LightningRods mid drive kit

[eTrailster]

I think all of these throttles including the Chinese cheapies are 5k Ohm potentiometers. It's a matter of quality, not difference in design. $15 for the cheap throttles, $50 for a Magura, $90 for the Italian made Domino.

http://www.electricmotorsport.com/domino-twist-grip-throttle-with-microswitch.html

Thanks Mike. I guess I'll be good then. Domino Twist Grip Throttle - 0-5K Ohm - IP67 Rated (Dust & Waterproof) $90.00 Shipped On: 2015-05-01. I ordered this from a golf cart part vendor because it is the IP67. IP67 = Dust proof & water proof to 10 feet under. IP54 = Limited dust protection & water resistant to splashes.

 

[Thud]

I think all of these throttles including the Chinese cheapies are 5k Ohm potentiometers. It's a matter of quality, not difference in design. $15 for the cheap throttles, $50 for a Magura, $90 for the Italian made Domino.

not exactly....most all of the uber chepo china throttles (twist or thumb) are in fact hall censored throttles.
http://www.electricscooterparts.com/throttles.html

gwhy has a nice durable design that uses a hall...(the hall advantage is no wiper wearing out or getting dirty & failing)

 

[stonezone]

gwhy has a nice durable design that uses a hall...(the hall advantage is no wiper wearing out or getting dirty & failing)

gwhy is a legend.

 

[LightningRods]

I got a visit this morning from my local sales rep from Gates Corp. We talked about belt drive of various types that could work well for the various stages of electric bike drives. I shared with him how we've gotten the power as far as the final drive where it's blowing up the bicycle driveline. Zero and Harley Davidson are using final drive belts that work great but are too wide for our applications. The sales rep is going to look into sourcing a motorcycle type final drive belt that is not as wide a gauge. Since we're not running the levels of power that Zero and Harley are we don't need that wide of a belt. It may also be possible to cut down the width of my current primary belt drives by using a higher spec carbon belt.

 

[r3volved]

Sweet! Has anyone busted a belt yet?

My ride is finally in motion. I'll need to try out my throttle tamer this weekend to eliminate the jitter. I have to get use to the full grip too...also thinking about the domino.

Rode to work today and got a lot of rubber neckers. A lot of comments from work colleagues

Tons of torque! It's going to blow my old chain sooner than later...that's probably my project for the weekend.

 

[LightningRods]

I'm not selling lots of replacement belts so I take that as a good sign. Mostly people wanting extras "just in case".

I need to get both Magura and Domino throttles to try out just so I know what to tell people about their value. Some folks are on a budget, some just want the best parts regardless of price.

 

[r3volved]

I'm already in balls deep...I want the domino
I just take my time so money isn't the issue.

 

[1BFC]

Pictures: https://drive.google.com/folder/d/0B0CrnjS7-w-iWG1iVWlUNHA3SUU/edit

Well I have had my BB kit installed with minor fit changes required to clear my 197mm rear dropouts on the Motobecane Sturgis Bullet for about 100 miles now.

Initially the 219 chain was difficult to keep tight but after the 5th or 6th adjustment it is holding exactly where I set it at. I think the key is having the two large bottom adjustment bolts pretty tight AFTER you do the adjustment and tighten the 6 smaller bolts. Once the 6 bolts are tight you go back and really crank on the 2 large ones and that prevents anything from moving. If you crank on the two large ones too much before locking down the 6 smalls, you over tighten the chain.

Try it if you have issues, worked perfect for me.

Total weight of my bike now is about 45lbs according to bathroom scale. I wear a 24s5Ah 20-30c turnigy lipo pack in a small camelback pack and have a range of about 25 miles at 10-12 mph Flat cruise. I limit the system power with a CAv3 and it shows about 1200watts needed to hit 30mph. Obviously there is enormous amount left on the table but I am ramping power up slowly.

The bike climbs anything with 26x4.0 tires.

To get the kit onto a 197mm frame required buying the 175mm BB kit from sickbikeparts. This included a longer crank "axle" that allows you to also use a 40t chaining from the same vendor that accepts the freewheel that Mike uses. You have to use the 40t from sbp because they mount the freewheel directly and not through the adapter Mike uses. using the adapter results in a chain line too far out for the 219 and no clearance for the 40t on the frame.

Using the 40t as the "adapter" allows you to just clean the chain stay with the 40t and still line up with the 219. To do this, you must bolt the adapter and 40t together using the bolt pattern for the freewheel. when you do this it will allow you to use the adapter as a drill bit guide to drill the holes into the 40t to mount the 219 sprocket directly to the 40t. The fit is perfect when you finish.

Ultimatetly Mikes sprocket adapter is used as a drill guide and then discarded as the 40t holds the freewheel and the 219 sprocket.

This sounds more complicated than it is and Mike has the pictures of what I did so he can support other 197mm frame builds.

I attached a picture of me holding the bike up. You can lift it over your head still but my ceiling is too low .

tech specs:

Bike: Motobecane Sturgis bullet
motor: LR Big Block
Controller: lyen 18fet
battery: 24s5Ah lipo
computer: CAV3
weight w/o battery: 45lbs
top speed: unknown (math predicts 57mph with 40t/11t final drive)
power needed to reach 30mph: 1200w

Even though this is a 100 volt setup, you can creep with the throttle and roll on smooth Thanks to the cav3. I haven't dialed in yet how to get "burst" acceleration that is smooth and low input latency to pop the front up like on a dirt bike, but I am not done adjusting the cav3.

Power to weight at full tilt 100v@100amp would be 3.5lbs/hp (or so). A Honda CR250R has 5lbs/hp.

No bike chain issues yet with power limited to 3kw or so. No slipping, no falling off, nada. It has been flawless. I can use 9 of the 10 gears with the 36t being just a tad too far off. You don't need this gear anyway....it climbs anything even in 40/13. Would likely climb at 40/11 but I know the 11 is not enough teeth to hold any power.

I won't be testing 100v100a....someone else can volunteer for that. I am limiting to 3kw and never, ever, worrying about motor heat.

Thanks for the support Mike. Happy to answer questions but may be sporadic at times due to my schedule.

 

[LightningRods]

Russ had a lot of challenges with this Motobecane because it has 4.8" tires and extremely wide chainstays that allow no clearance for chainwheels. I thought that fat bikes would be easy because of the 100mm wide BBs and conventional frame downtubes. No. The chainstays are the issue on fat bikes. I had a custom jackshaft made for Russ to line up with the super wide SBP square taper BB that he bought. Ultimately it all worked but with a lot of extra work for both of us. His bike now is like a halftrack hill climbing monster. If his bike can't climb a hill it tears it down and then Russ rides over the flat spot where the hill was.

 

[eTrailster]

I'm already in balls deep...I want the domino
I just take my time so money isn't the issue.

Now you guys want the Domino and I'm worried it may have been a mistake. It seems you need to solder a resistor in the power line in order to use a pot throttle with the CA-3. Not a big deal, but my soldering ain't pretty. Hey, maybe I can tell my wife I need a new soldering station.

 

[stonezone]

Even though this is a 100 volt setup, you can creep with the throttle and roll on smooth Thanks to the cav3. I haven't dialed in yet how to get "burst" acceleration that is smooth and low input latency to pop the front up like on a dirt bike, but I am not done adjusting the cav3.

If you dont mind, what are your throttle settings for your CA and which mode are you using?

super nice looking bike and build!

 

[Alan B]

I use a 560 ohm resistor in series with either the Domino or the Magura throttles to allow the protection against uncontrolled full throttle on a broken ground wire. This resistor could be installed inside the CA, though I usually put it in some heatshrink in the cable near the plug end of the throttle cable. The soldering is hidden inside the heatshrink so it is not visible, and a second layer of heatshrink over all the wires makes it part of the cable.

 

[r3volved]

For anyone with CA3 and full twist (from LR) and Lyen controller (from LR) - I found the following settings really eliminated the throttle dead spots and totally smoothed out my response curve:
MinVin - 0.90
MaxVin - 4.20
MinVout - 1.30
MaxVoit - 4.20

 

[1BFC]

I think your CAV3 throttle settings are going to vary quite a bit even if you use the parts you listed. I use those parts and have different settings. Believe it's the tolerance of the low cost electrical bits in the throttle that cause this.

Here's a short video of the bike in some urban riding. Nothing terribly challenging, but I haven't found anything at all that is a challenge for this rig.

http://youtu.be/YIXf27Cgojw

Might be a bit before that video is live, give it some time.

As for my cav3 settings, limit the amp ramp rate in throttle output menu and set the fast ramp to between 3 and 6. additionally, limit the VOLTAGE ramp as well. Amp ramp is something around 100. Voltage ramp is around 4-6.

If you don't get these right, setting the voltage ramp to 99 will still feel slugish. I find limiting the amp ramp to 50-100 (stock was 150) and reducing the voltage ramp made the biggest difference.

I find that i use the 3 speed switch a ton and that further smooths out low speed riding. I can creep at 1mph easily with everything set right.

On dirt with low tire pressures the cav3 showed 18xx watts to hit 30.8mph. This can be seen in the video near the half way point. Working backwards from 80% efficiency that is around 1400 watts motor power needed to achieve 30.8mph. On the asphalt path 450w = 20mph.

30mph + soft tires + dirt = 4x the power required to do 20mph. Pretty interesting.

I may try a larger primary 219 sprocket as I find with the 40t the bike feels best in the 13t rear. None of the lower gears are needed. Matter of fact, you can single speed this setup no problem. It's too bad I can't fit anything larger than 40t, but who is going to pedal at 30+mph off road anyway? My butt is hanging over the rear tire at those speeds

 

[LightningRods]

Great feedback. Thanks a lot for that. I still get e-mails from potential customers worrying about having their small front chainring. I try to explain to them the difference between a 300W mid drive and a 1500W+ mid drive. The low gears are not only unnecessary, they're almost unusable. Too much torque multiplication, too much instant throttle response. Another solution to jerky throttle response is to just grab a higher gear. It does much the same thing that the 3 way switch does.

 

[Samd]

Most throttles are hall sensor throttles. They output to the controller a voltage between 0 to 5 volts.

Most controllers trim about a volt from each end to remove noise, then rescale it. Depending on the combo you may not get full range because of the rescaling.

Some hall throttles have very poor resolution and are very chunky. You can literally feel the steps in throttle response. Some have 50 points or more and are very smooth. That can be good or bad - cruising at constant speed can be easier with a cheap@rse throttle but bad on high powered technical work.

Some throttles have a very linear distribution due to the magnet and hall used. As well as the geometry that dictates how the magnet swings to/fro the hall sensor.
Some have very nonlinear behaviour that ramps up fast.

Some controllers have different ramps for dishing up the current when going from standstill to WOT.

It's important to find a good combo and know how throttles work for your given conditions.

 

[teklektik]

Here's some illustrations that may clear up some of the throttle and CA V3 confusion:


Figure A shows the conventional hall throttle and ebike controller situation. These controllers accept a ZERO to WOT input voltage of about 1-4V because that is roughly what hall throttles deliver (this has nothing to do with noise - it is matter of the simplest design). The controller scales the input 0-100% over this range. The voltage output range of hall thottles slightly brackets this controller range so the controllers will not creep at ZERO and will actually develop WOT. This bracketing - both by design, lack of standards, and manufacturing variations can be sloppy which causes the dead zones at each end. In some case, the throttle voltage is so far off the controller may not see a high enough voltage to actually reach WOT.

Figure B shows the situation with potentiometer throttles. A primary difference is the expanded voltage range 0-5V. This is why resistors or trimpots are added to Maguras - to form a resistive divider and bring the ZERO and WOT voltages in line with the 1-4V range to match the controller (and hall) 'standard'. Failure to add such resistors leaves about a 1V (20%) dead zone at the each end of rotation.

Figures C and D show the hall and Pot cases mapped through a CA V3. The CA can be configured for any input range (ThrI->MinInput, ThrI->MaxInput) and any output voltage range (ThrO->MinOutput, ThrO->MaxOutput) and scales the configured input range to the configured output range. Generally the settings are made so the throttle min and max are detected slightly before they actually occur and the controller voltages are set a bit above and below the actual controller values to ensure full range and accommodate variations with time. In this case, pot throttles no longer require the external resistors because the CA addresses the expanded voltage range directly.

• It is foolish to copy these throttle settings from one CA to another!
  The adjustments are designed to compensate for differences specific to each throttle and each controller - using the settings from other bikes (almost certainly with different ranges) undermines the feature entirely. This can cause dead zones, creep at ZERO throttle, or failure to achieve WOT - all the problems the adjustment is designed to eliminate...

The CA V3 Guide outlines the one-time throttle setting adjustment procedure. It requires no meter or knowledge of voltage ranges and requires about 5 minutes to work perfectly. Easy.


Regarding the extra resistor for pot throttles on the CA:

• A common cause of bike runaway is a broken throttle Gnd connection which causes the throttle signal to rise to 5V. As can be seen in the illustrations, in the case of hall throttles, this error condition is easily discerned because the voltage rises above the max possible throttle voltage of about 4V. Not so with pot throttles where a failed Gnd sends the throttle output to 5V - exactly as does an intentional WOT. This makes the failure indiscernible from normal operation.
  
  The problem can be addressed either of two ways:
  • Disable the CA V3 input fault detection by setting ThrI->FaultVoltage less than ThrI->MaxInput or
  • Add a resistor to the high side of the pot to lower the intentional 'normal' WOT voltage making a 5V level identifiable as an error condition.

 

[eTrailster]

teklektik - You are an awesome communicator. Thank you.

I just read your Mundo build thread. Amazingly well documented. It may be the best I have seen.

 

[LightningRods]

Thanks for taking the time to lay all of that out Teklektik. Hugely helpful.

This is a great discussion and I'm going to keep digging into this to provide the best possible options for throttle and controller. I'm already getting e-mails asking if I have the Domino throttle and how it works with this or that controller. My current course is to work with Edward Lyen and to see how much the CA3 can improve the throttle ramp on his throttle/controller. Justin and the other good people at Grin Technologies are helping me tune this particular motor/controller/throttle combination on their CA3-DPS. I'm only going to switch horses if there is a clear reason to.

One advantage of hall effect throttles is that if they fail the throttle signal goes to zero and the motor shuts off rather than going WOT.

 

[eTrailster]

Domino arrived last night. It's not very impressive for $90, but I have nothing to compare. I was looking for a vintage Elsinore feel from my youth, but this isn't quite it. It seems to be durable sturdy plastic and the grips are comfortable. I like the range of motion and twist resistance. It will be several weeks before I will be able to test it and, due to limited experience, my opinions aren't worth much. However, I would advise others not to take the plunge before me or another wasteful idiot provides some feedback. Your money is likely better spent elsewhere on your bike.

If I want that old school MX feel, I may have to try the gwhy box and a cable throttle.

 

[LightningRods]

When you consider that the CA3-DPS is only $65 more than this fancy new throttle it does seem a bit excessive. I'm willing to bet that a CA3 tuned Lyen throttle will perform better than the Domino. If you have money burning a hole in your pocket and want both, that's fine. I certainly wouldn't get the Domino instead of the CA3.

 

[teklektik]

One advantage of hall effect throttles is that if they fail the throttle signal goes to zero and the motor shuts off rather than going WOT.

Well - there are several failure modes...

• A broken throttle Gnd wire will send the signal to 5V regardless of throttle type.
• A broken throttle Signal wire or 5V supply wire will send the signal to 0V regardless of throttle type because of the internal controller pull-down resistor.
• Regarding internal throttle failures:
  • An intermittent pot wiper will cut power as in (B).
  • The exact nature of a failed hall semiconductor device is unclear but will probably result in (B).

I believe that most throttle failures arise because of broken connections (A or B), not actual mechanism/electronic failures (C), but that may be a matter of debate. In any case, (A) is the really dangerous bugaboo and can be easily detected for either throttle type, either by the CA or by many controllers (if the throttle is directly connected without a CA).

• FWIW - the advantage of pot throttles is the improved linearity, although Justin reports very good linearity with the particular Wuxing hall throttle model he sells which utilize a long strip magnet to enforce a linear magnetic field over the rotation in lieu of the two magnets commonly used. That said, a crappy non-linear throttle will compromise the benefits of improved control from the CA Current Throttle since you will be in the garbage-in-garbage-out situation.
  
  Intended use is important as well. gwhy! is an aggressive rider and muscles the bike around with the grips. A motorcycle throttle can provide him with the physical robustness to withstand that type of use where plastic construction may not. On the other hand, I am a more sedate rider and the throttle does not take a pounding. I use a V3 and a Magura that has never failed in rain or snow for close to 10,000 miles. For me, the Magura is an excellent choice, for gwhy!, not so much. Pick components appropriate for the application - just sayin'...

 

[spinningmagnets]

The Cycle Analyst V3 has a ton of features that I don't know how to use, but for me...it is worth it to have due to the temp sensor allowing the CA to automatically limit (or possibly cut) the amps when the designated component is getting hot (motor or controller, whichever runs hotter on your set-up). A fried motor/controller costs enough to make this a highly recommended component as soon as funds allow.

If someone was planning on having a CA V3 from the beginning, that can make an "affordable" throttle a viable option.

 

[teklektik]

Justin and the other good people at Grin Technologies are helping me tune this particular motor/controller/throttle combination on their CA3-DPS.

They use the V3 on their StokeMonkey mid-drive with good success, but frankly there can be problems using Current Throttle mode with powerful mid-drives. Current Throttle works nicely with hub motors since it translates percentage of throttle rotation directly into percentage of maximum controller amps - 10% rotation with 50A max setting always gives exactly 5A to the controller. The result is smooth and easy control.

A difficulty can arise because the CA software PI controller has a hard time with the varying torque/speed applications per amp when in different gears - and the CA doesn't know about gears. So - the AGain adjustment can look out of whack in certain gears (affects surging/sluggishness).

As a fallback, if adjustment proves too challenging for the gearing and power of a particular mid-drive, you might try running the V3 in PassThru mode with a little curve shaping thrown in - instead of Current Throttle mode. (PassThru is the more familiar straight through operation with the addition of throttle ramping.)


If you use a resistive throttle (Magura, Domino, etc) you can add some shaping pretty easily as described in this post. As with any curve shaping, you get throttle resolution at one end and lose it at the other, but that can work out to assist in low speed operation. For example, curve 4 in the plot above scales the first 60% of throttle rotation to only 30% controller throttle voltage while leaving the full power (with reduced resolution) immediately available. This scaling technique is a hack and although simple to implement, has restricted curve shapes and is admittedly a PITA to setup/tune. Still - might be worth a look - particularly with a long-throw throttle like the Magura... (and the tiny $2 adapter can just be unplugged if it doesn't work out )

 

[stonezone]

just to add to the throttle discussion and the Lyen controllers:

http://endless-sphere.com/forums/viewtopic.php?f=16&t=14836

29. Bar Protect- This is a throttle fault protection feature. When active (1:Yes), it will shut off the motor if the throttle input is shorted to +5V. Normally WOT is around 4V, so if the wires short, or the hall sensor shorts, it should go close to 5V.

(it reads "drive saving" on my software but I'm guessing it's the same).