Hillzofvalp's 29er Thread... now with custom 63T (cromotor)

[hillzofvalp]

I CNC'ed a splined coupler that lets me mate a standard 11T or even 10T cassette cog to my White Industries freewheel. The two main reasons for making this were that I wanted to both avoid a derailleur because with horizontal dropouts it is redundant to have a derailleur AND to run a much smaller gear with a nice freewheel mechanism. It also let's me adjust chain alignment more easily and have arguably cheaper maintenance ($10 cog).

Aluminum is just a prototype.. but should work out well. I got the spline pattern very snug the first time around. The last cog on a cassette has much more surface area, so I'm not too worried about stripping the splines.

If anyone one wants one, I have the CAMing done and can spit another out in less than a couple hours, so let me know if you want to do business . If I ever did an RC build I would easily add the mounts for 130bcd road rings...

The reason the splines have rounded grooves is that the smallest endmill I had (1/8") would not get all the material if I didn't overcut. Worked out splendidly. Again, if anyone wants one.. maybe I can make a business out of it.




Pretty soon I'll have my welder done so I can finish my pack. I just got the weld pressure sensor working, but I have yet to finalize the main circuitry.

[hillzofvalp]

update: the coupler works very well. I'm very pleased with it. Solved all my issues with drive train.... AND I can pedal and contribute up to and probably beyond 30mph. Good ratio with the 11T.. perfect alignment.. no play (at least yet).

Welder is sorta working.. going to fix it when I come back in a week. Got a second hyperion 1420i for balancing the larger pack.. but I was disappointed to find that compared to the other 1420i I have, this one is reading a .1V difference from the same supply. That's a lot, especially for balancing. We'll see if it holds true for balancing. The other big disappointment is that I will have to split my pack to charge both packs. I figure now I won't be doing balance charging very often (since my pack is enclosed in)... and I'll have to finish up the 72V bulk charger project and stick with that for the time being.

OR I can put in contactors and shit.. yeah right.

[hillzofvalp]

update: gave in and got a Cycle analyst. I don't think I'll be disappointed.

[hillzofvalp]

http://www.flickr.com/photos/30556588@N ... /lightbox/

Pics above. Just lost my 500 word post typing on a IPad. Gist is that I she tested well at 14s in the snow... Working on finalizing everything with 20s.

[hillzofvalp]

I machined a lower profile controller (1.2-1.25x5.893x3.1") so that I can squeeze in another 5p block or two into my enclosure. Sounds like a lot of work, right, to just fit in 5-10 more cells, doesn't it? It is. Don't do it.

Side benefit is that I will get better heat transfer to my aluminum u channel framing, and the is more aluminum near the fets for more thermal mass over all.. I haven't decided between running a heat pipe to the outside of the frame, or making a specially shaped wedge that will allow me to sink it to its mounting position on the aluminum u Chanel. Also, i may consider oil cooling, as now I have better control over getting a good seal.

The nice thing about working with the 2 inch u channel is that there are a lot of materials already sized for it. I bought some Velcro for keeping the batteries from sloshing around (not tying them down because the snug fit keeps them I'm place. I also got some 2 inch 1/16" thick rubber adhesive backed strip to line the downtube and other areas to protect the battery shrink from contacting the frame and potentially eventually wearing off and causing a short (though the 100A fuse should take care of that).

I also got some grommets and very soft reusable waterproof seals for weather proofing.

I'm excited to have the pack balanced and wired-- I just got to think less about the cromotor!!!!! Need to make lid for controller enclosure and run 10 or 8 gauge instead of the stock crap. I'm also thinking about some graphics underneath the the clear polycarbonate...I want to make it cool to offset the stupid bulkinees


Does anyone know if there is a brake light trigger pin on the lyen 12 fet? I thought I read about it once but can't pinpoint a location

[hillzofvalp]

Figured out how to copy image links on iPad!!

While I had the controller apart, I went ahead and put in some 12awg deans for the phases and some 8 gauge that matches the battery pack. May cut down on some heat.

I haven't made a lid for the the controller yet, but I was just going to pound some .05 inch aluminum on it (with the board out of course) to get the conformity I want. I really just want an indentation in one area where I removed the battery pack cap. This will let me fit in everything in my triangle enclosure.

I ordered some sample connectors from molex... Ones that include both 24 signal pins and 4-6 blades for charge leads. The idea originally was to have 4 main leads: 2 for breaking into the middle of pack for Hyperion sync charging, and 2 for full potential normal leads. But then I thought, why not make the charge port also serve as an emergency disconnect.. So the cap for the receptacle will actually be another connector that shorts the mid-pack connection. This is why I tried to get a connector with 6 blades: 2 for main charging, and 4 for mid pack low resistance connection.
http://www.molex.com/pdm_docs/sd/464373007_sd.pdf

I want a cromotor, but I will have to wait until march or April when they are more affordable.

[hillzofvalp]

I'll have to say that 20S is a lot more fun--- up until the axle spins in the dropouts! I'm fashioning a torque arm and want to get some input (if anyone actually posts in my thread anymore). I am hoping to water jet some 1/8" stainless I have around.. Let me know if you have suggestions. I was going to bolt it threw the rack eyelet, one disc slot, and in one of the pictured gaps





I don't know how common this type of spin out is, but notice how it left the slot and jacked out the dropout! (nut is flush with axle end). Phases left unscathed amazingly. I was on slight hill stopped at an intersection. It happened from dead stop with my pedaling hard. The uneven torque on the axle from pedaling hard one one side probably caused it.. And I shouldn't have had grease on the nord locks. Wd-40 only I'm now told. Oops

[hillzofvalp]

I see 4-5 people coming to this thread but not commenting. I'll give u each 5 ES bucks if u comment.

[Idontwanttopedal]

Can't wait to see how a cromotor going to go in a er29. This motor making think twice about big rc build and try my first hub build

[hjns]

I don't need 5 ES bucks. However, if you get me another 10S Lipo....

With 20S and full acceleration you are in need of two heavy torque arms. So you noticed...

And with a voltage of 84-72V, I would recommend a kill switch and a maybe a fuse as well.

[hillzofvalp]

Yea.. I'm working on it.cant decide over arms or making something to epoxy on... But i made this prototype which I probably won't use at all

See the torque plate thread for me discussion on this

[John in CR]

Dude, it's steel. Just weld some extra thickness to the dropouts. Make it thick enough to run 5/16th good grade bolts for clamping and be done with it. Do you have an angle grinder? Welding the thick stuff is easy, so practice a bit on some scraps and then do it. You're into DIY ebikes. You're on a good track with a steel frame already. You're spending the bucks on all kinds of other stuff including $700-800 just for a motor. What's another $200 for some more metal working equipment and supplies.

John

[MadRhino]

Well said John.
I would add that 1/8 thick SS torque arm will not help much to hold a powerfull hub motor.
On my first build, the X5 spinned through 3/16 SS both sides after only 500 km.

Make them dropouts overkill.

[John in CR]

My first hubbie spun out too thin SS torque arms in the first 10ft of the first ride. The SS gave very little, but cut grooves in the axle like it was butter.

Think of the width of any decent wrench. Now visualize the bolt or nut heads that are only 10mm across, and then consider the entire force of motion for your bike being transferred through those little nuts as turning torque. Regular nuts and bolt heads are going to deform in no time, as will the wrenches, because it's literally thousands of pounds of force acting at that tiny radius. At a minimum you'd want double length nuts and custom wrenches with a width to match.

Just like phase wire thickness, you'd have to get pretty crazy to be overkill. On Hubmonster my 20mm wide clamping dropouts look kinda puny, and definitely aren't close to overkill, but I believe they'll get the job done.

Hubmonster clamping dropout.JPG (47.9 KiB) Viewed 606 times

[MadRhino]

...On Hubmonster my 20mm wide clamping dropouts look kinda puny, and definitely aren't close to overkill, but I believe they'll get the job done.

To someone who thinks that he can get away with a 1/8 SS torque arm, your dropouts might look way overkill.

[John in CR]

...On Hubmonster my 20mm wide clamping dropouts look kinda puny, and definitely aren't close to overkill, but I believe they'll get the job done.

To someone who thinks that he can get away with a 1/8 SS torque arm, your dropouts might look way overkill.

1/8 SS is exactly what cut through my axle like butter in less than 10ft of my first hubmotor ride. Of course my weight and the 25% grade didn't help matters. Thank goodness I had them well strapped on, because it was a front motor and the alloy dropouts snapped off like toothpicks, so had they not held the motor on, my hubmotor career may have started and ended with a faceplant in the asphalt.

[hillzofvalp]

in an effort to make charging and balancing a much more seamless process, I have received free samples of these 60A molex connectors. They have 24 signal pins I will use for balancing.. If I ever went up to 28s I could easily add another 24 signal block(or less if they have it). Samples we free.shipping was free. CAD model access allows me to easily 3d print a grommet/mount that will nearly seal the bike.

So the plan said now is to run 4 connectors to mid pack.. 2 on each side... Cause Hyperion chargers don't like to sync charge with pack connected at midpoint. The reason for using 2 is to minimize resistance and maximize current capability for say a cromotor or dual motor setup. The the controller will be bullet connected to the pack...And the outside pins of the connector wil be for positive net and negative net charge leads (bulk charge leads).

I got an extra male to use as a cap and emergency disconnect.it will connect the 4 pins together and seal the port when ready to ride. Still haven't figured out precharge method. Both halves of the pack will each receive a 100A fuse in the event of a connector failure. Is will try to upload an image of the mount that I'm constraining in Catia.. I thought of doing it out of Al but decided that wouldn't make any sense from a time standpoint, cost standpoint, or electrical safety standpoint. So glad my school has 3 3d printers...... And I can do projects on them through the EV club ( no personal projects allowed)


I also may note for Hyperion users: it appears the power leads connect BEFORE the balance leads. Disconnect should be safe also. Makes sense cause these were designed for server supplies generally

Edit: It may actually be worthwhile to Put a few cell logs on the bike and integrate them into the cap

[hillzofvalp]

Regarding torque arms. I'm still stuck between clamping the axle vs. thick torque arms. I don't really have space for thick torque plates from doc, cause I cut the axle on my crap 9c. However for cromotor I will most likely use a modified version of docs part

[John in CR]

When is someone going to come out with a cheap but reliable single cell charger? That's the route to seamless charging and balancing of Lipo, but no one uses it. Integrate Methods LVC/HVC protection board somehow for safe quick bulk charging and we'd be good to go. Hopefully these kind of solutions come by the time we can readily get NMC's. They need balancing don't they?

[hillzofvalp]

Yeah.. It's only a matter of time. This becomes a practical solution for lifepo4. In reality I don't need to balance every charge, but it's nice to log what's going on.. Cell by cell. This is more focused On having emergency cut off without a power consuming contactor. It just s happens I need to split the middle of the pack anyways.

With this type of connector you could more easily parallel a big bank of lipo for charging.. Only would require some diodes and a single 14s Hyperion for a full 28s Pack. Not ideal.

I would really like a smaller bms then the current ones. That's where ebikes are headed

[hillzofvalp]

The added benefit I didn't really realize fully is that I can. Now easily bulk charge with a lower voltage psu. 36V instead of 72. Same C rate but much more utilizing my 55A supplies.. Just make a special connector that links the right things.. Don't have to disconnect wires and open up 15 security hex screws

[hillzofvalp]

The reason I made the torque plates .5" is that I need .5" to actually thread a pinch bolt through. I will mill off 1/4" so that the net width is .5" instead of .75".

Also below is the port that I 3D printed. I am very please with the results. Molex has drawings for all their parts online, so it was very easy to use it as a reference. Fits perfectly.







Since it would be a pain in the ass to make a new top for this controller -- and since my enclosure is going to be completely sealed (already is everywhere except two spots at very top), I don't think I will go through the hassle. It looks cool this way too. It's heatsinked to the frame with thermal compound.



In this picture, the polycarbonate is on. In a future revision, I hope to reduce the side panels an the aluminum edges. Before I weld them together I will mill them down to about 1/4" high. Now, at least, I have a good reference template. When I first started, it was very hard to get the geometry/shape of the enclosure right without having a prototype.

Velcro strap holds the top most pack. The bottom long pack sits on 1/16" rubber.. snugly. All other packs are velcroed to each other.

I love how it looks! 20S5P A123

[Samd]

Anything gonna get hot and penetrate the wrapping at a flashpoint on that lipo?

[hillzofvalp]

It's actually A123 LiFePO4. critical areas are bumpered with rubber. Most of them are double and in some cases even triple wrapped with shrink.

Because it is well-tabbed, I don't think I'll have issues with a hot point on the pack, especially at 55A peak. I do need to test some of the tabs to see when they begin to saturate

I intend on adding another fuse for the second half of the pack... mainly because of my charging methods.. though I suppose it will also protect against this sort of failure, if the hot spot is very localized, anyways

[hillzofvalp]

been waiting in line for tesla motors for almost an hour at this job fair... Appropriate time to update my build thread.

Torque plates are complete (hopefully). See pics for how they will be welded on.

The molex balance wires were a bitch to connect... But should work great. I can now use network mode on the Hyperion chargers