A Newb and his Yuba/Stokemonkey

[kevbo]

My current job would need about 3hr/day of commuting time if I did it on a muscle bike (and probably 3.5 or more if I rode the Mundo) But I am thinking I can get that down to less than half with an E-bike.

The question then became shall I get an E-bike, or modify one I already have....six rideable, plus various broken ones. I rode my Yuba Mundo a lot when I first got it, but due to changing life circumstances, it has set virtually unridden for a couple of years. The Stokemonkey concept intrigued me, but I was doing OK on muscle power, then they were unobtainium for a while. Since kits are now available again, I decided to pull the trigger.

Someone is bound to bring it up: Yes I saw the thread on Yuba Mundo fork failures. I have a Big Dummy fork on it's way.

So I am installing this on the Mundo, and I ran into a couple of issues...no big deal, but thought I'd post it where others might see it if they run into problems, and also to provide some feedback to the vendor.

One thing this kit really needs is a packing list telling me exactly what should be included. I was (maybe?) short a primary drive chain, and the axle bolts (& washers) for the motor. Not a big deal to source locally...the bigger issue was just wondering if they were supposed to be included, and what else might be missing? My understanding is that some inginuity and improvising is to be expected with such a kit...so just because a needed part isn't there doesn't mean it is missing.

The next issues came up, I suspect, because the vendor is now supplying mounting brackets radiused to suit the Mundo seatpost , rather than just V clamping surfaces. The front plate (the one counterbored for the capscrew heads) on mine had a larger radius than the middle part. The supplied screws also proved to be a bit too long. Maybe the screw length worked with the old V'd blocks? I'd suggest that the next batch of clamps have tapped through holes rather than blind holes...that would make the screw length less critical, and the motor would mostly hide any protruding thread...also possibly easier on the machinist.

I thought about either buying some shorter screws but (metric screws are available, but still not trivial to source in the US...stupit bib box stores killing all the real hardware stores. Or I could trim the screws a bit, and hope there was enough thread left.

Then I got to thinking about that too-large radius, and decided to see what would happen if I added a shim. I found some 1/8" x 1-1/2" Al in my goodie pile, and used a vise to press it to shape using the clamp piece and a scrap of pipe.

Yay, the screws now fit. 1/8" is maybe a bit thick of a shim, as it doesn't quite bottom in the groove now, but close enough for me...also the screws are still long enough that they will bottom out if I don't get them evenly tightened, so the right shim would most likely require shorter screws.

Next issue: Somewhere I read that the clamp interferes with the front derailleur cable, but it will still work. Maybe with the old V-notch, but the radiused clamp would require the cable to go through solid aluminum.

Maybe I should get a bottom-pull derailleur? No go on the Mundo: The chainstays merge into a single ovalized tube that is the full width of the bottom bracket, so it isn't possible to run a shifter cable up from below. Also no cable stops on the downtube, and since it is massively oversized and ovalized, a clamp-on cable stop isn't an option.

I thought about drilling or notching the clamp to clear the cable, but there isn't a lot of meat between the seatpost groove, and the motor mount groove....then inspiration struck: If I incorporated a cable stop into the clamp itself, all would be good. So that is my solution.

 

[kevbo]

If anyone wants to duplicate this, a bit of advice is called for. When you have to drill deeper than 10 diameters or so, special technique is needed. This job requires drilling a 1/16" (1.5mm or so) hole to a 2" depth...32 diameters! That is not a trivial task, A 1/16" drill is just waiting to break on you, so do everything possible to stack the odds...I am passing along what I was taught by my machinist friend:


First off, use the right tools. If you can manage this using a hand-held drill motor, you are both highly skilled and lucky, and you don't need my advice. So use a drill press and vice bolted down solid to the table.

Speed: You need a bunch. 6000 rpm would be about right for a 1/16" drill. (100 sfpm) Most machines won't turn that fast, so just turn it up to 11. If it is a "real" drill press, then you will want to run it at the highest speed available . A dremel tool in a drill press accessory should work nice, and a low speed will be plenty fast. Use a 1/16" collet in a dremel to hold the drill, because the drill chuck accessories tend to be rather eccentric.

First off, the hole location: If you hit either the cross-groove or the the bolt clearance hole, you will most likely break the drill off in the hole. So measure carefully, center punch, and/or spot drill so that little bit doesn't walk on you.

Coolant: I like to use denatured alcohol (ethanol with enough methanol added you won't be tempted to drink it) on aluminum. It cools and prevents the chips from gumming up...doesn't really lubricate. An eyedropper full would do this whole job if you can get it into the hole without waste. Yes it is flammable, and the flames are not very visible, so use appropriate caution...or better yet, use a proprietary cutting fluid specified for aluminum.

Use a screw-machine length drill, or choke-up on a jobber-length to start with, this will help the hole start off straight. Once you are in a 1/2" or so, you can use the full length, which you will need every bit of. I had only about 1/8" in the chuck to finish, and you can see I dinged the top of the part with the chuck jaws. I actually had to switch to a slightly larger diameter drill, because my 1/16" drill wasn't quite long enough. If you try to drill from both ends, you will almost certainly leave a step , and there is also a high risk of breaking the drill when breaking through in the middle...and that is even if you measure both sides perfectly, because the holes will drift a bit with depth.

If the speed is high enough, it should take very little pressure to feed the drill...you should just barely be able to tell when it is cutting. If you are having to push significantly, something is wrong with your drill, so get a fresh one. Back off and clear the chips at the first hint of increased pressure or noise.

When you start drilling, just peck at the metal until you are deep enough the drill can't walk, then you can go about 1/10" at a time, withdrawing the drill fully from the hole to clear the chips...as long as you can still see the flutes. Once you are deeper than the flutes, only take maybe 25-50 thou at a time. If you cut too much at once and the flutes pack with chips, the drill will jam and snap off. You have been warned.

Once you have the 1/16" through hole, it is a simple matter to counter-bore it with an appropriate drill for the cable housing. Doing this while the part is still in the vice will keep the counterbore centered on the through hole.


I know I am wordy, but I have probably broken more tools than most folks have owned, and like to pass on what I learned the hard way.

 

[kevbo]

Well, I got the beast rolling last night, and went for a ride in the dark. The rubber wheel on my bottle dynamo had split, but I pressed the rubber tire from a model airplane wheel into service and had light.

It rode nice! I messed around some with the CA settings. Need to figure out how to set up the ramping...pretty much nothing for the first bit of throttle, then a big jerk. I'm running it with the throttle controlling amps for now....that makes the most sense to me.

I need to come up with a way to mount the batteries. I'm running two 51V x 6.6Ah LiFePO4 batteries in parallel. The huge panniers don't restrain them enough, and they want to go off the edge of the running boards (still inside panniers).

Any advice on setting up the CA for the Crystalyte SAW405 motor that came with the stokemonkey kit would be most welcome.

 

[footloose]

Nice solution with that shim, and beautiful job on the derailleur cable routing.
Thanks for describing how you machined that 1/16" hole.
I've always figured that slow rpm was best. Apparently wrong.
May explain why I've broken so darn many bits.

It is a shame what big box stores are doing to real hardware stores.
I tried to get some basic barb wire staples at a big box last weekend.
The kind I expected to scoop by the pound from out of a bin.
They didn't have them -- not in bulk, not prepacked, nada.
How lame is that?

 

[kevbo]

I've always figured that slow rpm was best. Apparently wrong.
May explain why I've broken so darn many bits.

With a tiny drills, if it is going slow, then it tries to cut too deep at once at any manageable feed rate. With bigger drills you DO have to slow them down.

I'm glad someone appreciated that novella.

 

[Chalo]

Thanks for describing how you machined that 1/16" hole.
I've always figured that slow rpm was best. Apparently wrong.

One concept you have to internalize to become a machinist is "surface speed". That's the speed your tool's perimeter is moving; the fastest part of the cut. The smaller the tool, the slower the surface speed is at any given rpm, so you have to speed it up proportionally. The ideal surface speed varies by material and the presence or absence of lubricant/coolant. You can look it up, or even use an online speed and feed calculator.

The other thing to get your head around is "chip load". That's how thick a slice the cutting edge is biting off as it turns. Too little can be just as bad as too much. For hand fed drilling, the main thing is to pull a continuous string of waste material, such that the cutting edge doesn't wipe the material without making a chip. Either be feeding or retracting the tool, but not letting the tool dwell on the material's surface.

 

[kevbo]

Well, I got the new fork on, and a few other odds and ends...far enough along that I rode it to work today. Prior to this, I had only done a few 3 mile uphill runs to take the pooch to the off-leash park. There was hella headwind, (~15 mph) which was no problem with stoke monkey helping. I was using 300-400W to maintain 15-16mph ground speed on the level.

I think I am going to set up the cycle analyst with a switch to select power limits 250W, 500W, and full power. 250W is enough help for my mostly downhill to-work. 500W should get me home no problem, and make the motor last forever. Mostly don't want to become too dependent on the motor: The goal was to make bicycle commuting feasable time-wise, not to let my ass get any fatter.

I can charge at work, so I'll have a fresh charge for the climb going home, and can park it overnight with ~50% SOC, which should keep the battery happy.

Turns out I work with another ES member, so we had a nice chat when he spotted the Mundo in my office.

I ordered a comm cable from Justin, so hopefully that will allow me to enable the temp sensor function in the CA. The outside of the motor is barely above ambient when I arrived at work, but I worry about what is going on in the copper.

Edit:added photo showing Surly BD fork, temporary dynamo mounting kludge, belt-and-suspenders brakes.

 

[Chalo]

You know... in terms of efficiency, weight, points of failure, and generally not being awful, a DC to DC voltage converter hooked up to your battery is a much better solution than a sidewall dynamo. There's a place for a sidewall dynamo, but a 21st century e-bike is not it.

Are you using a dual cable brake lever?

 

[kevbo]

Yes on the dual pull brake levers. The dual brakes idea came from hating to
Look at the canti-stud "warts".

I've had the bike about five years now, the bottle dynamo has worked well, but yes, a Buck converter is obviously the way to go. Maybe I'll make one, and give it a couple USB charging ports as well.

The Yuba fork had a couple of 5mm braze-ones on the inner leg of the fork, which was how the dynamo
Used to mount. Maybe they were for a front hub motor toque arm?

Anyway, The bottle dyno was all wired, and known good, and the phillips safelight it powers is superb. The worst thing about it is the mozzy noise, The motor is noticeably quieter than the dynamo. Bottles get a bad rap partly because there is some really bad ones, and people fail to align them just so for minimal drag....and of course Bart Simpson.

Mainly wanted to get it on the road and see if my battery sizing was going to work out for my commute. Yesterday I rode home about as fast as practical and that took a bit less than 9AH of 13.2 available. That included a light headwind most of the way, soI should be fine on battery.

So I need to work out battery mounting, motor drive mounting, chainguards, and powering the lighting via the propulsion battery.

 

[kevbo]

This may not be an original idea, but here is how to make a single cable gear sensor work for both F and R derallieurs.

The photo shows the gear sensor located under the top tube where both shift cables make parallel open runs. This should also work under the downtube on typical road bikes if the cables are cross routed as is common to make the housings off the bars a bit neater.

The gear sensor is designed to mount in a run of housing, not in an open run as I have done here. A bit of brake housing has been installed at each end.

The rear shift cable routes through the gear sensor in the normal fashion.

The front shift cable routes outside the sensor, and is secured to the housing bits on either side.

This causes the whole sensor to move along the RD cable when the FD is shifted. Note the loop of slack I put in the sensor cable.

The red bits are electrical butt splices that I crimped onto the FD cable to give the zip ties something to grab. The jacket on the brake housing is soft enough that the zip ties get a good bite. Make sure both housings are snug in the sensor. Double check before you crimp the butt splices (or whatever) to the outside cable that the position will not cause the cable housings to jam against the cable stops on the frame...check full travel of shifter.

 

[kevbo]

In addition to getting the gearsensor working, I fitted a Grin 6V output DC-DC converter, so have ditched the dynamo.

I also rigged the reed switch from a cheap cateye computer as an ebrake switch. With just the cable travel moving the magnet, it was very fussy getting it to make and brake reliably. So I rigged up a lever made out of a spoke to increase the movement of the magnet, and it is working FB now.

I also attempted to get higher gearing by fitting a DNP 11-32 freewheel. That was a no-go. The DNP sits about one cog farther outboard than the Shimano freewheel, and the derailleur doesn't have enough travel to shift to the highest cog. Also I think the chain pins would at least take the paint off the dropout if I were able to get it to shift. I toyed with ideas to to increase the derailleur throw (I think a little dremel work would do it) Maybe come up with a longer bolt and bushing so I could put a spacer under the derailleur. Maybe add some spacers under the locknuts and cold set the frame.

I wasn't liking those options. I checked ebay yet again, and lo and behold there were a couple NOS Shimano 11-34 buy-it-now at a fair bit less than an arm and a leg. So I snapped one up...it is currently making it's way from the UK. I think I may need to do something to get a little more chain wrap, and also keep the chain from dragging on the center stand if I run using the two smallest sprockets. (feasible on a longtail...though increases wear)

I'm thinking of adding a third jocky wheel like suntour did in the early MTB era:

Edit: I found a SRAM X4 derailleur new for less than $20. It is an upgrade over the X3 I have, and offers 45T of wrap, vs. 37 for the X3...that should be enough, but dragging on the center stand might still be an issue.

 

[kevbo]

Here are some photos of how I set up the eBrake switch:

The bent spoke in the foreground pivots at the elbow end where it is zip-tied to the frame.

It is linked (another zip-tie) to one side of the straddle cable for the rear brake. That is a swaged splice because I needed more than 330mm to clear the fender. The grey housing runs back to operate the disk part of my "belt and suspenders" brake setup.

The other end of the spoke carries a magnet to trip the reed switch:

The other side of the straddle cable supports the magnet so it doesn't grab the frame tube. Shown above in the un-tripped state.

When the brake is applied, the spoke swings forward so the magnet trips the reed switch:

It is hard to see in the photos, but the magnet moves by about twice the cable travel. This allowed easy adjustment (bending spoke) to get reliable on-off action.