Phoenix Rising: P1

[Kingfish]

AW, those lights look way cool!

Been a short while since I last posted an update so here it goes…
The Blacksmith’s hammer & anvil have not been silent, no. Like a Damascus Sword, the ideas have folded over and over again to forge a better beast from raw material and off-the-shelf product. In the past month there have been four prominent changes:

• DIY Torque Arm
• Fender upgrade, & Rear Rack added
• New water-resistant faring

The faring is the one topic not yet covered and has been the last modification applied. The image below is how the battery bag & bike previously appeared.




Black & White Faring
This new frame has a slightly different triangle shape with wider tube members, especially the top tube which has a taper that is widest at the front and narrowing as it heads aft towards the seat post. Technically the distance between the steerer tube and seat post is farther away to accommodate the rear shock. Where P0 had an exposed cable running across the top tube, P1 has the opposite condition running along the bottom of the down tube. The problems with the battery bag were pretty much the same as with P0 before adding the faring:

• It was wide enough to cause chafing of my winter trousers
• The batteries were getting wet (although for a time during the sub-freezing weather this was less of an issue)
• Excessive road splatter created a cleaning nightmare.

I knew it was unfinished; something had to be done, and I waited until an opportune weekend to do it. After stripping off the plastic bits and rope, the (formerly) top flap of each camelback bag was cut off, essentially slimming down the profile by nearly a ¼ inch per side. Next I added protective tubing to the bottom cable in precisely the same manner as with P0 using old Homebrewer’s spare tubing and Black Duct Tape. I also took the opportunity to restring the battery bag forward about 1.5 inches so that I could access the front shock air attachment for future adjustments. The image below displays this part of the story as completed and before the application of new faring materials.


Stripped, clean as a whistle...

I detest water spraying up onto my batteries. Borrowing material left over from the Epic Battery Bag Story, I took a large shingle of black 42 oz. Vinyl-coated Polyester and cut out matching Left & Right pieces that would cover the exterior, like flexible water-proof armor. Each side was uniquely fitted to completely cover the exposed front, bottom and sides, requiring minimalist sewing to create the conformity with three modest seams. Extra material was provided at the top where the halves met so that the two pieces could be sewn together. The remaining edges were punched and secured with grommets. Sturdy shock line was used to fasten it into place. Except for the bottom where fastened, the bags should now be well-protected from water. This heavy vinyl covering looks in the pictures as if having a diamond-plate pattern; that is in part deception upon the manufacturer for the reinforcing weave is heavy though the pattern is faux. It is tough stuff to punch through and does not easily rip or tear!

The next part of the armoring is the HDPE. The Black Vinyl does a great job at protection but I needed some bright color for better visibility. Yellow doesn’t look right on this bike, though White does. The covering is slightly smaller than what was used on P0, mainly due to geometry and best fit. Once trimmed up, it too was punched and secured with nylon line.

The last bit of HDPE took the most time. I used Black to create a better-fitting forward cowl and was able to keep it in one piece along the entire length of the bottom. The top of the cowling wraps over the top tube and is modestly secured with zippies whereas the the bottom uses Velcro. Funny stuff Velco; the forward section where it meets with the White plastic is stitched in place whereas the rear section is super-glued. I think if I had to do this again I’d forgo both methods and use Urethane instead. Regardless – there have not been any issues to report; the faring assembly fits flush and is quite aerodynamic: it feels like I’ve picked up another mph in the slipstream as wind rolls right over it.

The final addition to the faring is the 1-inch wide Reflective Striping. For some reason I only had a yard of this stuff so I put one along the trailing edge of the White faring to visually mask the overlay, and the remainder went on the front faring. It’s very reflective in all lighting conditions. The sticky-back is likewise very grippy even on Polyethylene.

Catwalk Gallery: Take a bow. Apologies for the filth – it is winter and I do ride the bike nearly every day.







Stop back soon: More mods are in the queue.
WOT, KF

[Kingfish]

Oh time to update this project! The last picture posted was the FWD version that was used during the Winter and Spring of 2011. In May, I began the redesign and construction of a Push me–Pull you FWD eBike and pusher Trailer in preparation for my next road trip to California. Below are pictures of that assembly prior to launch.


FWD eBike with two controllers to support the pusher-trailer when attached. This version is technically labeled P3 because I have completely rewired the whole bike from end-to-end with new handlebar controls, new motors, new 12-FET controllers, indicator signal lights, active taillight, integrated eBrake & Stop light, two CAs, and trailer cabling interface. All battery-operated components run from one battery, and use the DC-DC converter to power the auxiliaries. I also employed a key-lock switch to completely disconnect the electrical power. In addition, I replaced the Meanwell S/SP units for bulk-charging with the faster HRP-600 Series which required zero-modification. New hardware included disc brakes, custom torque arms, and a new Brooke’s saddle.

Later I swapped out the chainrings, crank, and bottom bracket for a Campagnolo 53-39-30 Road Racing chainring & crank, a Phil Wood BB, and a Campagnolo Braze-on Derailleur. In order to mount the derailleur I designed a custom MtB Braze-on adapter. This provided an available gear ratio as high as 53-11. The views below include this modification and are taken just before my first (false) start on my journey to California.


Top View: Note that the right pannier is not mounted (cos I have the camera out).


Rear ¾ view. In 2011 I changed out my Ortlieb panniers from yellow to black to match the B&W theme.

Alas – there were problems with the pusher-trailer, namely the framework twisted axially and created an extremely unpleasant situation above 5 mph. This configuration was therefore abandoned. However – I was able to keep the pulse alive on the road trip by removing the hub motor from the trailer and mounted it instead onto the eBike, creating a 2WD in the process. The trailer frame was stiffened by injecting foam into the tubing (this is all covered in the trailer thread). After a couple of weeks or reintegration and bug-solving the assembly was ready again and I launched a successful road trip to California and back as a 2WD eBike pulling a traditional trailer covering 2500 miles over 28 days. This eBike configuration I label as P4 because I had to make and hardware wiring changes that were unique and specific to the activity. The design continued to evolve On the Road as I resolved persistent nagging issues one-by-one. Below are pictures of this assembly from the road. Each motor has 2 hp programmed into the associated controller.


July 26th, 2011 - Spanaway, WA: First day on the road to California. Look how spiffy it was! Difficult to believe I am hauling 100 lbs. of LiPo.

Changes include the addition of fenders front and rear, rear torque arms, rear freehub was replaced with a hub motor with an 8-speed freewheel. The front wheel is 26 inches using the Mavic EX 729 Disc rim, the rear and trailer wheels are 24 inches using the Kris Holm Unicycle rims. All tires are 2.5 inch wide Hookworms. The Harley-Davidson tool bag is mounted over the rear tire, although I moved it over the saddlebags a couple of days later to reduce the handling problems and wobble: The rear suspension can’t take a lot of weight mounted above the axles.


This is picture of the assembly just before I crossed the Columbia back into Washington after 27 days On the Road. The rear fender was a pesky biotch the whole way cos the bike frame was not designed to support fenders. The modification I created split the fender into two parts: Rear over the tire, and a short front splash-guard which protected the linkage and front derailleur. When the suspension would pivot in the extreme, the rear tire would crash into it and make a horrible grating sound; most annoying. Eventually I had to remove the front section and that eliminated 80% of THAT problem, although there were many other issues with the rear suspension et al. Though essentially useless, the aerobars made the bike look even more badass and were a convenient place to hang my helmet


August 22nd, 2011 - Bremerton-Seattle Ferry crossing over the Puget Sound. Regardless of the pesky issues, this design worked very well. Although I had one flat tire, there were no show-stopping failures; lots of mods – yes, lots of tinkering – sure, just no failures.

At present I have the trailer disconnected and the Ebike is used for commuting. It will do 42 mph on a flat, although I prefer to ride at 32 mph.

Cheers, KF

[chroot]

*edited* Nevermind, I found your information in other thread. I am going ask my girlfriend do the custom sewing 2 of the backpack together as saddle bag.

[Whiplash]

COOL! Do you have any pics of the trip? What kind of power consumption were you seeing? 100# of LiPo?! How many AH total was that and what kind of range did you get? Sounds like a fun ride! Wish I could take a month long vacation and not loose my job! LOL!

[ambroseliao]

Fenders, both F&R are X-Blade, purchased from REI. I saw an interesting set over at another bike shop, might have been the Specialized dealer (yeah, it was cos I had to fetch the Specialized Rim Tape) they were made by Topeak.

The part that I don't like about mine is that they are not straight and true: There is a noticeable twist from front to rear like they didn't make the mold properly and when the plastic was released it still had pliability to it and cooled with a contorted twist. If you're mudding it won't matter.

I would see if you could get one mounted before you buy it to see how it looks if looks bother you (like they do me)

Cheers, KF

I have the same front fender on my bike and it is also twisted somewhat. I wanted to relay that I once left my bike sitting out in the sun all day long. When I got to the bike, the X-Blade fender was completely warped and distorted and looked more like roadkill. I took it off the bike and thought it was destined for the recycling bin. However, after a few hours in the shade, it returned back to it's original shape! Leave it out in the sun for a while and see if it will "undistort!"

Ambrose

[chroot]

Last time, When KF visited in San Francisco, I was surprised KF mentioned, and He carried 130Ah total of the LiPo on his touring ebike! Here the picture in other thread about his longest trip (Washington state to California then return Washington state. viewtopic.php?f=1&t=28573

COOL! Do you have any pics of the trip? What kind of power consumption were you seeing? 100# of LiPo?! How many AH total was that and what kind of range did you get? Sounds like a fun ride! Wish I could take a month long vacation and not loose my job! LOL!

[Kingfish]

Fenders, both F&R are X-Blade, purchased from REI. I saw an interesting set over at another bike shop, might have been the Specialized dealer (yeah, it was cos I had to fetch the Specialized Rim Tape) they were made by Topeak.

The part that I don't like about mine is that they are not straight and true: There is a noticeable twist from front to rear like they didn't make the mold properly and when the plastic was released it still had pliability to it and cooled with a contorted twist. If you're mudding it won't matter.

I would see if you could get one mounted before you buy it to see how it looks if looks bother you (like they do me)

Cheers, KF

I have the same front fender on my bike and it is also twisted somewhat. I wanted to relay that I once left my bike sitting out in the sun all day long. When I got to the bike, the X-Blade fender was completely warped and distorted and looked more like roadkill. I took it off the bike and thought it was destined for the recycling bin. However, after a few hours in the shade, it returned back to it's original shape! Leave it out in the sun for a while and see if it will "undistort!"

Ambrose

All plastic fenders have a slight bothersome twist to them. The Aluminum plated with CF have the least twist. For my most recent build, I was quite happy with the Electra Alloy Balloon Townie Fender Set; the Front was mounted on the front of the eBike and the Rear on the trailer; the Rear was too wide though for the rear of the eBike frame.

Last time, When KF visited in San Francisco, I was surprised KF mentioned, and He carried 130Ah total of the LiPo on his touring ebike! Here the picture in other thread about his longest trip (Washington state to California then return Washington state. viewtopic.php?f=1&t=28573

Link is Correct; also in my sig

Best, KF

[Kingfish]

Upgrades between November 2011 and September 2012

Greetings –

I haven’t posted on this thread in some time, and yet what I have done to this ebike in the past few months is shy of a complete new build, and it's tempting to call it P1-Mark II or something of that sort. Anyways, it’s the night before I finish the assembly and put the fairings back on, so I thought I’d spend a little time and talk about what’s new under the covers – which pictures and stories. Briefly, I have listed the items that I have modified, reworked, repaired, upgraded, investigated, or otherwise reviewed below. The most significant changes are in larger text.

• Hub motors:
  
  • RUST R&R, varnished, vented, new seals and bearings. Done.
  • RH Phase Wire through Axle upgraded to 12-AWG, and replaced damaged HE sensors. Done.
  • Correct Controller/Phase/Halls wiring for 2WD:
    With typical Infineon-based/clone Controller = A-B-C,
    Using 9C 28XX Series F & R Hub Motors having Left (RH) & Right (FH) exits,
    And corresponding Hall Effect Sensors,
    The best performing relationships are (color agnostic):
    
    • FH: C->M, A-A B-B C-C & C->H, A-B B-C C-A
    • RH: C->M, A-A B-C C-B & C->H, A-A B-C C-B
    • The relationships can be swapped for C & H, so long as the pattern is preserved.
• Battery Harness completely reengineered: The primary goals were achieved for the Commuter pack although some details remain for the secondary packs.
• Controller Signal Harness redesigned. Initially this was to resolve missing regen on the RH:
  
  • -EBS apparently only goes to GND on one controller. One idea was to explore putting a diode in-line, and a plan was made to do just that – however upon inspection of the PCB, a diode is already in place. The next probable cause was poor GND quality. Thus the rework attempts to move GND dependencies from the Controllers and instead use direct GND from the battery pack; reduce the length of wire and upgrade wire gauge.
  • This rework also includes the incorporation of the PCI Express 8-Pin Extension Cable; a more robust connector interface between shared functionality and the Dashboard. In this concept, there is no Master-Slave; both controllers are essentially identical with exception to Power: FH is de-rated by 20% to prevent spinout at the starting line.
  • The entire dashboard wiring & Black Box has been simplified.
  • +5V is now shared for Throttle; there is a diode in-line on the PCB, and we do not foresee a problem with two sharing the miniscule load created by the Throttle.
  • Except Throttle, all connectors have been replaced/upgraded with more robust quality.
  • All signal wires are now minimum 18-AWG.
  • All harnesses are now sheathed in Nylon-Braid Expando Cable Sleeves for protection, natural lubricity in a dynamic environment, and for consistent appearance.
  • The L/R Signal & Headlamp Dashboard Control also has a horn button; however the wires were overtly unattached to the connector. We anticipate the employment of a small light horn in the future; therefore the dashboard control is now wired to support the accessory when it is acquired.
• FH Controller Current Mod (was Master) Evidently I missed upgrading the traces, so these were beefed with 10-AWG twisted wire.
• HE Throttle was replaced following the accidental short of the RH Controller during testing; more below. Added two plug types to accept both EBikes.ca and Lyen/Molex type connections.
• RH Controller Replaced (was Slave) and upgraded to match FH Controller. Lyen calls this a Version 2 (MII) controller. There were slight changes noticed. I took the opportunity to match both controllers with identical Capacitor upgrades, trace-beefing, and wiring. RH does not have the R12 mod because of the impending Prototype project described below.
• Move RH controller to Left/Port, and reverse for FH. The Phase wiring may need to be shortened. This change eliminates all wiring passing through the fulcrum of the rear suspension.
• R12 Progressive Brake & Regen Mod investigated. Parts were ordered/received for the first prototype board. Construction is in progress. The new Controller Signal Harness has a new pigtail incorporated to facilitate attachment of the prototype directly behind the fairing for easy access, measurements, and adjustments.
• High-Powered LED Headlights Upgrade - Bust:
  
  • The idea was to swap out the 5-way flashlight for Single-Mode due to flakey circuitry: Hit a bump and the lights would change mode.
  • Problems with supplier and replacement product: DinoDirect took approximately 33 days to ship product. When the units arrived, they were much smaller than anticipated, and unfortunately cannot be parted out; they are 100% incompatible with original lights.
  • However, it was also determined that the flakey nature of the lights could be from the poor quality of the power connection. The automobile-style bullet connectors were replaced with recycled APP connectors. Given the low current and high stability of the fairing – we think the problem – at least with connection quality has been resolved.
• Bar-End Handlebar extensions (replaces Aerobars). Reoriented LED headlights to allow for closer spacing. This arraignment conflicts with the two CAs; see below.
• Unified CAs. Removed mounting hardware from RH CA and glued the top frame surface to the bottom frame surface of the FH CA. The two units now tilt/toggle together, and are more closely organized for better viewing. The new alignment better affects placement of the LED Headlights, and provides more room for the Handlebar extensions.
• Handlebar Hardware, Controls, & Instruments: Because the previous-style Aerobars were removed, there is room to reorient the LED Headlights: The mounting hardware was reversed and trimmed to extend the lights more forward on the handlebar, thus freeing up space directly behind them which in turn allows the newly unified CAs to angle for better viewing. Without this hardware mod, the CAs couldn’t have been joined. Mild trims were made to the Front Fairing to accommodate new instrument positions with the fit more snugly improved.

Present Status:

• The electricals are finished: New Battery Harness, Controller instrumentations, and DOT Indicators and subsystems are mounted and tied down.
• The motors have been mounted and tested. The disc rotors have been trued up and brakes adjusted. Tires are pressurized for touring. Both shocks are topped off.
• The Commuter Pack has been laid upon the bike and charge topped off, but is not yet tied down.
• The front and side fairings will be added next. By the time I got to mounting the batteries is was already close to 5 PM and I know it would be dark before the task could be completed, so we shall finish this last bit of construction in the AM and perform some test rides before moving on to the Prototype Brake/Regen project.

Links and briefs and pics to related stories will be listed shortly.

More in a bit, KF

[Kingfish]

Continuing...

This ebike went down for the count in November of 2011 when the RWD stopped working, and followed about two weeks later by the FWD. Second to that, I somehow managed to injure myself sitting cross-legged on the sofa playing a marathon round of XBox: Quadriceps Tendonitis was what the Doc said. I couldn’t stand to put any pressure on my knee, so working on the bike was out of the question for several weeks until things calmed down. Finally about mid-winter I was able to attack the hub problem.

Highlights of activity below.

RUST: DIY Hub Motor Repair & Restoration
Short story: Rust was removed, hub covers vented to eliminate pressure differentials that introduced moisture. The mods now allows for any moisture to weep out, and in combination with new varnish coat over the stators the chance for future rust is greatly reduced. Hub wires were upgraded on the RH to match the FH; both now 12-AWG:

From the 2011 Road Trip, the rear phase wires became hot and partly melted through the plastic sheath covering the individual wires, though not quite through the wire insulation. Still, there was some unforeseen wear and shorting that contributed to several pesky drive problems with the slaved Rear Controller & Hub. This was a must-fix.

These repairs took time, and seemed the correct action to take, however – either the rust or the repair process appears to have reduced the efficiency of the motors and they no longer perform at rated power (I need to test this to be sure but the freewheel test suggests so). The RH incurred phase wire damage from testing, and is in risk of failure; for now we are keeping an eye on the problem while we sort out a good fix… something like an armored elbow – like a two-piece cable gland (but they don’t make those).

Battery Harness:
The next big mod to the bike factored in my continued frustration with Anderson Power Poles, and I decided to replace the entire lot with regards to power distribution. The APPs are acceptable for the Phase connections; with contacts removed from the plastic case, they are the only connectors I know that can slither out from the nut, washer, and hub cover. All APPs between the Batteries and the physical controllers were replaced using EC5, HXT 4mm and XTS150 6mm Barrel Connectors. Previously the Main Harness was a kludge of 10- and 8AWG. No more: The Main Harness, the Utility Harness, and Trailer Harness are now all multi-strand Marine-Grade 6-AWG. All connections to batteries are via 8-AWG links, using EC5 connections. The Harnesses themselves are interconnected using the XTS150 6mm connections, and there is one HTX 4mm connector that uses 10-AWG wire to mate with the DC-DC Converter which supplies 60W of 12V & 3.3V power to the bike accessories and lighting. The 8-AWG links are mechanically swaged onto the 6-AWG harness using a Hydraulic Crimp Tool. The swages are actually 0.03 thick wall copper tubing; it’s stronger, faster to assemble, and more conductive than the tedious solder methods I have previously employed. The weight of the harness is about 3X of the original, and yet, I am certain that the resistance of the new harness is likely 10X less. My hope is that this will end the voltage sag issues that I’ve been experiencing when On The Road.


On the Left, the old nasty kludge of 10- and 8-AWG. On the Right, the rebuilt robust & streamlined Battery Bus.

There’s a second part to this mod in that each battery unit consists of a small assembly of three 5S batteries connected in series to make 15S. These “units” are then connected in parallel as assembled packs, and they connect to the Harness via the newly minted 8-AWG wires. The wires that connect the units together I call “collectors” for lack of a better term. The wire gauge from the batteries has been 12-AWG, although somewhere in the last year or two of production Zippy FlightMax changed to 10-AWG. For whatever reason, all my collector wires are 12-AWG and they are unified into a 10-AWG wire which connects to the harness. All of these collector-unions are this way; the physical union is wire-wrapped together and soldered, then covered with heatshrink in an incredibly long and tedious and laborious process. I need to change these all out to have the maximum benefit of the new harness as: 10-AWG silicone wire collectors into swaged 8-AWG unions: Altogether a work item of 28 collectors and 8 unions. Doesn’t need to happen right now, but certainly could be a rainy day project.

The Battery Harness is part of a larger system for the entire ebike and designed to serve me for cross-country. Below is the master EBike & Trailer Wiring Diagram for Main Power. (Apologies for the large picture.) I like drawing up my ideas and refer to them frequently during construction.


It can be viewed separately here...

Front Controller Current Mod:
For some reason I had not completed this mod before heading out last summer. The FWD never had issues; it was the Rear that would drop out during the most extreme loading. Very odd. Anyways, this mod was completed using 10-AWG stranded copper wire on the power and phase traces. The battery wires were upgraded to the same 10-AWG, although the phase wires remained 12-AWG: The FH is down-rated slightly from the RH by about 20% to prevent the front wheel from spinning during inclement weather.


Shiny Happy PCB

More in a bit, KF

[Alan B]

Amazing amount of wire in these ebikes!

How did you handle the balance leads? Any connections at that level?

[Kingfish]

Amazing amount of wire in these ebikes!

How did you handle the balance leads? Any connections at that level?

You don’t know the half of it

Except for the Triangle, I can get to any pack fairly quickly for balancing. However the Triangle is altogether another problem because it is fasten down by several tedious layers (and takes about ½ hour to load/unload). It requires a BMS in situ. I am not completely satisfied with my options and have been strongly thinking about what my next project shall be… I think we need to address this disparity good and proper: A self-balancing powerful management system.

During testing I was really taken aback by the large difference my Triangle Pack would see between cells – up to 0.7 volts! It leads me to suspect that small resistances, perhaps in the “collectors” wiring is instigating the disparity, but it really does hamstring Pack capacity and obviously damaging my assets.

A bit chilly this morning. I say - good weather for riding!

Need more coffee, KF
PS - More in a bit... <slurp!>

[Kingfish]

Controller Signal Harness Redesigned
This was a call to action and waiting for repair. When I first built the ebike I had an idea to co-locate the controllers at the rear in a Master-Slave arraignment. The problem was that I had wires passing through the linkage of the rear suspension – an incredibly dynamic area filled with motion and grime. Once it was deployed, I never cared much for it and it seemed like I had to frequently hack into it because of pesky issues, like no throttle on the RH controller, or missing Regen, or no Cruise.

Once I started down the path of resolving –EBS to GND problem good and proper, I saw the writing on the wall and spent a good deal of time copying it down and refining what I needed to accomplish. Then the search for parts began – another long saga (I’ll spare you). It was just a “prototype” mess that I created… and wanted to get away from and clean up the ridiculous bundle wires coming out of the controllers.


Corrected some issues and added features.

I didn’t like the Master-Slave arrangement cos I couldn’t debug the motors with the opposite controller without a special cable. I knew the key to reducing wires was to complete eliminate GND from the Controller to the Controls, and instead just pull a separate large GND wire up to the handlebar for the sink – which is going to work very well now that I’ve updated the Battery Harness. Most of the GND wires were already removed – I just took it one step farther.

Next, I had to settle on either a 8- or 9-pin connector. Molex is bulky, old-school. I sourced an 8-pin PCI Express extension cable (comes with M/F!) – the only problem was that it was only a foot long which meant splicing. <grumbles> For Signal Cable, I couldn’t find 8-conductor, so I used two 4-conductor 18-AWG wire. Now I have to tell you – that has to be the smallest 18-AWG I have ever seen and was a PITA to strip; I should have just gone to 16-AWG.


Document everything for reference; very important to keep from making errors during asembly.

I took both controllers and ripped out Throttle, 3-Way, Cruise, -EBS, and Vcc_L, then replaced with the PCI Express wires which were slightly larger than stock… a little difficult to thread into the through-holes of the PCB. The final touches included using the Nylon-Braid Expando Cable Sleeves which neatly hid the wiring. Now each controller is reduced to five wire assemblies: Battery (10-AWG EC5 and sleeved), Halls (USB Mini-B), Signals (8-Pin PCI Express and sleeved), CA (unprotected), and Programming (also unprotected). Both controllers are now – more or less essentially equal.

At the other end is the Dashboard: The convergence of instruments and controls and accessories. All flimsy wiring was removed and upgraded with better quality. The Dash actually suffered very little due to the protection provided by the front fairing, but it was still a rat’s nest. After rework… well, it’s still busy – but able to take a lot more abuse, particularly the Ebrake. I went through everything - even the Black Box, and lifted out part that didn’t work, shortened wires, beefed and padded protection – that sort of stuff. Also decided to add both types of Throttle connectors so I wouldn’t have to kludge something together in the field in case of failure.


My forte is Design and not Production, so I'd check off each small task as I went and kept the focus all the way to the end.

All of the related diagrams helped keep track of ideas and progress, and if something didn’t work – the docs were updated. I kept a daily journal/notebook that relied on manual technology: I could take it with me to the pub and enjoy a beer doodling out solutions. Except for the occasional leaky pen… nothing beats good ol’ manual technology for illustration.

I spent days assembling the harness, then hanging, then fastening it down. Before I began I had stocked up on Nylon zip ties; pity they can’t be recycled. Tedious work, but the outcome is greatly organized, and I’ve eliminated a source of several problems.


The most important diagram of the entire effort; all schematics lead to physicality. Apologies for the size; link to ful image here.

Weather Channel says it up to 52*F. I need an hour to finish prepping the bike. More in a bit.
~KF

[thickhead]

Thanks, Kingfish, for the detailed design/build info. The diagrams are great! This is a very timely update, since I am learning/thinking about what I want to do with my wiring between now and spring, and I really have little knowledge about electrical power system design and fabrication.

My situation is much simpler than yours (one motor, two battery packs, each with BMS), but for my commuter/hauling bike the goal is to have as close to perfect reliability as I can get. My "research project" is aimed at maximizing the amount of time I can comfortably depend on my bike as a car replacement (I'm in central PA, so I have the steep grades and weather to contend with). I already know what I have is not adequate, since even in dry weather I'm having intermittent contact issues with the wiring (particularly with the Hall and throttle connections to the controller, for some reason). So, much as in your case my engineering focus is on reliability/maintainability. I have been taking as inspiration your bikes (from P0 on) and telektik's 2WD YM.

Couple quick questions:

• You mention a hydraulic swager--what tool did you use? A quick googling comes up with some pretty pricey stuff. Maybe I'll have to get one for my lab.
• Are your harnesses weatherproof (well, weatherized) as is, or are they protected by your HDPE fairings? I'm awaiting HDPE sheeting that I'll use for weatherizing my Xtracycle cargo bags and for some fairing experimentation. I'm also thinking about using it for protecting the controller etc. I don't think I am ready to consider running the harness through the frame tubing... yet.

[Kingfish]

Hi TH

Swager: Good thread to read is Mechanical Crimping Large Gauge Wiring Harnesses. For my part, I came in late to the party. I puchased one off of eBay that can be had at Harbor Freight, but after I got the tool I realized the shortcoming of it right away. I think it's worth getting one that Harold in CR suggests which actually buckles the material correctly instead of flattening. Also, as long as yer paying for the tool - might as well have one that can go up to 000 cos if you ever want to build an eMotocycle or EV, plain ol' "0" ain't gonna be large enough

Weatherproof: Nothing is weatherproof. Not motors, not controllers, not connectors... although my beer belly is weatherproof. Best we can hope for is weather-resistant. The outer HDPE shell works very well to repel rain from the battery bag. The bag is marginally protected by weave and by Sno-Seal, but it only stops light moisture & fog; a driving rain will soak it, hence the HDPE + Marine Vinyl layer. Learned that the hard way; thankfully no shorts. The rubber interfaces where wires pass through at the controllers is problematic, and the larger wires (and USB cable) really tear it up. After assembly I buttered it all up with black Urethane sealant. I don't like Silicone cos it's hydroscopic. The Body man that I used for my truck gave me the lowdown on Urethane and showed me the very graphic differences, and I've been a die-hard convert ever since. Urethane cures to touchable within two hours, and in 24 it's solid enough to carve - if layered correctly.

For shielding cable runs, I like the Expando sheath for all the reasons I stated. The wires can still move within the conduit and I like that around the steering where bunches have to be secured in place - and yet be able to move to keep from breaking. Plus - I just like the look of black webbing

Best, KF

[Kingfish]

Quick note:
Battery pack mounted/strapped down; moving on to fairings.

Decided to take the opportunity to test each wheel one more time before covering up the wiring. I disabled/isolated each controller by disconnecting the PCI connector. Spun the wheel up to WOT, then selected Ebrake: Both controllers & motors are seeing Ebrake! So that’s one good bit of news right off the mark. The concept of providing strong GND return works

OK, back to assembly, KF

[thickhead]

Thanks, Kingfish, for the thread link and suggestions.

[Alan B]

Quick note:
Battery pack mounted/strapped down; moving on to fairings.

Decided to take the opportunity to test each wheel one more time before covering up the wiring. I disabled/isolated each controller by disconnecting the PCI connector. Spun the wheel up to WOT, then selected Ebrake: Both controllers & motors are seeing Ebrake! So that’s one good bit of news right off the mark. The concept of providing strong GND return works

OK, back to assembly, KF

Hopefully it will keep working even with big current flowing. Ebrakes are easy to opto-isolate anyway. Differential amps for the throttle might still be needed, but the brake signals are larger and less likely to be problematic.

[Kingfish]

Understood.

UPDATE

OK, at about 1:15 PM PST I mounted the P1-Mark II (or whatever) EBike. It's downhill from where I live; tested throttle lightly - good. Regen'd to the bottom with both controllers pulling 70-80W (I took it easy). Once at the bottom, I set my 3-WAY to Medium (86%) and raced to the top; front felt like it wanted to leap into the air!

Rounded the loop and regen’d hard to the bottom. All good. Waited for traffic to clear and bolted out down the road a mile to the next light where I could make my speed run. Checked MaxS and it was 35-36 mph (CAs differ slightly); this is very good and about equals what I had before.

Set 3-WAY to High (120%) and waited for traffic to clear. Then I hit it hard: I caught up with traffic and was about to pass when I reached the point to where I was satisfied my legs couldn’t take me faster and backed off; right turn Clyde onto my hill and bolted to the top WOT. Fricken awesome! Checked MaxS…

Lady & Gentlemen, I have a new personal level ground speed record: 47.1-47.4 mph (CAs differ slightly); that’s the fastest I’ve ever been on an eelectric, and less than 1 mph slower than my fastest ever on a bicycle (going downhill). I yam duly impressed! I’d wager it has more to do with battery power than anything else. I’m tempted to remove the APPs and go with bullets to remove the last skinny link in the chain.

So, Ebrake works on both, 3-Way works (all tested), Throttle obviously works. I haven’t tested Cruise yet. And on top of that – I’m winded and sweaty.

Well, with the first order of business out of the way I need to bolt to the Electronics Supply and fetch some resistors for the Progressive Brake/Regen Prototype. I have more pictures to post – including the one of the CAs MaxS, and shall get to that later this evening.

Gee, I wonder if I up the amps if I could make it into the 50-mph Club?

Like the wind, KF

[grindz145]

nice! dual 9Cs ftw.

[Alan B]

Series up another 5S pack for your speed run, amps alone won't likely make it to 50.

[Kingfish]

End of a great day!

Wow! The weather was crisp and mildly warm; I overdressed as usual and ended up peeling off a layer. At the top of the SR-520 hill a fellow cyclist caught up and we shot the breeze about electric bikes for a good long time. Made it to Vetco Electronics just as they were closing so I only had time to get the voltage divider resistors to knock 12V down to 5V for the Prototype, and 3’ of split-sleeve tubing to cover up the CA & Programming cables.

Then I stopped to eat at a place I’ve always seen from the road; last Friday over beers at Malt & Vine we got to talking about the best burger joints in town and BurgerMaster came up. So there’s where I went, except it’s old-style where they bring the food out to you and your car-load of pals. I’m on a bike, but there was a bench under the trees near the lot and I camped there and ate a ChickenMaster with the works and greasy onion rings. I can only eat onion rings like twice a year … and after the 3rd one… uhg. There was a raven watching me; I tossed it a bit of my bunn, and another showed up. OK, he gets a nibble too… and another shows up. Right, so I toss one to her… and another shows up. Then one walks over a bit closer and is pecking the wood like it’s pretending to eat – looking at me, so – I toss it a nibble… and another shows up. Then one sharpens his beak back and forth, and back and forth – and looking at me, so I toss him a nibble… and another shows up. It’s a party of ravens! I ran out of bunn but I had lots of onion rings. Now I’ve got ravens on the ground, ravens in the trees perched high and some perched low. One is perched directly above me, and I told him “Don’t you poop on me!” It was pretty cool cos none did. Not quite Hitchcock’s Birds, but still amusing. Boy did they fight over my space when I left!

So far the bike performed flawlessly and I was just thinking that on the circuitous way home when I noticed a high-pitched whine. Stopped at REI and picked up some more rope-clasps which I like to use for the fairings; one can never have enough clasps. At REI I figured out that it was the disc brakes that were singing, and not something worse – like motor bearings. My wheel-builder left REI last summer and he’s over at another local favorite bike shop, so I popped in there introduced myself (seems like high turnover in the bike world) and discovered my pal is on vacation for another week. Cool – he’s still there. I’m going to put the bike in for a tune-up: Inspect/replace brake pads, true discs, adjust the dishing on the RH so I have symmetrical clearance, and adjust my shifters. One hour with The Guy can do more than I can in eight.

Beyond the singing, I think it was as good of a day as one could possibly hope for a significant update. I am very pleased.

Here are some pictures of the latest ebike:


Last night just before it got dark I rolled it out to capture it before I put the batteries and fairing on. Kinda hard to see but the cabling is much better organized.


This is a close-up of the Dash: Unified CAs and Bar-Ends are the biggest changes.


Today I finished up the fairing work and rolled the ebike out for a vanity shot before the first ride around 1 PM. You know, I’m not one of those guys that has sexy-looking girlfriends hanging around to drape over my ride like a Vanna, but… I’ve got Bert Grant, and best of all – He’s Scottish and doesn’t complain one wit about wearing a skirt and showing some leg. So there it is, my finished Ride with the fairing, and Bert Grant holding one of my favorite beers of all time! What could be better? (other than wishing the brewery still existed).


About 10 minutes after the Bert Grant shot was taken, I set a new personal land speed record for level ground on the main road heading out of Redmond. The top CA is the FH, and the bottom – RH.

I have double-checked my records, and at no time have I ever gone this fast on an electric bike. Gawd I felt goosey afterwards! The bike can take more. Alan B, I think that I’ll try amps first cos I am so close… and I’m not pushing these motors very hard. If that doesn’t work, then we’ll do the 20S boogie.

I feel like a beer. Time for a walk.
Cheers to all good days, KF

[Kingfish]

Specs for the first ride:
Battery @ 62.1V; MaxS was about 47.25 mph. Total distance was 21 miles.

Today I put on another 16 miles. Tested Cruise and the new Bar-Ends, and both worked out exceptional well. I appreciate the bar-ends so much more than the aerobars; it just makes sense.

Overall the bike performed like a champ; definitely gustier and can keep up well with commuter traffic so long as we don’t go faster than 36 mph. I could use the high-current mod to keep up - but that chews up the pack too fast. The normal current settings are 18A Battery/45A Phase for FH, and 22A Battery/56A Phase on the RH, and then I run the “Medium” level of the 3-Way at 86%. The bike has plenty of power at that setting! Still, I can think of a couple more of improvements to squeak me into the 50 mph Club.

Back to work, KF

[Kingfish]

Pictorial on how I modify my Controllers

I’ve been asked how I modified my controllers to support 2WD. Oringally I had a Master-Slave arrangement which worked well-enough for the Road Trip in 2011. Now that I’ve had time to go through and reevaluate the systems I have decided to remove the uniqueness between the two, and apply another round of simplification. Below is my present system as of September 2012, and so far it has been flawless in operation; the ebike answers all bells without hesitation.

1. Start with a good controller.


This is the 12-FET 4110 MOSFET Extreme Modder Controller LYEN Edition (MII) that retails for $129. Let's see what's inside...


Front, stock.


Back, stock.

2. At this point I go through and begin removing components and wires in preparation for upgrade.

• Molex Connectors on Power, GND & Vcc_L are removed.
• All dashboard control wires and connectors are removed.
• Bullet Connectors of Phase wires are cut off.
• Hall Effect Signal wires and Power/GND are removed.
• The three largest Caps are removed.
• The solder and thin wires from the beefed traces on the backside are removed.

3. The unit on the bottom is partly modified: Removed +Vbatt & GND wires, Throttle, Cruise, 3-Way, Ebrake, & Vcc_L wires. It is going to replace the unit above that was also highly modified for the 2011 road trip (older 2WD version that I had "slaved").

• All through-holes are cleaned-up to allow new larger wires.
• Power & GND through-holes require extra reaming to accept 10-AWG.
• Traces on backside are beefed with 10-AWG stranded wire and soldered up good and fat.
• Added replacement caps.
• Soldered in USB Mini-B wires into the Halls/signal pads. Added zip-tie to backside of cable as a strain relief to prevent pull-through.
• Pulled all 8 wired of the PCI Express extension through, and soldered in 7 or them to the appropriate pads. The spare is tied off for future use.
• Pulled in 10-AWG for Power & GND and soldered into place.
• The backside of the board is then "leveled"; I trim leads or otherwise file down high points that threaten to short the board to the casing.
• The PCB is then cleaned of flux, and is inspected for continuity and shorts. Any issues are corrected.
• Tape off the backside for extra protection, and reassemble the unit for smoke-testing.
• Attach EC5 connectors to Power & GND.
• Attach 45A APP connectors to Phase wires.

4. Upgrade completed. Front.


Upgrade completed. Back.

5. The unit is ready for smoke-testing: We check power and controls before the motors are hooked up. Then follows Freewheel spin with each controller individually. Then ultimately - the test drive. This unit had a perfect first ride which is rare, but we'll take it as a compliment

The List of Mods:

• +VBatt & GND now 10-AWG, and uses EC5 connector.
• Halls signals replaced with USB Mini-B cable & connector.
• Throttle, Cruise, 3-Way, Ebrake and Vcc_L unified into a PCI Express 8-Pin Extension Cable. This greatly reduced the clutter coming out of the board. Note that GND is not included; I pulled Battery GND up to the handlebars for a direct return. The spare pin will be used for the UV signal when I hook up the Progressive Brake/Regen Prototype.
• The three large Chengx caps were replaced with high-quality low ESR caps (and having a known spec sheet).
• Traces beefed up on backside using 10-AWG wire.

Notes:

• The Phase wiring stayed as OEM 12-AWG because that’s the size I have going through the hub motor and I didn’t see much of an advantage to upgrade them. We are considering upgrading the APP connectors to either the 4mm or the 5mm bullet connectors for increased capacity.
• The two controllers are now identical; no Master-Slave. The only difference is that the FH controller does have much longer phase/Halls wiring, although it is possible to chop it short and just hardwire the FH so it extends back to the controller, and if so then either controller could act as Front or Rear – so long as I flashed them accordingly. Seems like a lot of work though. For now – KISS.
• FH Controller is presently set for 18A Battery/45A Phase, and the RH Controller is set for 22A Battery/56A Phase.

Enjoy, KF

[Kingfish]

Updates for the Month of October

Built new body fairing components
Here's the new and improved version of good ol' Prototype #1.



New Front Fairing – was the Bee Head style, is now the SvS version… that’s “Spy vs. Spy” illustrated by Sergio Aragones. At least – that’s what it reminds me of.


Read a lot of these when I was a kid...


Rather pointy there... I decided to redo the front after nearly making the 50-MPH Club. In my mind, I'm only 3 mph short - so if I improve the aero... you get the idea


The Dash is now very well protected.

Sliding Controller Armor – this is a small plate of HDPE that completely covers the wires exiting the controllers and it fastened to the frame. The “plates” slide independently inside the removable side plate, and adds both security and aero to the bike.

Center, Rear Cowling & Side plates – Normally I ride without panniers, so I decided to cover up the lossy rear rack for better aero. The side plates are made of white HDPE for better visibility at night.


All cleaned up nice and neat.

New Reflective Tape – on order; the original tape is worn out, torn off in places, and just plain dirty. The new white areas will be striped.


Altogether, the ebike is still has a very slim profile.

Progressive Brake/Regen Prototype:
This is an ongoing project with the singular goal of creating a variable ebraking experience in the same manner as mechanical braking.

Hi-Powered LED Flashlights rework:
The original idea was to upgrade them from 5-Way to Single-Click, but that was a bust after receiving parts from DinoDirect. During the Front Fairing reconstruction I took the lights apart, cleaned them up a bit, lengthened the wires, and increased the pressure on the electrical contacts – all of which has improved the illumination quality.

Turn Signal LED Indicator:
Really a brain-dead solution to alert me when I have left the indicator on after completing a turn. Took the switch housing apart and soldered in two 1N4007 Diodes to a 12V Yellow LED. Copy of the circuit is below. Works like a champ.


Wired up.


Block Diagram of modified control.

Replaced Phase Wire Connectors:
Replaced the APP connectors from the Controller to the Motor with HXT 4mm Barrel Connectors which appear to have about twice the current-carrying capacity. Combined with the new Battery Harness, I no longer see voltage sags, and in fact – the FH really wants to gallop forward; it’s quite the improvement!

That’s it for now. There’s more to do – but I also have to work for a living.
Safe Travels & Enjoy, KF

[The Stig]

I wonder how much the unsprung weight affects the handling... have you taken this thing off road much?
I'm considering building my first mid-drive for suspension performance and off road handling...