Suzuk-E 

1987 Suzuki RM125 Conversion - Part 2 ...

Dec 2008

Part 1 concentrated largely on tearing down the bike and liberating the motor from the gear-box.

Since I still didn't have a clear image in my head of how the batteries would be mounted I spent a few days just putzing about on some detail work like getting the front brakes working.

The original brake lever was missing, so the clutch lever was drilled out, cut to shape and filed to act as a replacement.

The images below are the motor RPM and rear wheel speed measurements taken to compare against the calculated values as posted previously.

The last column are the measured values adjacent to the calculated values for 4200 no-load RPM off the motor shaft.

All the measured values are proportionately equal to the calculated values that I feel comfortable that the math is close enough to say that it is correct in the spreadsheet.

With the math verified, I reviewed the data on how the original Dirt-E calculated out and compared it to it's GPS level ground max of 24Km/h and came-up with a Load RPM of approx 2500RPM. So the objective is to improve on that number through the use of the gear-box and re-worked motor combo in this build.

I decided to start preparing the gas-tank for housing the electronics, for some reason I wasn't sure if I really wanted to use the original tank, but it turned out to be a fantastic housing.

This is the part I'd been dreading, having to actually start cutting the frame.

Initially I'd hoped that I could have oriented the batteries identically to how they are on the Dirt-E project, but there simply isn't adequate clearance to the gear-box.

Another option was to stretch the frame about 2 to 3 inches, but that would have created issues between either the seat placement or the gas-tank positioning.

The T-square shown here is almost exactly the side profile of the battery bank if aligned side by each.

This configuration is not exactly what I wanted, but will have to do for now as I can see getting side tracked easily.

The hardware set in to place here can easily be cut-away and reconfigured in the future to accommodate newer battery technology of mounting schemes. 

Looking at the image to the left, I think that I'll add some diagonal bracing back toward the front of the rear shock mount before paint is done.

I can live with this set-up...

With the center of gravity nice and low the bike feels good in the short test runs it's been out on so far. This is especially noticeable as riding conditions are ice and snow, where a top heavy bike would be a real hassle to ride.

The motor mount is beginning to deform and actually has a small tear where the metal has been fatigued by the constant gear changing. 

The Drive sprocket is holding-up well and seems to have worn into shape quite nicely.

All the previous electric type bikes that I've been working on to this point have used very cheap switch hardware compared to the application that they were being employed for. The largest switches that I'd been using were 35A automotive bat type switches that worked ok, but always made me a little nervous that they weren't up to the higher currents that were being drawn through them.

The switch assembly pictured above is inspired by the mechanical switches that I've seen on 1950's Marketeer golf carts that are still working today, the white pieces are made from a 1/4" thick sheet of nylon type material while the copper bars are ground plate stock found in small(ish) electrical distribution panels.

The long horizontal bar slides between the 2 vertical bars to create a secure high current contact. The black strips over the vertical contact bars are 3 layers of inner tube rubber that holds the contacts together under tension. I originally wanted to re-use the clutch basket springs but they are just a little too large.

The switch is a little bit stiff but it makes a satisfying arc as the contacts are engaged and likely could withstand well over 200 Amps of current if this bike gets a larger motor in the future.

This is an array of 3 (three) 30Amp thermal breakers that are wired in parallel to create a 90Amp breaker should the controller or any of the wiring short out.

In previous tests of a single breaker it took several seconds at 75Amps to pop, so this should be able to allow peaks well over the 90Amp rating but provide the piece of mind that a dead short will be mitigated.

These breakers self-reset after about 15 or 20 seconds, which should be just enough time to kill the main power switch.

The copper busses that connect the breakers is fashioned out of household 1/2" copper pipe. The long leg of the lower buss bolts directly to the receiving contact bar on the main power switch.

The gas tank slash storage compartment is nice and neat...

The bike has been out for several test runs, in progressively colder and snowier weather... I'd hoped that there would be a break and thaw so I could check the bike against the GPS to determine the top speed of each of the 6 gears...

The GPS values would be compared to the calculated speeds to determine the "Loaded" RPM of the motor, and then the motor can be re-worked to squeeze just a little more out of it and I'd have an objective metric for comparison.

But the weather and my Home-brew controller didn't co-operate. The PicOx initially was working reliably, until I started to advance the current limiting towards the 60Amp Mark in the firmware... 

There is a huge difference between testing against a static load on the bench and having the bike bouncing and loading down the controller over varying terrain.

The image to the right is a test area that the bike is driven straight into and hooked directly to the PC for program tweaking...

After a few days of chasing noise issues that lock the controller as I approach 75Amps, I've opted to dump the PicOx for now and use a 100Amp (Peak) Yi-Yun YK42-4 commercial controller from TNCScooters.com. It was a tough decision, as I really wanted to use the PicOx, but this project seems to be just beyond it's capabilities.

The PicOx Ver 1.02C would be quite reliable At up to 36V 75Amps or 48V @ 45 to 50Amps Peak, so over the winter I hope to move forward on the Ver 2.xx Picox and apply what I've learned from this project... So hopefully this isn't the last of the PicOx Controller.

Over the next few weeks the Suzuk-E will be torn down, for paint, some minor revisions in the drive train, addressing the lubrication issues of the gear-box and refurbishing the MY1020 motor.

For more details on this section, check out the link for the Video below...

Back to Energy

Disclaimer (an unfortunate necessity)
All Rights Reserved theworkshop.ca © December 06, 2008