Safe's Electric Bike Project #001

[safe]

Grinding Error... Argh!!!!

My transaxle is made up of normal wheel bearings. The problem is that in the long process of hacking the thing together I managed in the last reconfiguration to weld and then grind very near the actual bearing surface. I "blew it" and allowed the grinder to get too close to the bearing surface and it's made a "pit" which now allows some play in the transaxle bearings.

I think at this point what I really need to do is replace EVERYTHING with a new hub. Take a generic front hub and then weld a freewheel cluster base directly to it and don't get anywhere near the bearings in the process.

Argh...

Everything is "fine" except for this "pit" that allows some play in the bearings. It's simply a manufacturing flaw, but as a result I'm probably going to reinvent the thing for a THIRD TIME...

You learn as you go...

It still works fine, but I hate imperfections like play in the bearings... I could try to weld and grind some more, but that seems like an endless process that does not have a good ending. If I had a lathe then I could do it, but without a lathe it's impossible..

 

[Lowell]

I bought a lathe for the shop, and then got a smaller one for the house so I could work on little projects anytime... :lol:

 

[EbikeMaui]

Cool bike. How much voltage are you running? What kind of motor, controller, etc..., are you using?

It's your basic "plain vanilla" 36V 750 Watt Unite motor with a typical unmodified 40 Amp limit controller. People seem obsessed with fiddling with the controller so as to bump up the current limit, but you're better off having gears and just "downshift" when needed. Better to downshift and hum along in the ideal powerband than throw current at the problem and turn your motor into a "space heater". :wink:

I notice that you are running a hub motor. That pretty much limits you in the gearing department, but if you wanted more power you could always add another hub motor on the rear. Given the weird ways that electric motors use power you would actually do better with two hub motors than with one both in power output and in overall efficiency. (it would be a wiser strategy than raising the current limit which is probably correct for what you already have)

Gears are the future though... or LARGE motors like an Etek...

Now this is something I can agree on. :lol:

 

[D-Man]

I notice that you are running a hub motor. That pretty much limits you in the gearing department, but if you wanted more power you could always add another hub motor on the rear. Given the weird ways that electric motors use power you would actually do better with two hub motors than with one both in power output and in overall efficiency. (it would be a wiser strategy than raising the current limit which is probably correct for what you already have)

I have taken your advice and installed a couple more hub motors.

 

[xyster]

I have taken your advice and installed a couple more hub motors.

Too bad you can't ride photoshop...

 

[TylerDurden]

Too bad you can't ride photoshop...

 

[safe]

The last few days I've been building a new "transaxle" after the old one developed a "pit" in the bearings due to some sloppy machine work I had done before. Often a small "pit" will work it's way out, but in this case it simply got bigger and bigger until it was no longer running very well. I'm also trying some stronger grease that is for bearings rather than the wheel grease I was using before. It seemed like the bearings were overheating and that was adding to the wear. The old grease would essentially melt away and turn into a liquid. This is version 3.0 of this "transaxle" and I hope it's the last. If this doesn't work then I'm going to have to get away from bicycle components and into something like a sealed bearing. My next project will not need a "transaxle" and I would not recommend it to anyone. Gears most definitely "yes", but a "transaxle" I have to say "no".

The new tires are "excellent" and give good grip and I've broken them in finally...

The results are that the top speed is back to where it is supposed to be at about 40 mph on the flat. I rode two "moto's" today, the second one was for about 10 miles and an average speed of 26 mph. Not too shabby...

 

[safe]

Average Speed on Today's Ride

On my (second) ride today:

Distance - 9.45 miles

Time - 20 minutes and 42 seconds.

If I did my math correctly that's an average of 27.391 mph.

 

[Lowell]

I don't think any normal bicycle driveline parts will hold up reliably to ebike sport riding with an average weight rider.

Good to hear your bike is back to it's old self though. Keep the work in the shop going on the new bike 8)

 

[safe]

I don't think any normal bicycle driveline parts will hold up reliably to ebike sport riding with an average weight rider.

The chain has been shown to be overbuilt for the demands of a normal bicyclist. They've recently been reducing the width of the chain on these cassette style hubs that carry the 8,9 and 10 gears on them. I'm pretty sure the chain will be fine. (assuming you use a pre-cassette width chain 3/32") The power output is also smoother than pedaling (peak pedal power can be huge, 1500+ watts, even if the actual average output is only 400 watts) so you can run the extra power without breaking anything.

The problem I had was a simple grinding error. I knew that I "blew it", but hoped that with time the bad spot would smooth itself out.

So far this new transaxle has been incredibly smooth and is showing no signs of producing any play. It's likely this will be it.

I'm optimistic that there will be no need to wildly diverge from basic bicycle components, however, there will be some areas that will need to change. The fork is one area... it's simply not acceptable to have flex in the fork or the frame at high speeds.

The wheel bearings are fine.

The chain is fine.

The fork and frame need changes. (normal bike frames and forks are too flimsy)

The rims needs better "grip" onto the tires and need to be produced with little ridges like motorcycle rims do. That's a trivial change, but it will keep the tires from slipping on the rim from heavy braking.

The tires need to be built better and have higher speed balancing. None of the tires I'm able to find are perfectly round and true. The 24" x 3" tire is considered a "toy" tire and not something that needs to be able to handle heavy stresses at 60 mph, but it will have to for this to be really well done.

I might switch from 1" headset bearings to 1 1/8" headset bearings if I have to, but so far I have been able to survive with 1" headset bearings. (at least there's a choice in the matter)

My freewheel cluster 14-28 made by shimano has been bulletproof, but the origianl junk chinese made one failed, so sometimes simply replacing junk with quality can solve the problem.

Right now my bike is about as perfect as it can be for what it is... (hope it lasts)

 

[safe]

Version 4.0

Can you believe Version 4.0?

The bearings are fine, but since I made this out of a stamped steel bicycle hub the shaft that holds the two spoke wheels (whatever you call them) was not held together with anything. So I was out riding and all of a sudden it was like I had a clutch and it was slipping. (the two halves of the hub were sliding past each other) I managed to make it home and the alignment was still okay, but I had to weld a thin sheet of metal between the spoke wheels so that they are now permanently connected. Seems that I managed to do it without messing up the alignment or the bearings. Whew... I dodged a bullet on that one.

New paint, new grease, remounted and went out for a second ride and all seems fine. The transaxle has been a source of a tremendous amount of effort... literally days and days (probably 10 days) of effort have been wasted on the transaxle.

 

[Beagle123]

What size are those batteries?

They are 38 Ah each, so the total weight comes to 86 lbs.

How is that possible? What kind of batteries are they? Brand? Model? My SLAs are about that size and they're 17 Ah. Do you have a link?

 

[Lowell]

15.9Wh/lb sounds right on the button for SLA...

 

[fechter]

SLA batteries come in all sizes. I've seen some 100 Ahr ones that were the size of a truck battery.

38 Ahr would be a typical wheelchair battery. I've got a couple of nice Hawker 26 Ahr ones, but they're too heavy for anything I'm building.

 

[safe]

I bought mine in what might have been a closeout sale. They are similiar to these batteries, but not exactly the same. I got three for $132. They are selling them for $50 each now. Mine are only rated 38 Ah verses the 40 Ah these are listed at. It's a lot of batteries and you would need to build a very, very strong frame and fork like I did to pull it off. Beagle123, if you are an inexperienced frame builder are you sure you want to trust your life to uncertain skills under such intense stresses? My advice is to stay away from so much weight... my next project will be 60 lbs lighter which will make everything easier to deal with...


http://www.batterymart.com/p-12v-40ah-sealed-lead-acid-battery.html

 

[Beagle123]

I stand corrected. They look much more compact on your bike.

I may take your advise on framebuilding. I'll run my plans by you when I get them together. I'm becoming more convinced to switch to lighter batteries (not SLA). I agree that it will make life easier. For one thing my mountain bike suspension won't need alteration. Bicycle parts will work perfectly (.i.e. tires/rims). Also, I was planning to cut pieces of aluminum myself, then hire a welder to assemble it.

I'd like to make a battery box run down the center of the bike like you have that's made of L shaped aluminum. But instead of connecting the top rail to the bottom with braces, I'd like to cover the entire box with sheet aluminum. This would make a box-shaped tube that would be very strong and resistant to rain etc. I would connect each corner on the front of the box to the tube that holds the forks, and cover those triangles with aluminum too.

The image included shows the design using 4 17ah SLAs. The tube could be smaller with Lithium batteries. BTW tires = 20 inches. Seat height = 33 inches

I'm still getting my ideas together.

 

[safe]

Steel is a great material for testing with. The reason this is true is that steel goes through various phases before failure. First you will be able to notice flexing of the steel while in use and in some cases you will be able to observe stress related effects such as cracked paint in and around the joints. The flex will tend to increase in severity until an actual break occurs. You have some time to notice a problem is developing.

Aluminum is a wonderful material. It is very light weight, it's very rigid, it does not rust. The negatives about aluminum are that you have less feedback from the material before it breaks. Aluminum tends to remain rigid right up until you get complete failure. Carbon fiber is like this only worse. So companies that design aluminum and carbon fiber frames need to do "destruction testing" where they place the part into a test device and literally break the part in order to learn what the limit is. They then factor in some degree of a "safety factor" and use that.

So I'm just warning you that you will become a "crash test dummy" if you are not careful. I've built frames that failed miserably before. My first frame was a "monocoque" design that used thin sheets of steel and the connections on the edged came apart. It flexed really bad too.

The "bottom line" is that for an amateur frame builder you are best served with the most conservative design. The absolutely most conservative design is a space frame using small tubes like I did. With so many small tubes if any one joint were to go bad you could notice that something was going wrong and be able to correct it. (with an extra brace)

I've gotten really lucky in that my frame has been bulletproof from day one and the forks as well. However, I have done this a lot before. (I used lower grade mild steel on my bike... but...)

My advice:

Stick to steel tubes... preferably 4130 CHROMIUM MOLYBDENUM:

http://www.airpartsinc.com/products/4130-steel-tubing-round.htm

Weight is simply NOT that critical on an electric bike so the difference in weight between chromolly and aluminum is insignificant.

 

[safe]

Version 4.0 Fails!

Seems that the bearings were fine, but the housing that holds the gear cluster is hanging off the end of the bearing.... so what's happened is the entire transaxle has bent. (the stress of the offset gear has twisted the entire gear cluster slightly) The result of this was pretty spectacular as I was going 42 mph in 6th gear and the chain jumped out of alignment, jammed, twisted the derailler and the chain guide, broke the chain and forced me to walk the bike home. (which was fortunately nearby) Luckily the chain didn't get stuck into the rear wheel or it might have been even more dramatic.

So it looks like Version 5.0 is in my future for the transaxle...

This time I'm going to get away from bicycle hub parts and really think the thing through from scratch. All I really need is two sturdy sealed bearings held within a tube that the gears bolt onto somehow. It should not be too hard, but it does mean that for that part I'm back to the drawing board. (and I no longer have a ride)

It's too bad because the bearings are really perfect this time, but with the drive gear now askew it's not really usable any more. If I try to throw it all back together it will eventually repeat the problem, so I'd just be breaking stuff if I did.

I've just passed 1550 miles.

This transaxle has been a real source of problems...

 

[xyster]

Version 4.0 Fails!

This transaxle has been a real source of problems...

Surrender your will to the temptations of the hubmotor and be done with it!

Might be a lot "safer".

 

[safe]

I've been thinking it's time to try out my existing 8 speed hub. The new project is getting going and it will take a month or two before that's progressing to the point of real testing. I've done enough of my homework to realize that the hub motor can never equal the geared bike if you are staying within the 750 watt limit... which I want to do more or less. A lot of the work has been spent on the transaxle trying the same thing again and again and every time the same bearings or hub components fail. Actually the last one was bent too and so the bearing failure on that one might have been related to the same problems as in this one. The transaxle will not exist in future bikes, so it's an area of creativity that doesn't advance my long term objectives.

I still like to ride... and so part of it is that I just want to keep this one alive until my new project comes online... :wink:

 

[CGameProgrammer]

So have you given up on using a transaxle in the long term, but will attempt to keep the current one operating in the meantime?

 

[safe]

So have you given up on using a transaxle in the long term, but will attempt to keep the current one operating in the meantime?

From day one I disliked the transaxle, but it was the only way to get the gearing right. With the new motor I've got (MY1020Z3) it has a geardown that drops the rpms to something that is easy to handle. When you are starting with a raw rpm of 3000+ rpms it takes some effort to get it down to the 300+ rpms that your rear wheel needs.

Gearing is great for motor size limited bikes... but the technical issues are not trivial to get right if you want the motor to drive through the gears.

I just want this bike to become "stable" so that I can ride it while I build the new one(s). The transaxle has been my principle source of problems for a very long time.

The derailler and freewheel gear set have been surprisingly good. I figured that they would have been the source of more problems, but ever since I installed a Shimano rear freewheel things have been really good on that end...

 

[xyster]

Why not just chuck the transaxle and stuff your gears between bolted plates spaced 1" or so and welded to the motor side of the bike?

 

[safe]

Why not just chuck the transaxle and stuff your gears between bolted plates spaced 1" or so and welded to the motor side of the bike?

That would involve some redesign... hmmm.... I might.... I've thought of doing that on some of my new projects, but I've somehow overlooked that on this bike.

Very good idea...

Even just keeping the transaxle mounts, but flipping the motor around backwards so that both chains are on the same side would eliminate the twisting torque that's causing all the problems. (and is sure to be adding friction)

 

[safe]

The Best Idea

...is to simply eliminate the transaxle entirely.

I can go 11/90 (scooter chain) or 14/114 (go cart chain) to the rear wheel and then use a Shimano Nexus 7-speed hub ($129) and adapt the sprocket and a disc brake to it. This particular hub is made of steel and looks by appearances to be easy to work with.

So I've nearly completed "Version 5" and if it fails as well (which it very well might) I'm going to trash the transaxle concept entirely and go direct drive from motor to rear wheel.