on. He may only get one of the parts completely made (in qantity). But, at least you guys will be able to see some progress.

Shiney Bling is coming soon!

Oh, Man, I should anodize a few drives. That would be very cool!

Matt

 

[deecanio]

i'd pay a bit more for black :lol: :lol: :lol: :lol: :lol: :lol:
nice one Matt.


cheers,

D

 

[recumpence]

I was thinking of black too.

I think anodizing would be about $50 or so. I will check on that this coming week.

Matt

:EDIT:

UPS just dropped off some Stock Drive bling! I got my prototype pulleys, sprockets, chain and belt.

Man, this thing is going to be a bit bigger than my recumbent drive. The large belt pulley is 4 inch diameter and the #25 secondary sprocket is nearly 5 inchs in diameter. This drive will be setup as 16 to 1 to begin with. I can go down in motor pulley and small secondary a few teeth each for a 24 to 1 ratio. Any lower and these parts will be larger. If the drive is setup lengthwise (remember, it is adjustable in layout) it will be 11 inches long overall (dual stage) or it can be setup in a triangle configuration for a more compact layout. It can be chainged in layout very easily. Single stage would be about 7 or 8 inches long.

My recumbent is running at only 7 to 1 ratio right now and it is fine. This drive needs to be 16 to 1 roughly to drive my mountain bike properly.

Matt

 

[MitchJi]

Hi David,

Hmmm, I haven't done much research on installing sensors in my motor. A sensorless controller should work fine if it's designed properly, but if you guys really need sensors I can get a cheap ebike controller and start experimenting with the best way to mount them.

The problem with sensored controllers is that they are usually fairly primitive. Motors need some fancy software to run them the best at all speeds/torques, and the sensorless controllers are very advanced in this area.

Don't bother with this based just on my suggestion. On his Cross Canada Thread Justin mentioned one of the features of his controller (not in production and wouldn't work with RC Motors anyway) is that it will switch automatically from sensored mode to sensorless when there was a problem (a common occurrence with Crystalyte hall sensors when exposed to moisture). The consensus was that controllers are sufficiently better in sensored mode that this is a valuable feature. If the HV and SV Castle Controllers work as well in sensorless mode (or close) as in sensored mode forget it. Or if nobody else requests it forget it.

More information on Justin's Controller here:
http://www.endless-sphere.com/forums/viewtopic.php?f=1&t=7134#p107193

 

[CNCAddict]

Hmm, yeah a controller that does the sensored/sensorless flip or one that even uses an encoder for input would be best. Something along the lines of http://www.granitedevices.fi would seriously rock, but his output current is pretty pathetic right now. Maybe I'll shoot him an email.

BTW, that Haro Sonix looks awesome...what did you pay for that? The thing about bikes with all these crazy shaped frames is that it doesn't make a universal drive system any easier to mount. But who doesn't like another challenge 8)

 

[MitchJi]

Hi David,

BTW, that Haro Sonix looks awesome...what did you pay for that? The think about bikes with all these crazy shaped frames is that it doesn't make a universal drive system any easier to mount. But who doesn't like another challenge 8)

I paid $430 including shipping.

I think its a good choice for the Serial Configuration I want to use (one chain from crank to gearbox,one chain from gearbox to rear sprockets) because there is no chain growth between the crank and the rear sprockets.

If I figure out a clean mounting solution on the swingarm it should work well.

 

[etard]

Matt,
This is all good news! Let me be the first to claim a RED ANODIZED drive (oh man, I guess I need some red rims with black spokes). HAHA, that would be epic....how much extra are we talking? Also, will you be providing final drive sprocketing, or should we be prepping our steeds for this? Matt, would it be too foward to request some sort of system to designate production records. I think it would be cool to have a laser etching, or stamp on your artistry, some sort of representation that you crafted this contraption.

CNC,
Will your motors come in different color options or is that way over the horizon? Are you guys working with through the motor axles, or one side only? Dave, same for you as Matt. Please consider this drive to be a piece of art, and your motor will be the heart of it.

You know, we are entering into the same realm as Gary Fisher, Ned Overand, and all the pioneers, except we have claimed energy back from the mountain, preserving its sanctum. Honored, I will be to one day pass gassers on the trail and spit a pebble or two in jest as I dissappear silently.

March on soldiers

 

[GGoodrum]

You know, we are entering into the same realm as Gary Fisher, Ned Overand, and all the pioneers, except we have claimed energy back from the mountain, preserving its sanctum. Honored, I will be to one day pass gassers on the trail and spit a pebble or two in jest as I dissappear silently.

March on soldiers

Well put... :wink:

 

[recumpence]

My guess would be somewhere around $50 to $70 per drive unless alot of them were done at once. No big deal. I used to have my Heli parts anodized all the time, but that was a while back. I have to see who does it in our area currently.

Hmm, engraving. How about anodizing the drive first, then engraving through the anodizing? I have one pulley for my helis that is made that way. Very pretty looking with the machined aluminum showing through the anodized surface. 8)

At any rate, I do not mind doing stuff like that assuming people want it. It just adds lead time and cost, but anything is possible. If nothing else, I can buy a simple hand engraver and freehand sign and date each drive. I even have number and letter punches somewhere that punch 3mm characters into aluminum.

This is alot of fun, to be sure. However, I have seen this with my heli parts too. There are a handful of guys who are really excited at the beginning. But, I need to make sure there is enough interest in this before I really go all out amking a bunch of stuff. For now, I am assuming there is modest interest in a high-end drive. So I will make small production runs of a half dozen drives or so at a time to fill the immediate demand. What may be best is to have Dave make a bunch of drive "Chassis" partsm then I can order pulleys and sprockets per order. That would keep inventory cost down and would ensure each customer gets what he needs specifically.

Oh, I went over my recumbent drive last night............... That drive is only 4 to 1 to the output sprocket plus 3 to 1 to the wheel with my Plettenberg gear for 40mph. That is easily accomplished with a single stage drive. Dual stage will make for some stump pulling reduction!

Matt

 

[recumpence]

Ground has been broken!

Dave just called to confirm a final revision 10 minutes ago. At this point, chips are flying!

Matt

 

[deecanio]

oooooraaaahhhhh

Hi Matt,

how many of each type are you putting together first time round?
etched through anodised would be soooo cool, i can see silver through black now 8) - but lets get the boxes done and working on bikes first hehehe ;P
11" is long in the stretched out format, can't wait to see how your going to make it changeable either - its amazing it can do that, will give a lot more mounting options.
im so eager for those first built box pics im chewing my knuckles now.


Cheers,

D