Bonanza "Bulldozer" Dual PhaseRunner AWD

[amberwolf]

It appears to be the same speeds for resonance regardless of voltage, IIRC, with 36V, 48V, 60V, 72V, and 84V, which are what I have experimented with using my NiMH so far, and also with a 48V Vpower/CammyCC pack. I'd have to actually test and note down exactly what speed it occurs at with different voltages, to see if it is exactly the same or if it does change at all, but I don't think it did.

I didnt' pay too much attention to it at the 72V and 84V as I was really looking for what speeds I'd get and what power levels used, as well as trying not to rip the front wheel off. :lol:

But anyway, whatever RPM it's at for 26" wheel at ~15 and ~20MPH, it "buzzes" there.

What I suspect based on other things I've read is that since these XC controllers are basically trapezoidal PWM at best, rather than sine, there is a point at which current pulses from the PWM line up with interactions from the magnets and poles due to the RPM vs the PWM frequency (not the commutation frequency).

But I dont' really know enough math to figure that out, and I'd probably mess it up anyway, even on the off chance I'm actually right. :lol: (which is slim)


With a sine controller, it probably wouldnt' be as noticeable (although it might still happen, because I don't know what is physically moving to create the sound).

 

[Alan B]

It appears to be the same speeds for resonance regardless of voltage, IIRC, with 36V, 48V, 60V, 72V, and 84V, which are what I have experimented with using my NiMH so far, and also with a 48V Vpower/CammyCC pack. I'd have to actually test and note down exactly what speed it occurs at with different voltages, to see if it is exactly the same or if it does change at all, but I don't think it did.

I didnt' pay too much attention to it at the 72V and 84V as I was really looking for what speeds I'd get and what power levels used, as well as trying not to rip the front wheel off. :lol:

But anyway, whatever RPM it's at for 26" wheel at ~15 and ~20MPH, it "buzzes" there.

What I suspect based on other things I've read is that since these XC controllers are basically trapezoidal PWM at best, rather than sine, there is a point at which current pulses from the PWM line up with interactions from the magnets and poles due to the RPM vs the PWM frequency (not the commutation frequency).

But I dont' really know enough math to figure that out, and I'd probably mess it up anyway, even on the off chance I'm actually right. :lol: (which is slim)


With a sine controller, it probably wouldnt' be as noticeable (although it might still happen, because I don't know what is physically moving to create the sound).

Interesting. If it does not change with voltage then it is probably a commutation frequency effect as the voltage will change the PWM duty cycle and harmonics. But commutation effects are related to speed and wheel size with a given motor.

 

[Alan B]

Test Ride

Beautiful weather here today so I decided to test another stretch of my commute route. Home to the base of Wildcat drive (I previously tested the climb up Wildcat). This is a very popular bike route, especially on a Sunday.

I dug up a #10 extension power cord with PowerPoles and tossed a pair of fully charged 6S 5AH Lipos into the trunk pack since I was not sure the pair on the bike would make the distance (prudent choice). The extension DC power cord is one I had, it is waay too long for this. I needed three feet and this must be ten or more feet long. But it is well made with #10 wire.

A mile or so from the house I pulled up behind a road biker, and followed him awhile. He was pretty strong, holding about 17 mph up a slight grade. He checked over his shoulder periodically. I didn't follow closely. As we reached the climb at the reservoir along the dam he geared down and dropped to about 12. It is perhaps three to five percent on this grade. There is a wide shoulder on this road, great for bikes. I waited for a few cars to have a clear lane and then pulled out and passed him, pedaling pretty hard and full throttle making 18 mph. (9C 6x10 hubmotor on 45 volts). I maintained both pedaling and full power up the hill by the dam and kept around 17 mph most of the time. Then the road levels out and alternates between slight climbing and slight descending. Very nice view of the lake. Cars travel about 50 mph on this road so you can hear them coming nicely. Never saw him, or any other bikes going my way for the 7.8 mile trip to the crossroads. Passed quite a few on the other side.

Home to Wildcat:
7.76 miles
2.75 amp hours
16.5 watt hours per mile
moderate pedaling

At the crossroads I stopped and rested. Talked to a young road biker (university english major). He said my rig looked nice, he had no idea that it was a motor (he thought it was some kind of internal shifting hub). It did not seem to bother him at all. I told him about my plans to commute with it. I was wearing jeans, light jacket and welding gloves. So not your usual biking attire. He thought that was great.

One older couple drifted by and the woman asked me what the "stuff" was pointing to the controller, batteries and wires. I told her that was for the motor, and her husband told her "don't even think about it", and something about cheating. I told them I was setting it up for commuting 13 miles with 1500 feet of vertical and they thought it was okay for that. (He said something about no fat chicks). Interesting how folks react. Most have NO IDEA about the existence, appearance or capability of these motors, and don't even recognize them at all.

I reconfigured the power to come from the pair of fresh 6S 5AH batteries in the trunk bag. On the way back I pedaled very little. The view is even better on this side of the road nearer the lake. No pedal bikers overtaken or passed me on the return trip.

Number of e-bikes seen on this trip: 0
Number of pedal bikes seen on this trip: 30 (most on the other side of the road)

On the return trip:
7.8 miles
2.96 amp hours
17.4 watt hours per mile
138 watt hours
17.7 mph average
26 minutes 19 seconds

On this return leg I took a slightly longer route through a nearby neighborhood.

So this 7.8 out of 13 miles represents over half my commute. Not bad at all!

Bike just after returning from test ride:

Better view of extension power wiring leading into trunk bag:

Motor wiring (note Hall red wire about to pull out!!):

Better views of Axle hardware:

This shows the Axle hardware. From the motor out:
1) thin D washer
2) dropout
3) D spacer / torque strut
4) Torque Arm
5) NordLock washer set (2)
6) Nut

Edit - note that the washer between the torque strut and the frame dropout appearing to bend the strut in the above photo has been removed...


New XTR V Brake (which works really well), and the Schwalbe Big Apple 2.1 rear tire and better rim and spokes on the new rear 9C 6x10 rear motor:

Trunk bag with too much wire:

Two 6S 5AH batteries in Trunk bag (still in factory red bubble wrap):

All this fresh air gave me a sneezing fit, so I will go take a benadryl and have a nap.

 

[Alan B]

Diagnosis by Photography

Looking at the photos in the last posting I saw several things I did not like. One was the washer behind the torque strut that appears to be bending the strut. The earlier spacer was a little thicker than the current one, so the washer needed to come out.

Another was the aforementioned loose red hall wire (the hall supply voltage). This actually explains something that I didn't mention on my test ride. On the return, about 400 yards from home I was stretching and taking it easy as the motor worked on this slight upgrade. As I moved there was a sudden moderate braking effect that subsided quickly. I was confused as to the source but decided my heel had touched the rear wheel brake disc. While this may in fact have happened (there was a mark on the shoe rubber), this was not likely the source of the effect. That hall wire was right there and it probably came disconnected for a moment as my shoe moved the connector about, then re-connected. It was basically just touching the pin before I repaired it. At least that's my current hypothesis.

I replaced the pin and did a better crimp job. The others appear to be fine. I also reworked the connectors and placed them all below the chainstay, well tie-wrapped and protected.

To get that superfluous washer out from under the torque strut I had to loosen the wheel nut. THOSE NORDLOCKS REALLY WORK! As you loosen the nut, tension rises, then it "pops" and drops. Amazing!! THOSE SHOULD BE STANDARD EQUIPMENT on e-bikes. If you look at the photos above you can see the ramps between the washers that increase tension as the nut is loosened. This insures that it won't loosen, as it is in a "valley" and it takes more force to move either way. So it just stays put.

So having good photos of the equipment and inspecting them really helped out here.

 

[Alan B]

The region near the axle is not a good place for connectors. I'm thinking that all connectors should be in the (soon to be shielded) triangle area. So the wires on the motor would be lengthened to perhaps 18 inches (half a meter) or so. Changing a tire would entail cutting a few tie wraps along the chainstay to release the cable but this seems like a good trade and avoids having water and stress on these connectors near the axle. The axle area would also be a lot cleaner with a single cable coming down the chainstay and into the axle.

 

[Alan B]

Connectors at the hub are gone. Took all the daylight after work to complete it so I didn't get photos. Seventeen pieces of heatshrink, two layers per wire plus one overall glue-lined armor shrink tube insure that this connection is likely to stay waterproof and mechanically secure. Now if I need to remove the wheel I'll unplug it at the controller. Tested and working.

Note that for future reference the phase conductors are marked 1.5mm. The cable from the hub motor and the cable from the controller appear to be the same exact cable with 3 phase and 5 hall conductors. They came from different sources but probably have a common root source.

The next task is to increase the voltage. This requires changing to the new controller, and rewiring a few things...

 

[Alan B]

Test Run

Work to inspiration point and back. This completes the route home to work to home, in pieces. I'll have to dig up the other parts data to get the whole route.

Work to Inspiration Point:

2.2 amp hours
3.8 miles
27 watt hours/mile
15.6 mph average
104 watt hours
moderate to light pedaling
12 mph up 10% grade pedaling moderately (worst grade of commute)
exit through upper gate (better route, shorter)

Inspiration Point to Work:

2.1 amp hours
4.4 miles
22 watt hours per mile
107 watt hours
light pedaling
return route is longer due to upper access gate not available for entry at lunch hour
return through lower gate

12S1P 5AH Turnigy pack, 50V charged, changed packs at Inspiration Point (midpoint) so each direction was fresh pack
9C 6x10 in 26" Schwalbe Big Apple
throttle generally full when not downhill, off downhill
top speed under power about 20 mph level

 

[Alan B]

Commute Trip Measurement Summary

The route is about 13 miles each way with about 1500 feet of vertical. Work is higher than home. The CA measured 14 miles so I need to do some calibration there, but the amp hours have been calibrated and agree pretty well with the charger.

Morning Commute

7.63 amp hours
26.4 watt hours per mile
343 watt hours total

Evening Commute

5.26 amp hours
18.2 watt hours per mile
237 watt hours

Total Commute (round trip)

12.9 amp hours
580 watt hours
22.3 watt hours per mile

This data based on three test runs added together, each of part of the route; one part with 9x7 motor and two with 6x10 hub motor in 26" wheel. All test with 12S1P 5AH 44V nominal Turnigy Lipo.

This is interesting data, it is less than I estimated. I plan to go to 18S 66 volts on the 6x10 motor, so power consumption will increase somewhat. Most of these were with light to moderate pedaling (I'm not a strong cyclist at this point so not a huge contribution, but it did seem to make a difference).

Based on this I can make the round trip commute with 6 batteries. 3 batteries is not quite enough for the morning commute, but 6 is enough for the whole round trip. I'm currently planning for 12 batteries so could just make 2 days commuting on a single charge, theoretically. I expect power consumption to rise with the increase in voltage.

But the fact that a commute with this much vertical can be done with so little battery is amazing!

 

[Alan B]

Lipo Balance Parallel Boards

The PC Boards I designed arrived today. They look excellent! Need to send in the order for the connectors and solder them in to make sure I made the dimensions right.

These boards allow four batteries to be paralleled at the JST-XH balance plugs by plugging them into this board, and a fifth connector or pads are available to connect to the charging connector for the pack. I'm thinking of using something like a DB-15 for the charger to carry both the charging current (on a few parallel pins) as well as the balancing connections.

The boards were ordered 3/14 and arrived on 4/6. So about 3 weeks. Not bad!! I may have some extra boards, too.

 

[Alan B]

Keyswitch Design

I've been considering different keyswitch designs. I want a keyswitch that is foolproof. No steps to remember. No relays. Very low power consumption. Handle 100 amps and 100 volts. Puts only a few milliamps through the switch. No sparks. No current surges that are hard on caps. Protection if they key is accidentally left on.

I made a couple of designs, but too many parts.

So here is a better design.

In this design the keyswitch powers a voltage regulator at a few milliamps. This powers an 8 pin micro. The micro controls one FET with a precharge resistor, turning it on first. After a programmable delay it triggers the two main current FETs and applies full power. When the key goes off, the micro and the FETs shut down.

But there is another feature - R5 and R6 divide the battery voltage down and give the micro a voltage sample. So if it has 100 volt range with 1024 ADC counts that will yield about 0.1 volt resolution. With this voltage information it can do two things. One is to look for variation in the voltage, indicating the motor is being used. If there is no variation for a period of time it can shut down, say after 20 minutes or whatever is programmed. So the system protects against leaving the key on and running the battery down. One other protection is low voltage. It can easily shut down if the system voltage is too low. It could also take an input from the LVC or BMS and shut down the FETs. So rather than have FETs in the BMS, just have one set in the master on/off control and use it for that functionality from wherever it is needed.

So how does that sound? Looks like about 15 parts (not including connectors and fuses).

 

[Alan B]

Peparing for stepping the voltage up

So now that we have the new 6x10 9C motor mounted we need to step the voltage up. The stock controller won't handle this voltage, so the Lyen 12 FET must be installed. Tonite I changed the phase connectors to match the existing wiring. I am planning to keep most of the Lyen controller connectors stock, but the phase wiring connectors I did decide to change.

I'm going to mount the controller using the DX Engineering saddle block clamps on the seat tube. I got some bolt hardware for them tonite. Next I need to make some brackets as the controller mounting ears don't match the hole spacing on the clamps.

I'll need to adjust the 12S1P battery series cable to 18S1P and make an extension so I can temporarily run batteries in the trunk bag or panniers.

Then I'll get to figure out if Lyen's color coding on halls and phase wires matches 9C/methods/ebikes.ca. Anyone know?

 

[dogman dan]

Swap the blue and the green, on both the halls and the phase wires.

 

[Alan B]

Thanks for the tip dogman, and thanks also to Lyen who sent me a PM about it. Apparently Ebikes.ca uses the Crystalyte convention on their 9C as well. Apparently Cycle9.com also uses this color code, though the connectors were slightly different at the motor the hall connector sexes were reversed.

Last night I decided to go ahead right away with a "simple test setup" for the battery mount for the voltage step up phase. This will use the DX Engineering saddle blocks and a piece of wood or plastic and mount six 6S 5AH batteries. That is all I have (of that size) right now, and it is a very useful amount of battery. That is 18S2P or about 63-76 volts at 10 amp hours (call it 700 watt hours). I want more later, but that much is very easy to mount. This simple mounting configuration won't hold more than six batteries but that is fine for now. That is adequate to barely make my commute round trip (and if I charge at work it is only about 60% discharge on the more difficult inbound part). There is also a popular ride not far from my house called the three bears that goes by a couple of lakes and has a lot of cyclists on the weekend. It is 18 miles with three difficult hills (hence the name). That should be well within this six battery setup's range. Lots of cyclists on that route to have fun with. Maybe carry some extra tools and help some out. Like to give some positive PR on ebikes when I can. Maybe make up some free ES business cards for those showing an interest.

Hopefully I will get this done this weekend, time permitting. Maybe in time for a good ride! Perhaps the folks over in SF will have another Marin ride.

I ordered the connectors for the paralleling board, and some DB-15's for charging connections. Should have them next week sometime. Current plan is to set up a DB-15 connector for each 6S parallel bank. This connector will include four parallel lines each for charging plus and minus; and single wires for the balance lines, so a single plug to charge each bank and three plugs in total. Three isolated chargers or one charger alternating on one bank at a time may be used with no reconfiguration of the pack for charging required. The only risk is multiple chargers that are not isolated, so I will have to avoid connecting more than one charger to the same power supply, which is not a difficult rule to follow.

I do need to find a solution to moving the shifters off the handlebar to set up the Magura throttle and opposing grip. I may just put both shifters on the left side to start with (which was a suggestion here as I recall), but eventually I think a parallel bar with the shifter would be nice. Are there such things made for bikes, or do I have to conjure up something myself?

Thanks for your suggestions,

 

[kold kanuck]

I like your idea for the key switch/anti spark pre charger and i even think i understand how it works. Which can only mean that you did something wrong :lol: :lol:

Glad to see your making progress.

 

[Alan B]

I like your idea for the key switch/anti spark pre charger and i even think i understand how it works. Which can only mean that you did something wrong :lol: :lol:

Glad to see your making progress.

Thanks. Seems slow here. But I think this weekend it will pick up. Motivational juices are starting to flow.

Tonite I went to the hardware store and picked up some crimp tubes and hardware. While I was there I prowled the wood aisle. Found some 1/4" hardwood in 2.5 and 3.5" widths that looks good for mounting the controller and batteries. I'm charging the two newest unused batteries now, having recharged the first four after the test this week. I removed the first two batteries that were on the bike and prepared them and the second pair for mounting on the board. I'm using 3M vet-wrap to constrain them together and to the red oak boards. This stretchy ace-bandage like wrap sticks only to itself and provides compression to hold everything in place without making an adhesive mess on the batteries and hardware. It is easy to remove later for upgrades and service.

 

[Alan B]

Basic Triangle Mounted Setup

18S2P LiPo 700 watt hours, 9C 6x10, Lyen 12 fet

Here is my Visio CAD layout for this design.

Under the top tube is the gutter to keep the triangle dry. It mounts to the to tube with saddle clamps showing as rectangles around the tube here. This will be fabricated later. The two small squares just below the gutter are the mounts for the triangle covers. These and the covers will also be future work.

On the seat tube is the controller. Still considering whether to drill the controller or fabricate some brackets to mount to the saddle blocks.

On the down tube is the battery pack. This is six 6S 5 amp hour Turnigy LiPos, two deep. Planning to vet-wrap them to the red oak plank under them, and bolt that to the saddle blocks. At the lower left block the small square is the third mount for the cover triangle.

There is a fair amount of room for wiring and airflow. Wires can route in and out at all three corners. Most will come from the handlebars at the upper right, and the motor wires will go out the lower left.

The batteries are 2 inches deep, so the pair plus the side covers will be about 4.5" plus a bit for the vet-wrap. So it is narrow enough for good pedaling room.

Weight Estimate

12 pounds for motor
12 pounds battery
2 pounds controller
2 pounds mounting
2 pounds wiring, CA, throttle and misc

30 pounds above plus the bike

Perhaps 70 pounds grand total with bike and accessories

Sure adds up. Hopefully this estimate is not low. (measured later at 63 pounds)

 

[Alan B]

Progress Report

Battery mount made and installed. That Red Oak 1/4" by 3.5" board is very stiff. DX Engineering saddle block clamps appear to be working well.

Controller mount made and installed.

Magura throttle installed and wired, handlebars reconfigured.

Hall plug installed.

Phase wires hooked up.

Next is the power wiring... Back to the shop.

 

[Alan B]

Working on plans for the battery wiring. Picked up a few parts at Fry's tonite for this, working on a clean installation that is at least somewhat reconfigurable.

I did clean up the wiring with a few tiewraps, as far as it is. Picked up a large assortment of black tiewraps.

The other thing I need to do is wire up a couple of variable resistors to adjust the Magura offset and full speed settings.

 

[gtadmin]

Progress Report
Battery mount made and installed. That Red Oak 1/4" by 3.5" board is very stiff. DX Engineering saddle block clamps appear to be working well.
Controller mount made and installed.
...

Hi Alan, I'm following this build with interest. Will you post a piccie of the battery and controller mount please?

Cheers, GT

 

[Alan B]

Great to hear there are a few folks out there following all this. Nice fiberglass job on the Aprilia, by the way. I may have to try that some time.

But for now I've got to keep it simple. I did take a few photos yesterday but was too lazy and tired last night to put them up. So here goes. It is 5am and dark here so good to lay in bed with the laptop and work on this now.

The sun in the side yard was pretty bright so I turned on fill flash but didn't really mess with the photos much. They should at least be adequate to see some of the detail.

First an overall shot:

Zooming in a bit to the triangle. Six batteries (6S 5AH) vet-wrapped to red oak 1/4 by 3.5" board which is then bolted via DX Engineering saddle block clamps to the downtube. The downtube is flattened on both ends, the flexible saddle clamps are complying with this distortion at the lower end nicely. The saddles were notched with a side cutting drill bit to allow the shift cable to come through.

I was concerned that the Magura throttle might be "un"stealthy, but it is actually pretty stealthy. Feels pretty nice. Looking forward to trying it out.

Shifters and rearview mirror. The mirror was removed and has not been readjusted, it must be a bit high as that is the roof of the second story neighbor's hours just a few feet away! I'm not happy with the shifters, but had to do something to clear the bar for the Magura. Looking for ideas on that, some kind of parallel bars perhaps. Probably will want both shifters on the left side, and get the matching Magura handle on the regular bar.

Moving in on the battery mount from a lower angle. Here you can see the bottle mounts under the board. The vet wrap is really locking down the batteries. They are not budging. Time will tell if this is a good solution, but it looks great now. Just need to complete the wiring...

The cable to the rear hubmotor is the standard cable that comes out of the hubmotor. I removed the two connectors and spliced it with solder and plenty of heatshrink, include overall glue-lined armor shrink. This whole run should be redone with 12 gauge from inside the motor at some point.

Moving in even closer to show the controller mounting. Here I made an aluminum plate with holes to match the saddle block. The controller flange is pinched between this and the saddle clamp. I may change this but it seems to work so well that I will give it a try. The compliance of this saddle clamp material makes it "grip" the controller. It seems rock solid now.

Lower end of the controller and a few wires. Have to work on that a bit more.

Today's challenge is to make a simple clean battery series-parallel setup and get this thing on the road for some testing!

 

[gtadmin]

Great to hear there are a few folks out there following all this. Nice fiberglass job on the Aprilia, by the way...

Hi Alan, thanks for the compliment I'm going to use what I learnt doing that in building my battery box for the cruiser + I like your mounting blocks idea. I also like your keyswitch circuit but I only have a basic understanding of electronics - I'll leave the design to the experts!

I felt that nobody was interested in my cruiser, but it gets quite a number of views even if nobody posts so I keep updating it when and where. Plus it's documentation.

Finally, thanks very much for the piccies 8)

Cheers mate,
GT

 

[kold kanuck]

Glad to see the saddle blocks worked out, they are such a much more elegant solution than using u-bolts. Look so much more positive in contact area plus i imagine they will provide some vibration and impact damping. Curious as to your impression of the new throttle and its operation compared to the hall one you stated with. Hope you can come up with a satisfactory shifter method and once again thanks for the documentation and photos.

 

[Alan B]

Glad to see the saddle blocks worked out, they are such a much more elegant solution than using u-bolts. Look so much more positive in contact area plus i imagine they will provide some vibration and impact damping. Curious as to your impression of the new throttle and its operation compared to the hall one you stated with. Hope you can come up with a satisfactory shifter method and once again thanks for the documentation and photos.

The saddle blocks are a vast improvement over u-bolts. A u-bolt must dent the tubes to get any real grip, whereas these spread the force without denting. Since they are compliant they really grip. I don't think they will give a lot of vibration and impact damping, maybe a tiny bit, but they are really firmly on there.

Progress report and photos coming up shortly. The Magura throttle is nice, I still need to adjust the range. I put it in directly and it has a lot of deadband at startup. There is no deadband at full throttle, though, or very little. The feel is great and it is more ergonomic since I can position my hand at any angle.

The bike is running again, obviously... More soon...

 

[Alan B]

Progress Report

I spent a few more hours today on the bike. Essentially hooked up enough wiring to make it run. I based the battery series/parallel wiring on a four position 75 amp screw terminal block. It is sitting on the top of the batteries.

Moving in a bit closer you can see the terminal block and the wiring. It will be vet-wrapped in approximately that position, so it won't be visible normally. The fuse is in the upper foreground. The main power SB-50 is on the right. Behind it, not visible, is the precharge PP30.

Here it is all plugged in. The one plug not in is the controller "on" line.

Here is the overall view, ready to roll. Obviously I need to make side panels, but that is for another day.

Ride Testing

I took it out for a quick spin as it was getting dark and dinner was near ready. It quickly climbed up a steep street near my house. I should have documented that hill climb before. It is pretty steep. The 9C 6x10 at 75 volts with Lyen 12 fet controller in default configuration went right up this hill. About 15 mph no pedaling. Drawing about 3kw.. It was not that easy on the 9x7 at 48V. Or the 6x10 on 48V!! It definitely has more muscle!!

On a short straightaway top speed run it got to 26 mph. That's about right.

Acceleration is way up. Again about 3kw.

Power consumption was about 60 watt hours per mile for a bit over a mile. This was pretty worst case, jackrabbit acceleration, speed runs and steep full throttle climbs. Motor and controller were cold at the end.

I like it!

 

[Alan B]

That reminds me

I need to check the current accuracy with the new controller. Both are 12 FET so perhaps it is the same. It seems about right. I also need to recalibrate the tire diameter slightly. My earlier testing shows it is a little off in distance/mileage.

Motor Wiring Hookup

I had to search the motor phase space on this build. Interesting process. The Lyen motor tester showed me that everything was working but didn't directly show which phase combination was the one. It did help though.

On my new setup it is not easy to change the phase of the hall sensors, so I wanted to minimize that. There are six combinations of hall, but due to rotation symmetry only two combinations are truly different. One is forward hall rotation, one is reverse. You can see this with Lyen's tester. Fechter also diagrammed a simple hall tester with three LEDs and three resistors here on ES.

For each hall combination there are six combinations of motor phase. These are currently powerpole 30's on my setup so they are easy to change.

So the process becomes:

Test each of the six combinations of motor phase wiring (for one hall wiring combination). If none of those work smoothly in forward direction at low current, swap any two of the hall sensor wires (leaving ground and power alone) and try the six more combinations of motor phase. Twelve total. Not thirty-six. The other 24 are simply rotations of the first 12.

On the motor phase, make sure to explore all six combinations. Assuming three colors:

1) colors matching through (one combination)
2) one color matched, other two not matched, for each of the three colors (three combinations)
3) no colors matched (two combinations)

After All That, it looks like Dogman's suggestion was right on. For some reason it didn't seem to work the first time I tried it.

Final Motor Wiring for Lyen 12 FET to Methods 9C 6x10

Controller---Motor

Halls
y---y
g---b
b---g

Phases
y---y
g---b
b---g

Resonances in the motor

Remembered last night that I didn't hear any resonances in the motor. Have to look for those. On that test hill I accelerated from 0 to 15 mph so I would have thought they would have been noticeable. Unfortunately I have to work today, maybe more testing tonite if the weather holds.