Hi-Efficiency Recumbent Commuter Conversion

Show off your E-bike creation here.
User avatar
jkbrigman
10 kW
10 kW
Posts: 815
Joined: Oct 28, 2011 12:34 am
Location: North Carolina, Southeast US

Hi-Efficiency Recumbent Commuter Conversion

Post by jkbrigman » Dec 31, 2011 1:55 pm

This thread documents the conversion of a Short Wheelbase - Above Seat Steering recumbent bicycle.
swb.jpg
swb.jpg (54.65 KiB) Viewed 7588 times
Mission Objective:
30mi/56km range with less than 80% pack discharge or,
70mi/112km range with 80% pack discharge.
Cruise speed 18-25mph/32-40kph with 30mph/48kph burst capable if needed.

Bike Upgrades:
Planet Bike fenders
Schwalbe Marathon tires front and back
Tube liners for puncture protection
Rear rack
Cateye Mini-8 cyclocomputer
Headlight and taillight

The Build Kit (from methods/ebikes.ca):
9c 2810 motor in a 26" rim
36-72V controller
8 speed freewheel w/11T minimum
Direct-connect cycle analyst
Thumb throttle
Torque arms

Advantage and Motivation:
Using this bike for daily commute to pass stopped traffic in the bike lane on the commute route. Traffic backs up for over 5mi/8km, traveling less than 30mph/48kph, stop-and-go. At cruise speed, this bike may be able to achieve a commute time equivalent to an automobile over the same route at the same time.

Energy Budget:
At an estimated energy budget of 15Wh/mi or 10Wh/km this bike will require a minimum of 450Wh to go 30mi/48km. The plan is to build or buy a battery capable of at least 800Wh of energy storage for an 80% utilization. The design calls for a 1000Wh battery. This is possible on 72V, 15Ah. (18S-3P LiPo or 20S-1P A123 flat packs - oversized at 20Ah).

Bike Selection:
The chosen bike has a steel frame, huge sprockets on the front (up to a 52T) and tiny (down to 11T) on the back. The bike came stock with Kenda Kwest 1.25" 100PSI/6.8bar tires, which are being used successfully for ebike uses by some on the forum. Those tires will stay on the stock rims, using Schwalbe Marathon on the new rims for durability and puncture resistance while in use as an ebike.

This bike has no suspension components and uses V-brakes, therefore speeds over 30mph/48kph are not planned. Gearing on the bike (52T front, 11T rear) is intended to provide some power assist at cruise speed. Regen is planned and will come into play on 4 significant hill points in the commute - 2mi/3km of regen is estimated possible in either direction.

Through pre-project polling, members of the E-S forum rightly pointed out two key concepts: 1) That the commute will take upwards of an hour or more each way. A recumbent would make the 2-3 hours in the saddle each day easier and more comfortable. 2) Wind resistance of the recumbent is lower, therefore contributing to the potential for a lower energy cost, higher speed ride. I agreed with these points and switched focus from converting an ill-equipped upright bike to converting a moderately-well-equipped recumbent.

This bike uses a 26 inch rim in the rear and a 20 inch rim in the front. Electric drive using the front 20 inch wheel would be simple and lower-cost: no modifications to the chain drive whatsoever and the stock rear tire/wheel would be reused for ebike use. However, with the smaller diameter 20 inch wheel, rotational speed is slower than the 26" wheel, so more than 72 volts is required to reach the desired maximum speed. This voltage exceeds the intended voltage range for the controller. Furthermore, stresses and wear on the front forks and steering column would be out of spec for the design of the bike. Therefore a conversion of the 20" wheel was ruled out in favor of a 26" rear drive. Potential is there for a future dual-drive upgrade, particularly if research shows dual-drive could save energy on hill climbs.

Open Issues as-of this posting:
Still trying to figure out what battery to use. Advantage to the A123 packs is simplicity, completeness (cell_man includes a BMS with his batteries) and reliability. "000's of cycles" are what the A123's are advertised for, which provides a lifetime measured in years with good battery care. Charger purpose-built for the battery can be purchased from cell_man. Charging is simple and direct: buy two chargers, one for home, one for work, for daytime charging to make the return trip home. Disadvantage is cost. A pack for my required mission runs around $1k before adding in charger and shipping.

LiPo has a cost advantage and incredible current capability. It is also a flexible battery choice - you can change your pack to other voltages or trivially add Ah capabilities by paralleling-in more packs. This can be important if you have no clue how much current you will consume on your desired travel route. However, the disadvantage for LiPo is BMS (LiPo "hobby style" packs do not come with a BMS), cabling and expensive charging equipment. There are solutions to these problems available on E-S, allowing LiPo packs to be charged simply and reliably, but to get there is a "mini-project" with added time and expense upwards of $200.

The cost difference of the A123 vs. LiPo bricks may be little-to-nothing in the end. However, the ability to change the battery configuration (adding Ah) could become important. I do not know the answer to this question at this time, because there is no quantitative answer. Everyone's riding environment is different.

A project goal is to add fairings - particularly a front fairing and a tailbox. Front fairings are painfully expensive - a Zzipper product with hardware would run somewhere around $500US.

Ultimate goal of this project is to reach a minimal Wh/mile or Wh/km number, thereby opening the possibility for doing some long trips on the bike before having to swap out the battery or recharge - would like to exceed 100mi/160km on a charge if possible.

Open Questions on Batteries:
- do you "bulk charge" first at high voltage, then balance the pack, later, or intermittently?
- do you use a BMS of any kind permanently built into your batteries?
- if you DON'T have a BMS integrated into the battery, then how do you balance your cells?
- if you are using mean wells, do you have to mod the meanwells or do you just hook 'em up in series and go?
- if there is anyone charging 18S3P LiPo inexpensively, I seek your comment with data and pictures if possible
- suggesting I look at an existing thread is A-OK: there's an amazing amount of data here and my search-fu isn't so good

Questions and experienced comments solicited!

Thanks for reading! - James
Last edited by jkbrigman on Jan 22, 2012 10:51 pm, edited 21 times in total.
eRecumbent. 9C 2810, 6-FET,CA,18S3P LiPo, Kenda Kwest. 33mph max 12.8Wh/mi (22mph avg) Phat Bike Commuter HS3540, 6-FET, 18S3P LiPo, Schwalbe Marathon 47mph max 25 Wh/mi (28mph avg) Basic LiPo Charging Thread , Add Throttle to Multispeed Bike
dogman: "I bet if you just don't ride like a jerk, you can do whatever you want up to the speed where everybody looks at you like this. :shock: "
tomjasz: "Romans go full dork. They could give a hoot what 'Merkins do."

User avatar
Alan B
100 GW
100 GW
Posts: 7026
Joined: Sep 11, 2010 7:43 am
Location: San Francisco Bay Area, USA
Contact:

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by Alan B » Dec 31, 2011 2:13 pm

Great start James.

User avatar
amberwolf
100 GW
100 GW
Posts: 23948
Joined: Aug 17, 2009 6:43 am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by amberwolf » Jan 01, 2012 2:48 am

jkbrigman wrote:Open Questions on Batteries:
- do you "bulk charge" first at high voltage, then balance the pack, later, or intermittently?
- do you use a BMS of any kind permanently built into your batteries?
- if you DON'T have a BMS integrated into the battery, then how do you balance your cells?
- if you are using mean wells, do you have to mod the meanwells or do you just hook 'em up in series and go?
- if there is anyone charging 18S3P LiPo inexpensively, I seek your comment with data and pictures if possible
Based on the reported behavior of "good" RC LiPo (if the hobbyking type stuff can be called that; still it is better than a fair bit of other stuff reported here on ES and elsewhere :lol:), I'm planning on the following once I get to actually using the RC LiPo I am accumulating:

--Bulk charge, and periodically check the balance state of cells after a long ride. If they seem to be badly out of balance (tenth of a volt or more), then recheck them during and after bulk charge. If theyr'e still out of whack, it's time for a balancing. :)

--No BMS per-se, but a cell-level LVC to notify me it's time to stop pulling power from them. Most likely it will be the Methods LVC/HVC boards, as I have a set now for up to 18s, and up to 4s for a lighting pack. They will also be used to stop bulk charge should any cell really go overboard for some reason. ;)

--Balancing would be done manually, with a "single cell" charger. (in my case, that means an adjustable voltage/current lab supply, since that's what I already have laying around).


I'm not using MW, but AFAIK with at least some of them you may have to add a current limiter to one; there's a thread on how to do this by Fechter, I think. No link though, sorry. :( No answer to the last question, either.

User avatar
Alan B
100 GW
100 GW
Posts: 7026
Joined: Sep 11, 2010 7:43 am
Location: San Francisco Bay Area, USA
Contact:

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by Alan B » Jan 01, 2012 6:12 am

I am charging 18S2P at the moment.

Here is a minimum configuration that I am using to charge this ebike:

At home I use a Turnigy 8150 RC charger, it will do 150 watts or 6 amps into 6S which is 25 volts. It is a balancing charger. I connect it 3 times - once to each 6S block. The battery remains series connected in 18S at all times. This charger requires a 12V power supply, I'm using a Meanwell 350 watt. If a 15 AH pack was drained it would take about 3 hours to charge each section. A higher current charger would reduce this time as would a lesser draining. This is a balance charge so it can take longer if the cells are out of balance.

At work I use a BMSBattery bulk 4 amp charger. This unit runs from 120 VAC. A 15AH empty pack would take about 4 hours to charge.

In any case it is not a good idea to drain the pack more than 80%, and if it is drained that far a balance charge is recommended.

Each of these two chargers cost about $40, and the 12V Meanwell was about $60 not including shipping.

One improvement would be to purchase two more Turnigy 8150's and two more power supplies. Then all three sections could be balance charged at the same time. Separate power supplies are required since the battery remains in series and their ground potentials must be isolated.

For LVC I use the Cycle Analyst right now, but I am planning to use Methods HVC/LVC modules as well. The Cycle Analyst does whole pack voltage detection while the Methods boards do cell level protection. Both pull down the throttle when there is a problem.

Note - the BMSBattery chargers are not UL listed, so not something you want to use at the workplace. Maybe even not a good idea to use at all.
Last edited by Alan B on Nov 09, 2014 10:16 am, edited 1 time in total.

toft
10 W
10 W
Posts: 81
Joined: Jun 10, 2010 5:12 am

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by toft » Jan 01, 2012 7:05 am

I'll be watching. Recumbent could be my next conversion. Nice thread:).

User avatar
jkbrigman
10 kW
10 kW
Posts: 815
Joined: Oct 28, 2011 12:34 am
Location: North Carolina, Southeast US

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by jkbrigman » Jan 01, 2012 10:44 pm

Thank you guys! (Alan B and amberwolf), much, much appreciate your feedback. Toft, stick around. Even if I fail, there will be lessons and learning published.

A Battery Start
I've initially resolved the question of "what battery". I'll purchase a small number of low-capacity LiPo packs, a small charger, a couple cables, and do testing with that technology to characterize the energy per mile required by the recumbent ebike. I say initially because I'm unwilling to close the door on other battery technology - I simply need to figure out how much energy the commute route will take.

I have a general feel how the bike should perform, but no knowledge precisely how much Wh-per-unit-distance is required. It's vital to figure out how much energy I'll need to have available to "do my mission". For the type of use I intend, an A123-style pack IS a better choice, but A123 packs require an up-front investment of $500 or more and are not easily reconfigurable. (this is not exactly true - you could open up the enclosure, add or remove cells as needed, and either modify or change out the BMS. But the point of buying an A123 or PingBattery type of pack is to NOT open it up, but to purchase exactly what you need from the start.)

Why LiPo Packs?
It's (somewhat) cheap and reconfigurable.

You have to be careful about how you charge the battery. If you break it down into parallel "bricks" of relatively low voltage, you can charge a brick with a simple 11-18V power supply and a hobby-style LiPo balancing charger. This kind of approach is OK for learning about how to use LiPo or other Li-chemistries, but I have yet to determine if it's a reasonable long-term approach. I'm expecting to have to "mess" with a lot of cables, and expecting not to be able to charge the entire pack all at one time - I expect to have to use the pack on the bike via high voltage/high current series wiring and charge the pack via low-voltage parallel wiring.

ggoodrum has a nice BMS he sells, usable for LiPo, that will eliminate cable jockeying while caring for individual cells. You could probably build a nice LiPo pack with his stuff. Problem is, I'm no better at understanding how much of his hardware I need to buy until I understand how much battery I need to buy.

What about charging?
This is the tipping point. I can charge LiPo packs "cheaply" because I have some low-voltage DC power supplies around the house. I can't charge a sealed, complete A123 pack because I don't have a DC supply that can go as high-voltage as I need in the battery.

I already have several power supplies of 15-19VDC at 90-150 watts. This is suitable for feeding power to a small, hobby-style balancing charger. I have pestered several kind souls on the E-S forums to find out how they charge, and two members I've come to trust use LiPo bricks and balancing chargers. I'll go the same way just to get the wheels turning and collect some data, then invest in the longer-term battery later.

Another idea that will help lower charging cost: you can drive a hobby-style balancing charger with something even as small as a laptop power supply: those units put out 18-19V at around 2-4 amps. It's not much, but it will work. The connector on the laptop supply probably won't mate with what's on the balancing charger, so you'll need to cut the connector off the laptop power supply and match it to the balancing charger. This kind of setup will charge a large pack very, very slowly, but it will work. I have an unused laptop power supply and intend to experiment with this idea.

Some people are also converting PC power supplies. You could drive a balancing charger with the 12V line from a PC power supply.

What about a BMS?
I will use the balancing charger to keep the cells in balance and the low voltage function of the Cycle Analyst to monitor overall battery discharge. I am counting on the balancing charger to help me spot a bad cell during the discharge cycle. When I have fully characterized the energy needed for the commute, then I'll go with the final battery using a full-on BMS, regardless of the chemistry.

Isn't LiPo Dangerous?
LiPo can be dangerous if the cells are abused.

There are clear risks that have been demonstrated and described in the forums by running a damaged cell or imbalanced pack. But balancing chargers are (somewhat) low cost and readily available. Information available in the forum on how to use a balancing charger with LiPo batteries indicates they can be safely and repeatedly charged and discharged.

What's LiPo Going to Cost?
I've spent around $200: three 5000mAh 6S 20C LiPo packs, one balancing charger, some misc. connectors. I'm shopping parts now to make the parallel balancing connectors (JST-XH for charging only) 4mm bullet plugs/jacks for the high-current harness and an Anderson Powerpole for connection to the motor controller. I do not expect this battery to be big enough to make the entire commute, or even just one of the to/from work legs. If you want more detail on exactly what I've bought, post a question in this forum or PM me.

What Next?
First I'll learn to make a reliable battery, then how to reliably charge it, then see how far I get with it.

JKB
eRecumbent. 9C 2810, 6-FET,CA,18S3P LiPo, Kenda Kwest. 33mph max 12.8Wh/mi (22mph avg) Phat Bike Commuter HS3540, 6-FET, 18S3P LiPo, Schwalbe Marathon 47mph max 25 Wh/mi (28mph avg) Basic LiPo Charging Thread , Add Throttle to Multispeed Bike
dogman: "I bet if you just don't ride like a jerk, you can do whatever you want up to the speed where everybody looks at you like this. :shock: "
tomjasz: "Romans go full dork. They could give a hoot what 'Merkins do."

User avatar
jkbrigman
10 kW
10 kW
Posts: 815
Joined: Oct 28, 2011 12:34 am
Location: North Carolina, Southeast US

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by jkbrigman » Jan 09, 2012 11:35 pm

Took the recumbent out for a short 10 mile ride just to remember what it feels like to ride it. Looking forward to the conversion. The test ride reminded me that mirrors are not an option, but a requirement.

Turnigy 500mAh 6S 20C bricks have arrived. Took 2 weeks for them to get here "from the USA warehouse". The weight of the ordered bricks are heavier than I imagined - the negative side effect of my derating exercise. I can see why ggoodrum built a 12S2P plug'n'play accepting "only" four of the bricks: https://endless-sphere.com/forums/vi ... 14&t=30021

I'm wondering what is possible with the smaller, four-brick pack. Anything larger requires the expensive BMS Lite system https://endless-sphere.com/forums/viewto ... 95#p501227

Balancing charger was ordered about a week ago - had to buy the "nice" one, as the lower-cost Turnigy Accucel-8 is sold out - another add-in cost of the LiPo. No idea how long it will take the charger to get here. The bricks will have to wait till then to get a charge - the power supply I'll use with the balancing charger isn't high enough voltage even to attempt a bulk charge, (6S LiPo is 22.2v nominal, minimum 24V supply) - much less a balancing charge. I won't attempt a bulk charge for the first time - the cells need to be verified no damage in shipping. I'd rather know the cells are balanced from the start.

I finally got to see up-close what the bullet connectors look like - good thing I ordered some from ebay to make harnesses -they are not compatible with conventional old "banana plugs". (does anyone know if the 4mm bullet plugs are compatible with something else, like EC3? )

Got the Schwalbe Marathon tire installed to the drive wheel Had to adjust the V-brakes to clear the massive Schwalbe tire. Fenders fit over the Schwalbe, thank <insert diety here>..... Got the 8-speed freewheel installed to the motor, but it won't clear the dropouts. It did not come with the stainless washer, so I've got to order one.

Note to self: order JST-XH connectors from digi-key to make balancing PCB's. Moral of the story? ebike builds go slow - every detail matters and the parts are not easily available....
Last edited by jkbrigman on Jan 10, 2012 8:51 am, edited 3 times in total.
eRecumbent. 9C 2810, 6-FET,CA,18S3P LiPo, Kenda Kwest. 33mph max 12.8Wh/mi (22mph avg) Phat Bike Commuter HS3540, 6-FET, 18S3P LiPo, Schwalbe Marathon 47mph max 25 Wh/mi (28mph avg) Basic LiPo Charging Thread , Add Throttle to Multispeed Bike
dogman: "I bet if you just don't ride like a jerk, you can do whatever you want up to the speed where everybody looks at you like this. :shock: "
tomjasz: "Romans go full dork. They could give a hoot what 'Merkins do."

User avatar
Alan B
100 GW
100 GW
Posts: 7026
Joined: Sep 11, 2010 7:43 am
Location: San Francisco Bay Area, USA
Contact:

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by Alan B » Jan 10, 2012 12:01 am

If you are going to get Methods HVC/LVC boards they have the JST-XH's for up to 7P already installed. You might want some JST-XH extension cables, plus the double female from Methods to balance charge through the HCV/LVC board.

User avatar
jkbrigman
10 kW
10 kW
Posts: 815
Joined: Oct 28, 2011 12:34 am
Location: North Carolina, Southeast US

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by jkbrigman » Jan 10, 2012 1:29 am

Alan B wrote:If you are going to get Methods HVC/LVC boards they have the JST-XH's for up to 7P already installed. You might want some JST-XH extension cables, plus the double female from Methods to balance charge through the HCV/LVC board.
Alan - I think you just answered a lingering question I had: can methods LVC/HVC boards charge through the balance connector WITHOUT the high-current harness? I thought they could NOT, so I put off the methods boards and planned to make my own...
eRecumbent. 9C 2810, 6-FET,CA,18S3P LiPo, Kenda Kwest. 33mph max 12.8Wh/mi (22mph avg) Phat Bike Commuter HS3540, 6-FET, 18S3P LiPo, Schwalbe Marathon 47mph max 25 Wh/mi (28mph avg) Basic LiPo Charging Thread , Add Throttle to Multispeed Bike
dogman: "I bet if you just don't ride like a jerk, you can do whatever you want up to the speed where everybody looks at you like this. :shock: "
tomjasz: "Romans go full dork. They could give a hoot what 'Merkins do."

User avatar
veloman
1 GW
1 GW
Posts: 3402
Joined: Sep 14, 2009 1:06 am
Location: Austin TX

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by veloman » Jan 10, 2012 3:02 am

Are you considering adding a front fairing and/or rear tailbox? Adding both you would probably go from 25mph to 28mph realistically. The rear tailbox would hide your hub motor and make you look very stealth, plus added cargo room.


I will one day do a recumbent build. I'm held back by where I live is not conducive to fast non-stop riding, and lots of cars. In the future more bikeways will be built though.


Your build log is very organized. I need to take notes for my threads, lol.
Mush! Mush you electrons! Push harder!
I offer electric powered lawn mowing service in Austin http://efficient-ventures.com/

User avatar
Alan B
100 GW
100 GW
Posts: 7026
Joined: Sep 11, 2010 7:43 am
Location: San Francisco Bay Area, USA
Contact:

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by Alan B » Jan 10, 2012 8:02 am

jkbrigman wrote:
Alan B wrote:If you are going to get Methods HVC/LVC boards they have the JST-XH's for up to 7P already installed. You might want some JST-XH extension cables, plus the double female from Methods to balance charge through the HCV/LVC board.
Alan - I think you just answered a lingering question I had: can methods LVC/HVC boards charge through the balance connector WITHOUT the high-current harness? I thought they could NOT, so I put off the methods boards and planned to make my own...
Charging is generally done using both the heavy wires, and balancing through the JST-XH. It is a function of the charger's features. Most of these LVC type boards have parallel connectors for balance connectors. If you are going to get HVC/LVC boards that have parallel JST-XH connectors then there's no need to make your own parallel boards. I made my own JST-XH paralleling boards because there were no readily available LVC boards when I needed them, and because I wanted to use DB type connectors for my charger balance connections. On the newer build for my son I made a small adapter board and used the Geoff57 LVC boards. So the four battery JST-XH connectors go to the LVC board, and a fifth JST-XH connector goes from the LVC board to my DB adapter board. The high current wiring harness includes a PowerPole for the charging current. So two plugs, one for current, one for balance for each "block" of parallel batteries. On my 18S bike a "block" is 6S, and uses a DB9 for balance, and there are three "blocks" to charge. On my son's bike a "block" is 10S so a DB15 was required, and he only has one block to charge.

Most chargers put charge current through the heavy wires and do balance through the JST-XH. The small wires on the JST-XH are only good for a few amps. Some small chargers will do everything through the small wires, but those are probably not what you want.

So just a quick review. The LVC function is a safety net for discharge and clamps the throttle. The HVC function is a safety net for bulk charging and shuts down the bulk charger. My son's bike doesn't need HVC since it is 10S and always balanced charged every time.

On my 18S bike there are two charging scenarios. One is "bank balance charging" and each 6S bank is charged separately. They are still in series all the time, so a 25V difference exists between their grounds so care is required - either charge only one of three "banks" at a time, or use three chargers each with separate power supply. I use one charger and plug it in three times. For bulk charging I use a 4.15V per cell 4 amp bulk charger that shuts itself down. The HVC function opens up the charging circuit (when I get that part done) for an extra safety net. For monitoring checks I use a celllog wired to a DB9 plug so I can quickly plug it in and check six cell voltages at a time on a bank. So three plug-in cycles are required to observe all 18 cell voltages.

The Cycle Analyst is really the first line of defense against overdischarge. Setting the pack low voltage per cell to 3.6 provides good protection, and watching the amp-hour discharge allows staying away from even getting that low. The board level LVC is the third layer of defense against overdischarge.

So in summary we have defense in depth. On charging there are two layers, one is the charger itself and the other is HVC protection. On discharge we have three layers - amp-hour metering, pack voltage monitoring cutoff and low cell voltage cutoff.

Note - plugging in three times to charge gets old after awhile.
Last edited by Alan B on Nov 09, 2014 10:19 am, edited 1 time in total.

User avatar
jkbrigman
10 kW
10 kW
Posts: 815
Joined: Oct 28, 2011 12:34 am
Location: North Carolina, Southeast US

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by jkbrigman » Jan 10, 2012 9:38 pm

veloman wrote:Are you considering adding a front fairing and/or rear tailbox? Adding both you would probably go from 25mph to 28mph realistically. The rear tailbox would hide your hub motor and make you look very stealth, plus added cargo room.

I will one day do a recumbent build. I'm held back by where I live is not conducive to fast non-stop riding, and lots of cars. In the future more bikeways will be built though. Your build log is very organized. I need to take notes for my threads, lol.
veloman, thanks for the props, much appreciated. I'm trying to write down all the information either shared with me by the awesome bro's on E-S or that I've learned on my own. I hope to post pics soon of a rear DD motor "box opening".

Yes, I definitely want to do a front fairing and rear tailbox. The whole idea of this bike is to see how far, how fast, and how efficient I can go. But front fairings I've seen (zzipper.com) are expensive ($500). I think I'll have to make one. Tailbox ought to be relatively easy to make from coroplast.
eRecumbent. 9C 2810, 6-FET,CA,18S3P LiPo, Kenda Kwest. 33mph max 12.8Wh/mi (22mph avg) Phat Bike Commuter HS3540, 6-FET, 18S3P LiPo, Schwalbe Marathon 47mph max 25 Wh/mi (28mph avg) Basic LiPo Charging Thread , Add Throttle to Multispeed Bike
dogman: "I bet if you just don't ride like a jerk, you can do whatever you want up to the speed where everybody looks at you like this. :shock: "
tomjasz: "Romans go full dork. They could give a hoot what 'Merkins do."

User avatar
jkbrigman
10 kW
10 kW
Posts: 815
Joined: Oct 28, 2011 12:34 am
Location: North Carolina, Southeast US

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by jkbrigman » Jan 10, 2012 10:13 pm

Alan B wrote: On my 18S bike there are two charging scenarios. One is "bank balance charging" and each 6S bank is charged separately. They are still in series all the time, so a 25V difference exists between their grounds so care is required - either charge only one of three "banks" at a time, or use three chargers each with separate power supply. I use one charge rand plug it in three times. For bulk charging I use a 4.15V per cell 4 amp bulk charger that shuts itself down. The HVC function opens up the charging circuit (when I get that part done) for an extra safety net. For monitoring checks I use a celllog wired to a DB9 plug so I can quickly plug it in and check six cell voltages at a time on a bank. So three plug-in cycles are required to observe all 18 cell voltages.

The Cycle Analyst is really the first line of defense against overdischarge. Setting the pack low voltage per cell to 3.6 provides good protection, and watching the amp-hour discharge allows staying away from even getting that low. The board level LVC is the third layer of defense against overdischarge.

So in summary we have defense in depth. On charging there are two layers, one is the charger itself and the other is HVC protection. On discharge we have three layers - amp-hour metering, pack voltage monitoring cutoff and low cell voltage cutoff.
Alan - great charging summary! Thank you a ton for explaining what you do and how you do it. I plan to modify and improve the way I charge and discharge the pack when I figure out exactly what the pack needs to look like to accomplish the mission...so I'll be asking more questions AFTER I get the pack and an initial balance charger built!

I've been reading your Greyborg build. It's the most complete and current (har) information I've ever seen about high power ebike building - thanks for that thread!
Last edited by jkbrigman on Jan 16, 2012 12:04 pm, edited 1 time in total.
eRecumbent. 9C 2810, 6-FET,CA,18S3P LiPo, Kenda Kwest. 33mph max 12.8Wh/mi (22mph avg) Phat Bike Commuter HS3540, 6-FET, 18S3P LiPo, Schwalbe Marathon 47mph max 25 Wh/mi (28mph avg) Basic LiPo Charging Thread , Add Throttle to Multispeed Bike
dogman: "I bet if you just don't ride like a jerk, you can do whatever you want up to the speed where everybody looks at you like this. :shock: "
tomjasz: "Romans go full dork. They could give a hoot what 'Merkins do."

User avatar
Alan B
100 GW
100 GW
Posts: 7026
Joined: Sep 11, 2010 7:43 am
Location: San Francisco Bay Area, USA
Contact:

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by Alan B » Jan 14, 2012 12:06 pm

How is it coming, James?

I've never ridden a recumbent. Was doing some reading on it, I like the idea of better comfort for long rides and better efficiency. Is it controllable enough to deal with traffic with confidence?

User avatar
jkbrigman
10 kW
10 kW
Posts: 815
Joined: Oct 28, 2011 12:34 am
Location: North Carolina, Southeast US

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by jkbrigman » Jan 16, 2012 12:30 pm

Alan B wrote:How is it coming, James?

I've never ridden a recumbent. Was doing some reading on it, I like the idea of better comfort for long rides and better efficiency. Is it controllable enough to deal with traffic with confidence?
I wouldn't say it's going "great". Hitting a few snags that are adding up to really stop my efforts. (more about that later)

Recumbents
The SWB (Short WheelBase) Recumbent is a mixed bag when it comes to stability. Straight line and shallow curves are great. Turning an acute curve is more difficult than with an upright diamond bike. If you turn the front wheel too far, your foot can hit the front wheel and you can lose balance. The wheelbase of the bike is so incredibly short that I would not say it "helps" stability on where it's placed. If the front forks and steering had been placed another couple inches forward, I think that would have been better. Problem with that is, your feet on the pedals would not have cleared the front wheel at all. It's not so bad that I am worried for my safety - you ride the bike enough, you learn what to do and what not to do.

If you ride in shorts (lycra or otherwise) in sunny weather, sunscreen is a requirement - your legs are stretched out perpendicular to the sun's rays for maximum exposure. First time I ever rode the SWB, I got a bad sunburn.

Mirrors are an absolute must on a recumbent. It's impossible to turn your head and look back without stopping.

There is some transmission of vibration through the rear wheel into the back of the recumbent seat, but it's not terrible. The time I noticed this, i was running 100psi in the rear on a dirt road. Otherwise, no problems and the heavy padding of the seat is superb for cushioning road shocks even though I lack any suspension at all.

If I were buying a new recumbent to convert, I'd go with an LWB: Long WheelBase machine. Although many recumbents have the "feet hitting the front tire" problem, you can find LWB models that don't, so they are more stable. Caveat is that I can't comment with authority, as I've never ridden one. Notable names in that field are
Sun http://www.sunbicycles.com/index.php and RANS http://www.rans.com/products/bicycles

(Side note: IMHO I do not believe the "crank forward" semi-recumbent designs offer any efficiency advantage compared to a typical upright diamond bike. That's just my opinion on efficiency - I have looked at the Townie bikes with great lust - I think a crank-forward with disc brakes would make a great conversion. The retro "Juicer 48" https://endless-sphere.com/forums/viewto ... =7&t=34845 essentially became a crank-forward recumbent with the seat post mods done by the builder. But the builder didn't have any impressive comments to make about the range and efficiency of the bike.)

Having said these things I must point out that the ride comfort is superior to an upright bike. You can hop on this bike after months of not riding and go for miles without suffering any lower back pain. There's no "butt or crotch problem" with this bike like there is with the saddles on upright bikes. I took the bike out a couple more times before working on it and traveled some of my well-known and well-timed ride routes. Even though the recumbent isn't quite as good going uphill, the bike kicked the crap out of my upright bikes for ride speed and time to complete the course. Without even trying, I bested my previous-best upright diamond bike ride on my 10 mile loop by almost 2 minutes. On a loop that short, a 2 minute improvement is normally near-impossible for me. All I can figure is that I didn't lose much time going uphill and I must have gained immense time on the 3 miles of downhill over the 10 mile loop. I've just added a speedo to the bike so I may be able to better confirm that theory.

Installation of the motor and battery may improve stability. Putting that big, heavy motor down on the axle and mounting the battery pack low will move a lot of weight down on the axles.

The Build
The stock torque arms don't match up to any bolt holes on the bike, so I've got to either modify them or make new. Everyone is out of stock on the freewheel tools, so I can't transfer the one from the bike to the motor, which was the original plan (the one from the bike will fit without modifications - the one you buy with the motors is a couple mm wider, necessitating the spacer washer. This part was said to come with the gear cluster, but it doesn't, actually) I can show the motor wheel on the bike (it's massive!) but it's not really road-ready yet.

No one has the rear freewheel tool - it's out of stock everywhere. I plan to go to the local bike shop today and get a stock shimano freewheel tool. I'll have to drill it out to get it to fit over the motor shaft.

Bullet connectors ordered from China have taken a painful amount of time to arrive. Tracking indicates they are in the state, could be here this week. LiPo charger has left Hong Kong but tracking gives me no better or newer information than that.

I have some things to tell, so I plan to post a couple new pics shortly - parts unboxing and the usual added stuff to the bike - fenders, rack, lights and speedo. I'm suffering some delays waiting on parts and I'm planning the wiring harness builds, but I haven't gotten any parts in yet, trying to be patient, but I'm starting to worry. Any shipping time over two weeks gets me nervous.
Last edited by jkbrigman on Dec 17, 2012 11:25 pm, edited 1 time in total.
eRecumbent. 9C 2810, 6-FET,CA,18S3P LiPo, Kenda Kwest. 33mph max 12.8Wh/mi (22mph avg) Phat Bike Commuter HS3540, 6-FET, 18S3P LiPo, Schwalbe Marathon 47mph max 25 Wh/mi (28mph avg) Basic LiPo Charging Thread , Add Throttle to Multispeed Bike
dogman: "I bet if you just don't ride like a jerk, you can do whatever you want up to the speed where everybody looks at you like this. :shock: "
tomjasz: "Romans go full dork. They could give a hoot what 'Merkins do."

User avatar
jkbrigman
10 kW
10 kW
Posts: 815
Joined: Oct 28, 2011 12:34 am
Location: North Carolina, Southeast US

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by jkbrigman » Jan 16, 2012 1:25 pm

Alan - you were asking about the build, I thought I'd add one more comment. Unfortunately, I've learned that voltage costs more - lots more - when you are building up a LiPo battery. Conversely, adding Ah capacity (putting bricks in parallel) is relatively cheaper.

I say this because:
- most BMS units or chargers are designed to work with lower-voltage (6S to 12S) packs
- high-current and balancing harnesses cost more for higher-voltage (12S) packs
- You can buy balancing harnesses "all day long" for 6 packs, but not for 9 packs in 18S
- Power supplies that can serve as cheap bulk chargers come in 24V and 48V. 72V and up are harder to find and more expensive.
- If you choose to balance charge using hobby chargers, you've got to go with multiple chargers, one for each parallel bank

I'm pushing ahead with 18S because I think the higher voltage will pay off in performance and range and I'm expecting the 2810 will simply lose less power in heat than the 2806.

JKB
eRecumbent. 9C 2810, 6-FET,CA,18S3P LiPo, Kenda Kwest. 33mph max 12.8Wh/mi (22mph avg) Phat Bike Commuter HS3540, 6-FET, 18S3P LiPo, Schwalbe Marathon 47mph max 25 Wh/mi (28mph avg) Basic LiPo Charging Thread , Add Throttle to Multispeed Bike
dogman: "I bet if you just don't ride like a jerk, you can do whatever you want up to the speed where everybody looks at you like this. :shock: "
tomjasz: "Romans go full dork. They could give a hoot what 'Merkins do."

User avatar
Alan B
100 GW
100 GW
Posts: 7026
Joined: Sep 11, 2010 7:43 am
Location: San Francisco Bay Area, USA
Contact:

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by Alan B » Jan 16, 2012 1:59 pm

Great to get an update.

I'm waiting 3 weeks now on the borg frame, hard to have patience. Installing a garage door opener distracted me for a couple days.

In my reading on recumbents I came across a comment that surprised me a little. It makes sense after thinking about it but initial reactions were surprising. I did more reading on it and a number of things came together.

Bikes are more stable when the weight is high and forward. Surprised? The wikipedia article on bicycle dynamics goes into it in some detail. But the bottom line is that weight to the rear should be low for best control, but weight forward can be high for better stability.

On your torque arms, did you see what I did? I don't like hose clamps, or slotted arms that depend on tension to stay in a particular spot. So I use the ebikes.ca torque arm but make my own very simple torque strut. I try the arm on the axle in various ways until I find a spot that allows me to make a strut between the end of the torque arm and an existing hole in the frame. The holes for mounting fenders or racks have worked out well for me so far. Then I make a torque strut from aluminum or steel 1/8" by 1/2". It is very simple, having two precise holes one for the torque arm and one for the frame hole. I make the holes small and ream them to just big enough to minimize any free play. In my case I didn't have enough room for a nut near the freewheel so I threaded the strut and did away with the nut. It looks very simple but is actually very strong. At this radius even aluminum is strong enough for the 9C's torque.

One other thing to consider is the reaction force to the torque arm. This reaction force vector is perpendicular to the torque arm. This should not be pushing the axle out of the dropout slot. Ideally it should push in the strongest direction for the dropout. If the torque arm is vertical this force will be horizontal which is a good direction for vertical dropouts.

Wiring the battery up is a bit of a chore. Getting nicely made wiring harnesses from icecube does make that problem go away. From what I've heard the other commercial harnesses aren't great, but I have not tried them.

User avatar
jkbrigman
10 kW
10 kW
Posts: 815
Joined: Oct 28, 2011 12:34 am
Location: North Carolina, Southeast US

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by jkbrigman » Jan 22, 2012 7:12 am

LiPo balancing charger has arrived, the "iCharger 106". (I preferred the Turnigy Accucel-8 - $30 cheaper - but it was not in stock.) The good thing is that the iCharger 106 is a 250W balancing charger, so it will use every single milliwatt of power my 150W 15V supply will provide. The supply is a surplus unit I built into a basic radio shack box. I added binding posts for + and - output, AC input, power switch, fuse and AC line filter. It works like a champ plugged into the iCharger. There were no additional connectors or cables required, the charger has a DC supply cable with two banana plugs on it that fit the binding posts perfectly. They are color coded, but not polarized: therefore you have to pay attention to match red-red and black-black when you plug in the balancing charger!

The iCharger comes with a few short cables: a very short charging cable with alligator clips, a thermal sensor, a USB cable and a tiny cd disk of firmware. None of the cables are usable for charging LiPo packs: the vendors went with alligator clips because they can't predict what you will be charging with it. You could jam individual bullet connectors into the alligator clips, but that's not mechanically or electrically sound.

I bought 50 pairs 4mm bullets off ebay. I desoldered the alligator clips and installed the bullets (being careful to match plug and receptacle to the battery pack) and covered them with heat shrink tubing. Again, the are not in a plastic polarizing housing - so you have to pay attention when you plug in.

I connected up everything with power off - hit the power switch on the power supply and it all came to life. Selected the LiPo balancing charge cycle and turned on the balancing charger. It defaults to a nice, leisurely 2A charge rate and had the pack up to full charge in about 2 hours. The fan on the charger only intermittently came on and it never got hot. I continuously watched the pack to make sure there were no problems or faults and it all charged up without incident.

I've made a charging thread over in "Battery Technology", complete with photos of the setup from unboxing to charging:

https://endless-sphere.com/forums/viewto ... 50#p523523
Last edited by jkbrigman on Jan 24, 2012 11:16 pm, edited 2 times in total.
eRecumbent. 9C 2810, 6-FET,CA,18S3P LiPo, Kenda Kwest. 33mph max 12.8Wh/mi (22mph avg) Phat Bike Commuter HS3540, 6-FET, 18S3P LiPo, Schwalbe Marathon 47mph max 25 Wh/mi (28mph avg) Basic LiPo Charging Thread , Add Throttle to Multispeed Bike
dogman: "I bet if you just don't ride like a jerk, you can do whatever you want up to the speed where everybody looks at you like this. :shock: "
tomjasz: "Romans go full dork. They could give a hoot what 'Merkins do."

User avatar
jkbrigman
10 kW
10 kW
Posts: 815
Joined: Oct 28, 2011 12:34 am
Location: North Carolina, Southeast US

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by jkbrigman » Jan 23, 2012 9:20 am

Does anyone know of a good source for the plastic polarizing housings for the 4mm bullet connectors? I got a bunch of the bare connectors off ebay:
http://www.ebay.com/itm/50-sets-4-0mm-4 ... 1c2232116e
that I'm using to make cables right now.

They are usable as-is, adding heat shrink to give the parts a little more stability and protective insulation. But the plastic polarizing housings would add more mechanical robustness and protect the connectors from shorts. So I'm looking for the plastic housings. Any comments/thoughts?
eRecumbent. 9C 2810, 6-FET,CA,18S3P LiPo, Kenda Kwest. 33mph max 12.8Wh/mi (22mph avg) Phat Bike Commuter HS3540, 6-FET, 18S3P LiPo, Schwalbe Marathon 47mph max 25 Wh/mi (28mph avg) Basic LiPo Charging Thread , Add Throttle to Multispeed Bike
dogman: "I bet if you just don't ride like a jerk, you can do whatever you want up to the speed where everybody looks at you like this. :shock: "
tomjasz: "Romans go full dork. They could give a hoot what 'Merkins do."

User avatar
Alan B
100 GW
100 GW
Posts: 7026
Joined: Sep 11, 2010 7:43 am
Location: San Francisco Bay Area, USA
Contact:

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by Alan B » Jan 23, 2012 9:25 am

The pins need to have the right grooves to lock into the housings. They are always sold as a set that I know of. EPBuddy may have them, or Hobbyking. Also eBay.

User avatar
jkbrigman
10 kW
10 kW
Posts: 815
Joined: Oct 28, 2011 12:34 am
Location: North Carolina, Southeast US

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by jkbrigman » Jan 23, 2012 4:32 pm

Alan B wrote:The pins need to have the right grooves to lock into the housings. They are always sold as a set that I know of. EPBuddy may have them, or Hobbyking. Also eBay.
I found out disappointing news last night: you can't heat-shrink over the 4mm bullet "socket" - it makes it too fat to fit into the plastic receptacle for the mating connector. :-(

I fear you are right (that they are always sold as a set). I can't find anything with only the housings on eBay. I'll try EPBuddy and Hobby King! I wonder: could something like this be made from a cut-and-drilled piece of HDPE?
Last edited by jkbrigman on Jan 24, 2012 12:24 am, edited 1 time in total.
eRecumbent. 9C 2810, 6-FET,CA,18S3P LiPo, Kenda Kwest. 33mph max 12.8Wh/mi (22mph avg) Phat Bike Commuter HS3540, 6-FET, 18S3P LiPo, Schwalbe Marathon 47mph max 25 Wh/mi (28mph avg) Basic LiPo Charging Thread , Add Throttle to Multispeed Bike
dogman: "I bet if you just don't ride like a jerk, you can do whatever you want up to the speed where everybody looks at you like this. :shock: "
tomjasz: "Romans go full dork. They could give a hoot what 'Merkins do."

User avatar
Alan B
100 GW
100 GW
Posts: 7026
Joined: Sep 11, 2010 7:43 am
Location: San Francisco Bay Area, USA
Contact:

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by Alan B » Jan 23, 2012 5:12 pm

The two pin and plastic body set is about 0.30 at HK.

You can probably do the icecube trick of shrinking over the connector and wire to make a hood that has space inside as needed for the connector body tube.

User avatar
Alan B
100 GW
100 GW
Posts: 7026
Joined: Sep 11, 2010 7:43 am
Location: San Francisco Bay Area, USA
Contact:

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by Alan B » Feb 11, 2012 10:37 pm

How's progress here?

I didn't quite make the Mega ride today, unfortunately, but I did make progress and managed a pedaling test after dark with the hubmotor installed. Tomorrow I should be able to complete the wiring and do an electric test run. :)

For details see bikeE thread in my sig.

User avatar
jkbrigman
10 kW
10 kW
Posts: 815
Joined: Oct 28, 2011 12:34 am
Location: North Carolina, Southeast US

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by jkbrigman » Feb 11, 2012 10:47 pm

Alan B wrote:How's progress here?

I didn't quite make the Mega ride today, unfortunately, but I did make progress and managed a pedaling test after dark with the hubmotor installed. Tomorrow I should be able to complete the wiring and do an electric test run. :)

For details see bikeE thread in my sig.
Alan - sorry about the Mega ride - Will check your thread for more details.

I need the freewheel tool - going to see if I can drill out a stock shimano tool as has been previously described here. Have also hit snags where the mounting hardware for the rear rack doesn't quite work on my recumbent, so I'm going to have to cut down a couple of the M5 allen screws that came with the rack. Photo update of the machine coming soon ...
Last edited by jkbrigman on Feb 26, 2012 1:11 am, edited 1 time in total.
eRecumbent. 9C 2810, 6-FET,CA,18S3P LiPo, Kenda Kwest. 33mph max 12.8Wh/mi (22mph avg) Phat Bike Commuter HS3540, 6-FET, 18S3P LiPo, Schwalbe Marathon 47mph max 25 Wh/mi (28mph avg) Basic LiPo Charging Thread , Add Throttle to Multispeed Bike
dogman: "I bet if you just don't ride like a jerk, you can do whatever you want up to the speed where everybody looks at you like this. :shock: "
tomjasz: "Romans go full dork. They could give a hoot what 'Merkins do."

User avatar
Alan B
100 GW
100 GW
Posts: 7026
Joined: Sep 11, 2010 7:43 am
Location: San Francisco Bay Area, USA
Contact:

Re: Hi-Efficiency Recumbent Commuter Conversion

Post by Alan B » Feb 11, 2012 10:54 pm

I sure know how that goes. I have a kit of axle washers, D washers, D spacer washers, NordLocks, Bottle mount socket screws of various lengths, connectors, wire, pins, and leftover bits and I still had to make several hardware store trips and grind and file various parts to make them fit. Order a few extra bits, it will pay off sooner or later.

Post Reply