Bonanza "Bulldozer" Dual PhaseRunner AWD

[Alan B]

Street Hill Climb Data

Yesterday I used a local street (Brentwood) to test the climbing of my new 75V 9C 6x10 setup. I had one request to measure the grade for some reference. I had already planned to do so, and here it is.

I took a pair of compasses that have vertical angle measurement capability. I made several measurements both of the sidewalk by contact, and by sighting to eye-height targets up and down the grade. I got readings from 7 to 9 degrees. To convert from degrees to grade take the tangent and multiply by 100:

7: 12%
8: 14%
9: 16%

So until I can take a more accurate measurement I'm calling this a 14% grade. On this grade my 75V 9C 6x10 accelerated to 15 mph. I calculated the power required to climb this at 14% to be approximately 1500W, and the indicated power was about 3000W. So it appears the motor is operating at about 50% efficiency, very roughly, and generating about 45 pounds of thrust.

That is pretty good, I think, for a hubmotor.

 

[Alan B]

In other news

The DigiKey order came in, and the JST-XH sockets fit my Balance Paralleling PCB perfectly. So that is a go. I will use 3 boards for this 6S pack. I also got DB-15 connectors to use for the 6S one-plug charging setup. This allows 7 pins for balance and four pins each for charge plus and minus.

I was hoping to get some 6S SMD LVC boards, but those have not become available yet.

If anyone else is interested in these simple up to 5 socket JST-XH parallel boards, send PM. I have a few extra and could get more if needed. They will work for any JST-XH pin count up to 7, which means up to four each 6S packs can have their balance leads paralleled by simply plugging in. The traces are sized fairly small so if there ever is a KFF event the board traces should open and save the balance leads from damage. Plastic boxes, heatshrink or conformal coating should be used to insulate the board. The fifth socket is intended for the balance charging connection, but all sockets are the same so it could be used in some other way.

 

[Alan B]

Lyen 12 FET Controller Information

The default configuration is 30A battery / 80A phase, from Lyen. This is consistent with the 3kw peak that I've been seeing.

The model I have uses 4110 FETS for up to 100V. They are available in the for sale thread here on ES. Very quick service and he does repairs, though I don't plan to blow mine up anytime soon! At this power level it has a lot of safety room.

 

[Alan B]

Test Run

Did another test run today from work to inspiration point. Last time I did this with the 9C 6x10 motor at 50V. This time same motor but 75V. Leaving work is a very steep hill. I took some more measurements and it varies from about 10% to nearly 15%. The new configuration climbs this easily, making over 15mph on it w/o pedaling. This trip took 185 watt hours this time to travel the 3.9 miles. Last time with 48 volts this trip took about 100 watt hours, but I went slower and pedaled more. The extra voltage gives a lot more performance but takes a bit more power. In terms of the commute from wildcat to my house is mostly downhill, so I easily have enough energy to make it.

Based on the higher energy consumption I don't think I have enough battery for a round trip unless I conserve and slow down. But I have plenty to blast one way. So charging at work is going to be key.

I was looking at the possibility of adding 3 more batteries to get 18S3P for 15 amp hours. It is possible but would be a squeeze. I am thinking differently about the cover for the batteries now, and the new ideas would take less space in the triangle and barely leave enough room for a few more batteries. The wiring would be tight, though.

I'm pretty impressed with the power. I tried a standing start and the front wheel started to come up. This doesn't seem to happen if you are rolling first, but you have to watch it from a stop.

The throttle has a lot of slack, as expected. This bunches the actual useable range near the end. This still needs work. When it is spread out the control will be even better, and it is already quite good.

I discussed charging at work with various folks there and there seems to be no problem doing that. I can charge in my office or near it in some space I have. I'm considering a bulk charger at work. Perhaps balance charge at home and bulk at work. The BMSBattery chargers look interesting. They have a 4 amp 75V charger as well as other capacities.

I'm still looking for a nice small kill switch that would mount to the handlebar. I may have to make something, such as a handlebar mount for an ON-ON-ON toggle switch set up to be OFF-ON-OFF like a standard motorcycle kill switch.

I am also looking for bar ends that will take the grip shifters. I want to get those off the left grip.

I let a friend at work try the ebike today. He was very impressed. He's thinking about an Ezee geared hubmotor. After feeling the power of the 9C I'm not sure what his plans will be. He doesn't need the hill climbing that I want since it is pretty flat where he lives.

I don't know how well it will climb in the dirt, but on pavement it does great!!!

Thanks for all the input on this project!!

 

[Kin]

Going through your thread, I have had difficulty finding where you started making the decisions on LiPo equipment. Do you think you can point me to the right page range? I was lookng through the first 3-5 and last 3-5 pages.

 

[Alan B]

Going through your thread, I have had difficulty finding where you started making the decisions on LiPo equipment. Do you think you can point me to the right page range? I was lookng through the first 3-5 and last 3-5 pages.

Originally I bought a set of 16 amp hour Headways for 16S at 50V that I was planning to split into two low mounted rear pannier packs. Then on page 8 I bought a pair of 6S 5 amp hour Turnigy Lipo to use for testing. On page 11 I started considering using Lipo for the main pack since it would fit into the triangle. I also got the 6x10 motor which required a voltage increase at about the same time and that made the Headways even less practical.

For me the issue has been getting the batteries into a convenient location on the bike. The Lipo offers more power density in both size and weight. I just could not get the Headways to fit in the triangle and I really decided that was a better location than the original low rear pannier plan. One part of that problem is that I ordered the 16 amp hour Headways. These are too long to mount transversely in the triangle, they would take over seven inches of width. Some folks buy the 10 amp hour units and that helps some. My constraint width wise to fit between the cranks is about 5.5 inches. Even the 10 amp hour Headways are a bit long for that. My Lipo pack is four inches in width. I can touch my knees together over the top bar and still not touch anything on my setup. The batteries are the widest part and four inches is absolutely not a problem.

I like the easy maintenance of replacing a cell that the Headways offer with the screw terminals (though the way the spacers get locked in with the connection bars it is actually a lot of wrenching to work on them). The availability of the cells is a bit poor, and the hazmat shipping fee is unfortunate. I also am not as satisfied with the voltage sag they have and the fact that they are really overspecified. If you want long life the lifepo4 chemistry is going to last a lot longer (if properly tended) than the cobalt oxide of the lipo.

But if you want a small light pack that comes with connectors and balance leads installed and has little sag then the Lipo is great. Since my bicycle pack is so sensitive to size and weight, and since better technology will be out in a few years I plan to upgrade batteries anyway. So having a pack that will last 10 years is not really required for my bike. Well tended Lipo will last three to five years or more, and that is adequate for a small bike pack.

The lithium manganese (Makita Konions, etc) are another good choice if you are willing to make a cell welder. It is a bit of work to test each cell thoroughly and you don't want to have to work on a welded pack, but they are excellent if you know exactly what you want and are not going to be reconfiguring it. Or one could come up with sub-packs and weld them and combine them (like the RC Lipo packs) if you need to maintain flexibility. The lithium manganese are in between lifepo4 and cobalt oxide lipo for energy density. So if you want the most then Lipo is currently ahead.

A123M1 lifepo4 cells are another great choice if you don't mind welding. With the nano technology their current capacity and voltage sag characteristics are really amazing. But the energy density is low similar to the Headways and they have the pack welding and reconfiguration issues like the Konions.

I have and use all these battery technologies in various applications. For fitting into the triangle the best options are the Lipo or the packs welded from small cells such as the A123M1 or Konion type lithium manganese. The cobalt oxide Lipo has the greatest energy per unit weight and volume. The small cylindrical cells actually pack a bit better into the triangle so at the end of the day it depends on what you want and how well you know it.

Good luck with your selection,

Thanks for asking,

 

[Kin]

Wow, Alan,

thank you for theresponse. It was great, and really thorough. In some ways, I was even looking for a simpler question though. Well, for the bike I am (trying to go about) fixing for some one else, who has enough money, I am likely to recommend an a123 packs from Cell_Man because A) as evidenced be the ebike's condition they don't know much how to take care of technology [actually it's just that ownership changes hands on a 2-4 yr basis] B) It's a low-range-needed-use, high-current crystalyte motor (so high C yay).

I was going to go through wherever you made the LiPo decision, in part really only to look at a potential LiPo shopping list. For my own ebike, a 20" folder, I recently began debating going LiPo because A) I (as opposed to the other bike's owners) can take enough care of it, B) I like the lower weight and volume (I want to keep a somewhat "neat" look, and keep weight down for stairs), C) I like the modular design, so I can start off with a 15s (3*5s) 5000mah pack, but buy more bricks in the future as range needs and purchasing ability increase. 15s because 14s seems awkward (5+5+4), meanwhile 12s is just a bit slow on my 20" bike per calculations on the ebike.ca simulator (48v is good for me on the simulator, 15s is a bit more).

The problem with LiPo has been finding the appropriate charging equipment. I am definetly on a budget, but would like to be able to charge my batteries in one sitting even if that entails connecting and reconnecting packs [I.e I'd rather not charge brick by brick every 40 minutes]. One downside is that I might not save much weight, given I'd have to carry a charger and power supply for the charger if I want to charge during the day. This is a bummer. I also don't fully know if I need to buy a charger and balancing unit, or just one, or something else. SO I was going to try and found how you found those answers to the LiPo charging questions, especially since your purchases seem to be very recent compared to some other people who switched over.

Now that I have read through most of your build thread (it is great!) I have narrowed down by elimination you must have made the choice between pages 3 and 10 .

Thanks for your help and documenting so well on this thread!

 

[Ykick]

...12s is just a bit slow on my 20" bike per calculations on the ebike.ca simulator (48v is good for me on the simulator, 15s is a bit more).

If you buy a 20" rim wheel hub motor it should be laced with a faster wind. 2qty 20" wheels I've purchased used 10/6 (2806) instead of the more common 9/7 (2807) that comes with 26" rims. Bottom line, there's not that much top speed difference between 26" and 20" sizes as long as you get the faster motor in the 20" rim.

Also, a Lyen edtion controller will provide for extra speed per volt there too. I still can't believe how 12S runs almost as fast as 15S did on the stock eBikeKit controller.

My 15S experience is very mixed. I like the performance with a stock controller but everything's much more complicated and expensive than 12S. Think about it, you gotta buy extra bricks and a stock 48V Meanwell can't quite reach 15S full charge voltage. 12S @ 50V, no problem for practially any 48V supply adjustment. Balance charging is easier, etc.

Sorry to butt in, 'hope this helps...

 

[Kin]

Sorry to butt in, 'hope this helps...

Actually that helps quite a bit. I forgot about the Lyen speed increase, and I did buy a lyen 9fet controller. I was using the ebike.ca simulator of a 9C 2805 20" at 44.4v (12s) trying to reach ~30mph max, but being just a bit short. With the 20% increase it should be easily within reach. So meanwell is a single 48v source, can I use that to charge the batteries in series on a normal basis, and only balance wen needed? It seems you were suggesting meanwell as a charger not charger power supply.

 

[Ykick]

So meanwell is a single 48v source, can I use that to charge the batteries in series on a normal basis, and only balance wen needed? It seems you were suggesting meanwell as a charger not charger power supply.

Yes, I use the 48V MW as charger directly to the 12S pack. I added a Fetcher mini limiter to stock SP320-48 so it doesn't go into hiccup protection mode and adjust to a tad under 50V. Averages almost 7 amps output so the pack is full very quickly. I never leave it unattended but I'm not hovering over it either.

A 12V MW is probably handy for an RC charger too but I've never gone down that road yet.

I love my recent 9 FET Lyen edition. Installed this past weekend and wow, what a difference! Seriously thought I mixed up packs and used a 15S instead of 12S.

 

[Alan B]

More on Lipo

OK. The inexpensive way to get started is the RC chargers. I bought the Turnigy 8150's. I'm using one of those to charge my Lipos now. I charge one at a time at the moment. I bought 6S extension cables from ebay since the stock ones are too short to easily reach the charger from the bike, and I made an adapter to go from the charger's output connector to the 4mm Turnigy battery connectors. This is not a good long term solution since too much reconfiguration is required to charge each battery individually. It was fine for 2 batteries on the test setup to get started.

Next I need to make it practical to charge 18S2P on the bike. For the first phase of this I am going to use the paralleling board I made and a cable to a DB-15 for each parallel bank, so a single plug carries balance and charge currents (safe and easy). This is an easy setup to implement and fairly convenient to use. In my case I would plug in the charger three times (there will be three DB-15's on the bike), or have three chargers for my three banks of 6S. This is not quite optimal for convenience but it is good for balancing and it is not too difficult. Three chargers would make it even easier, but note that each charger must have its own isolated 12V power supply as they are sitting at 0, 25, and 50 volts above pack ground since they are still in series while charging.

The next upgrade here would be to get the SMD LVC boards from geoff57 (I don't have enough room for Gary's new boards). These boards would replace my paralleling boards and provide the paralleling function, but in addition they would provide Low Voltage Cutoff to the controller to prevent overdischarging the pack. In the meantime I'm watching the Cycle Analyst integrated current and making sure I don't over discharge the pack. The LVC would provide an additional insurance. This is not absolutely necessary but nice to have.

The next level would be to get a bulk charger for the total pack voltage. I'm looking at the BMSBattery alloy chargers. This would be a single plug bulk charging scheme. If I did this at work, and then balance charged at home (say with three Turnigy's) then the pack would stay well balanced, according to experience of others. Many folks balance charge rarely once their pack stabilizes.

It might make sense for me to get two of the bulk chargers. That way I could stage one at work and use the other either at home, or take it when I might need to opportunity charge along the road. It is also good to have a spare in case one fails if you are using it a lot (such as commuting).

The next level would be to have onboard HVC (High Voltage Cutoff) as well as LVC detection. Then a bulk charger can be used but there is more protection against a single cell overcharging.

The top level is onboard bulk charging with LVC and HVC - a full BMS. Then you have to commit to the weight/bulk of the onboard charger. Easy on a car or motorcycle but not so easy or perhaps desirable on a bike. Might be better to bring along a small bulk charger when needed. It would be great if a really small and light onboard system could be there all the time though.

 

[Alan B]

Bar-end Shifters

Spent a few days reviewing NIF (The National Ignition Facility, http://lasers.llnl.gov/). Very impressive facility and machine. Hopefully in a few years we will be charging our ebikes from clean fusion energy. On the way home I picked up a set of bar-ends as recommended at the start of this thread to mount the twist shifters to.

I installed the bar ends right away to see if they were going to work. No test ride yet, but they look like they are going to be an excellent way to clean up the handlebar clutter. I'll do some testing and get some photos this weekend for the thread here.

I suppose the next thing on the agenda after the bar ends are all settled and the left Magura grip is installed will be the charging setup. I need to put a few of my balance paralleling boards together and wire up the DB-15's for charging.

 

[Alan B]

Shifters on Bar-ends with Magura Throttle, Grips and Mirror

 

[Alan B]

Building Balancing Boards

These are easy boards to build, but I still use my surface mount workstation. It makes it easy to see the joints without bending over the bench and pulls all the smoke into a filter. You sit upright and look through the low power binocular microscope. It has a working distance of about six inches so there is plenty of room above the board for tools.

When doing surface mount the box under the board is a preheater (heater elements are covered now) and the hot air is used to do the soldering after paste and parts have been placed.

This gear is inexpensive clones of the commercial gear, or one can find used gear to keep costs down.

I assembled these boards for 3P, leaving out the 4P connector. Only 2P is on the bike now, but there is room for 3P so perhaps later I'll upgrade. The connectors are all in parallel so it doesn't matter, but I plan to bring the charge balancing cable in on the leftmost connector and use the other three for batteries. The traces are all on the bottom so we only have to cover that side to avoid shorts. The connectors "snap" into the board so it is easy to solder them as they stay in even when inverted. I plan to make this board available via the BatchPCB marketplace so folks can get them.

 

[gtadmin]

Hi Alan,

"Creative" solution for the shifters (looks like a pair of ?? though), how do they feel in use? Impressed by your workstation , I have one like that (actually no, I use the kitchen table and a magnifying glass )

Looking at the side view, if you tapped some holes in the tube mounts you could attach some side panels. I looked at buying some (tube mounts), they were about $3 ea but postage was $37

Cheers,
GT

 

[Alan B]

I have not done a test ride yet, perhaps tomorrow. The batteries are a bit low and hard to charge right now. But I should at least do a quick run and test out shifting. Testing on the stand was fine, and I adjusted the derailleur.

Actually there is one issue I don't understand. The 6 speed twist shifter only covers 5 gears. Pretty weird. I'm thinking about changing back to the original 9 speed to see if it might be a better deal. It would match the front shifter and cover the range (and then some). Just have to see how many gears are "usable", and how many are in-between.

On the side panels, the saddle blocks are not as wide as the batteries, so that would have to be handled. There is not much to tap into since there are long bolts inside all along where you would tap. I'm thinking a nylon cover would be much easier to do. Don't really need a cover when the weather is good if you want to show off the batteries, but hiding them would be good some of the time. Could have a nylon cover in the trunk that could be applied when needed. Stretch loops sewn into the corners might hold it down to some hooks.

Bummer about the price. These were about 5-7 bucks and the shipping was about the same for the whole package. Arrived in 3 days or so. You can make saddle blocks and that would be ideal as they could be wide enough and be tapped on the side, but it would be hard to match my oval tube at the bottom. Ones I made of delrin would not be so compliant. In any case these work so well they are worth the cost for me.

On the bar-ends perhaps longer ones that are curved more would work out even better. I just grabbed something that I knew would work, they didn't have much selection at the LBS. These are short with just a little angle at the mount.

 

[gtadmin]

Hi Alan,

... The 6 speed twist shifter only covers 5 gears. Pretty weird. I'm thinking about changing back to the original 9 speed to see if it might be a better deal. It would match the front shifter and cover the range (and then some). Just have to see how many gears are "usable", and how many are in-between...

The shifters have to be matched with the cassette and derailleur as there is different spacing between the cogs on the cassette depending on how many cogs are mounted. The shifter I got from the LBS is a SRAM, but it is supposed to be compatible with the SIS derailleur I'm using - we'll see

On the side panels, the saddle blocks are not as wide as the batteries, so that would have to be handled. There is not much to tap into since there are long bolts inside all along where you would tap. I'm thinking a nylon cover would be much easier to do. Don't really need a cover when the weather is good if you want to show off the batteries, but hiding them would be good some of the time. Could have a nylon cover in the trunk that could be applied when needed. Stretch lops sewn into the corners might hold it down to some hooks...

Figured there wasn't a lot to tap into and that you would need to space the covers out, but I was thinking that they would provide some protection for the batteries if ever the bike went down. To that end, the attaching bolts could be as small as 4 or 5mm in dia and the covers could be made with a bulge. You would need another tube mount on the top tube as well.

 

[Alan B]

It would be easy to add a mount to the top tube.

Hitting the batteries in a crash is going to be difficult. The frame tube is 2 inches wide, the batteries only 4 inches, so only one inch sticks beyond the tube. With the cranks, handlebars, seat and tires it cannot hit on a flat surface. A curb or protrusion might reach them. But they are surprisingly small mounted inside the triangle.

Supposedly the six speed stuff is one size, the funny spacing started later, but clearly there is something going on. Perhaps this derailleur has a different rate, but it came with sram shifters and this is an sram shifter. Anyway it is off so I'll probably try the original shifter and see how it does. I did that originally and most of the gears worked fine, so perhaps I should have stayed with it, but I thought I could get some improvement.

 

[gtadmin]

...Hitting the batteries in a crash is going to be difficult. The frame tube is 2 inches wide, the batteries only 4 inches, so only one inch sticks beyond the tube. With the cranks, handlebars, seat and tires it cannot hit on a flat surface. A curb or protrusion might make reach them. But they are surprisingly small mounted inside the triangle...

OK, as long as you're comfortable with that risk assessment. I'm just a bit paranoid about puncturing a LiPo I guess

 

[Alan B]

...Hitting the batteries in a crash is going to be difficult. The frame tube is 2 inches wide, the batteries only 4 inches, so only one inch sticks beyond the tube. With the cranks, handlebars, seat and tires it cannot hit on a flat surface. A curb or protrusion might reach them. But they are surprisingly small mounted inside the triangle...

OK, as long as you're comfortable with that risk assessment. I'm just a bit paranoid about puncturing a LiPo I guess

It is a valid concern. One approach would be to put some tough plastic plates right on the batteries, armoring them. This would be 2x6x1/8" plastic panels right on the batteries. Would be easy to do.

One concern is getting the batteries off the bike in case of a conflagration. On friday Lyen had a fire develop in his Lipo pack while riding. It was quickly removed so the bike was not damaged. With mounted batteries this is more difficult, especially with hard sides. I was thinking that with some stainless steel cable and pull rings I could make a quick ripcord setup to free the batteries with one pull, and with another pull ring yank the connectors and disconnect from the controller and drag the pack away from the bike with the wiring. Local armoring of the batteries would not preclude this, but a hard cover over the whole thing would have to be quickly removable somehow so as to not get in the way.

 

[gtadmin]

Agree it's a quandary, but my understanding of Lyen's incident was that it was real cheap LiPo and it was below LVC. With the stuff we use (25C LiPo, LVC protection) that should be less of an issue, but it's still a risk.

The other issue with easily removable batteries is of course security. I'm not sure how to solve this (yet!)

 

[Alan B]

Shifters and Bar-Ends

Did some testing and adjusting of the bar-ends for mounting twist shifters. They work great! Much easier to shift! I tipped them up toward the horizontal so they are perpendicular to my arms in riding position. Feels better. Now I can really use the gears which will be required for some bike paths and so on.

Magura Throttle Spanning

The Magura throttle is a bit "compressed" with a lot of "deadband" at both ends when you connect it straight to a controller expecting a Hall throttle. So I picked up a pair of multiturn PC board trimpots (5K) and put one in each the blue lead, and the black lead. With the bike on the workstand I adjusted the pots to eliminate deadband at both ends of the throttle, spanning out the throttle range. Now I have fine control of speed/power (about 2-3 times more rotation, uses the full range of the Magura). Much better control and feel. This is very useful for those cases where you want just a little power and still need good control.

I put the trimpots on a small piece of prototype pc board and put that inside a plastic 35mm film can. Later I'll do something better mechanically, but this works and is drip proof.

Motorized Bike "Death Race"

On an unrelated note, congrats to ES member PaulD who won the Tuscon Motorized Bike "Death Race" on an EBike last night! See the thread about that if you have missed it. Very exciting! And sympathy to Amberwolf who injured his ankle. Hope it is well soon.

 

[Alan B]

System Weight

I am often asked the weight of my ebike. Today I weighed it on a digital bathroom scale, differentially.

The configuration at weigh-in:

Aluminum mountain bike, REI Bonanza AL, 2000 model
Nine continents 6x10 rear hubmotor, 6 speed
hayes front hydraulic disc brake system 203mm
XTR rear V brake
12 FET Lyen controller
Magura throttle and grip
small bar ends
mirror
18S2P Turnigy 25V 5AH batteries
Topeak rack (trunk not onboard)
Thudbuster
Cycle Analyst
Schwalbe Big Apple rear and Marathon Plus front tire
Rhyno Lite rims
13 gauge spokes
No lights, lock, tools or pump onboard

And the total is: 63.2 Pounds

 

[Alan B]

Charging Cable

Setting up a 6S charging cable:

DB15 (female on battery pack)
pin : purpose
1,2,9 : negative charge current
3 : 0v balance
11 : 4v
4 : 8v
12 : 12v
5 : 16v
13 : 20v
6 : 24v
7,8,15 : positive charge current
10,14 : not connected

Basically this is done by using the three pins at each end of the DB15 for charging current, negative to the lower numbered pins;
Skipping one pin on each end from there,
Then going through the balance leads starting with 0 volts at the end closest to the negative charge current.

This keeps the voltage between adjacent pins to 4 or 8 volts.

 

[gtadmin]

Hi Alan

Charging Cable

Setting up a 6S charging cable:

DB15 (female on battery pack)
pin : purpose
1,2,9 : negative charge current
3 : 0v balance
11 : 4v
4 : 8v
12 : 12v
5 : 16v
13 : 20v
6 : 24v
7,8,15 : positive charge current
10,14 : not connected

Basically this is done by using the three pins at each end of the DB15 for charging current, negative to the lower numbered pins;
Skipping one pin on each end from there,
Then going through the balance leads starting with 0 volts at the end closest to the negative charge current.

This keeps the voltage between adjacent pins to 4 or 8 volts.

I read somewhere that the pins on these connectors were rated to 3A, which with your setup allows upto 9A of charge current. Is this enough?