2 motors, 2 controllers, 1 throttle + 1 switch

I'm about to install a 2nd motor on my bike. The large direct drive on the rear is perfect for speed, but it needs help on the hills. On the front I'm adding a freewheeling geared hub, to hopefully give my speed demon a mountain goat climbing alter ego. I'd rather not use 2 throttles, and I can't use one in the normal manner I've seen, since I want to use the geared hub only intermittently.

My thought is to wire the 2 controllers in parallel to the throttle. Then just add in a momentary push button switch on the throttle hall signal wire for the geared hub motor's controller. In effect the single throttle will typically control only 1 motor, unless I push the button for dual motor mode.

Are there any holes in this plan?

John

just one. some controllers require the throttle signal to start from off and ramp up to full. if they are suddenly on, the sence a fault and shut off. I.E. if you just short the throttle wires together, it won't run.

But its a minor snag. you can let go of the throttle, hit the button, and twist again. Or you could build a small circuit to ramp up the voltage from the button to simulate a throttle being turned.

Then again, the controller might not have this safty feature at all. hard to say.

Thanks DrunkSkunk, my controllers have the feature for hi throttle protection on power up of the controller, but not on the throttle I know because I once forgot to power off when connecting a new throttle and the tire spun up when I accidentally touched the + wire to the controller sensor wire. I really plan on backing off the throttle to push the button anyway so as not to have as much impact on the planetary gears. They appear far beefier than the Bafang gears, but they're not metal.

John

Brilliant, John! Make one a geared free-wheeler and the other wheel a typical "speed wheel", and so the geared isn't interfering when the bike is in "speed mode"(No lost energy or complications) but yet you have all that extra hill-climbing ability(probably at least 3 times as much as the normal 1-x-lyte wheel user)!

That's... just... awesome. But, yes, also expensive.

John, I recently set up my recumbent bike with two Bafangs (Not documented yet, but will soon, both Bafangs purchased from and laced by JohnRobHolmes). The front wheel is a 20" Bafang and the BOB trailer wheel is a 16" Bafang. Unlike your plan, I made each motor 100% independent, with the throtle for the trailer on the left side. I haven't tested enough to make any specific conclusions, but I belive throttle balance when using both motors together will be important.

My intention is to use only the front motor for normal conditions for maximum efficiency on long rides and switch to the trailer motor on steeper hills. Will probably seldom use both motors together.

Anyway, my concern is that with just the one throttle on two very different setups your faster motor may be trying to pull faster than the other motor is designed to run, resulting in more work for the faster motor than if the other motor wasn't even there. When riding my bike without the trailer (just the front Bafang), even though the motor freewheels, when I release the throtle (at top motor speed) the bike pedals easier.

Looking forward to your tests with two motors. Good luck.

swbluto,

Expensive? I paid about the same price for 2 motors and 2 controllers as the retail price of a Puma and below retail of a X5.


Rassy,

On the flats I'd have to agree, but on significant hills the direct drive slows way down even with pedal assist. As it is slowed to 15mph and slower, it's efficiency takes a big dump. I also don't want to dump the whole load onto the little geared hub with it's spindly wiring, so running in parallel makes sense uphill and for takeoffs. I think what will happen is that yes, the DD will want to go faster, but the geared hub will end up only helping it go faster than it does alone or with pedaling too. I think the end result will be that I'll be able to go up quite steep hills and maintain 20mph+ without straining either motor or myself.

If this works well enough, then maybe I can get away with just a small tire for the front geared hub on the special purpose bike I'm building. If I still don't get the hill climbing "pull" I need, then the geared hub goes into the chainline as a bottom bracket drive to help climb the mountains. That's probably the route I'll have to go due to the wide variety of grades for some long uphill climbs. While most of long climbs are in the 5-7% grade range, there are some 1 and 2km segments at a solid 15% grade, so without gearing I'd have to build for the 15% grades due to the 375-400lbs of me, bike, and batteries to haul up the hills. The direct drive stays for it's speed and efficiency on the flats, along with electric braking and maybe regen too for the long downhill segments.

Other than just my special purpose bike, I want to come up with a simple and cheap drivetrain for use on roads, so weight down low isn't much concern. I want to far exceed the typical ebike performance with lackluster acceleration and WOT is how you ride. We don't drive cars that way, and I consider extra acceleration and higher than typical riding as a top end speed to have important safety benefits. In addition, with normal riding only at 50-70% of capacity, then a truly bulletproof system seems easy. If I can't pull it off with the hub motors I have, then the next step is something like a Mars or Agni motor on a bike. I believe the right vehicle to be between a motorcycle and a bike, and I'm headed toward street use and more utilitarian than the electrified downhill bikes.

John

John in CR said:

swbluto,

Expensive? I paid about the same price for 2 motors and 2 controllers as the retail price of a Puma and below retail of a X5.

Click to expand...

The ebike market must've changed recently or you just have some really good sources.

When you say "retail", I take it you mean from the same retailer and since ebikes.ca is the widely accepted one, I'll choose them. Ok, so the price of two low amp controllers is about the same as one high-amp controller.. so we can say the controller situation is pretty comparable. But the price of a crystalyte 40x hub and a free-wheeling geared hub is about the same as a 530x? Last time I checked, two crystalyte 40x's were about as expensive as one 530x and a single free-wheeling geared hub was as, if not more, expensive as a 530x! But, I may not be au courant on the market of geared hubs. Where'd you get yours? I think I may need to snatch me one. :wink:

You know, I'm liking this idea so much, I might actually buy a geared hub. So... pleeeeease, tell me your sources! :wink:

Frodus did something similar on his motorcycle and it seems to work out quite well.
You can split the throttle signal to two controllers no problem.
If you want to have one switched, you can put a swtich in series with the throttle signal to that one. Most controllers have a pull down resistor on the throttle input so if the wire comes disconnected (like with a switch), the output automatically drops to zero. The switch in the throttle line can be very small as there is not much current in that line.

Are you planning to have the dd motor on full time?

You might be able to make something like a trigger switch to activate the geared motor. You need something that is easy to hold onto while pedaling up a steep hill. Alternately, you could just use a toggle or rocker switch to turn it on, so you don't have to hold it. When the dd motor overruns the geared motor, the current drain from the geared motor will drop to no load (a few amps). I won't be a big deal if it stays on a few seconds when you don't need it.

I want to do some torture testing of this geared hub, so yes I'll have an on/off switch on the DD throttle. My thought is in normal operation to just have the DD on full time, since it doesn't make sense to me to drag the DD around just adding more load on the geared motor. Is there a reason why I shouldn't use the DD full time? Could a variable pot on the DD throttle line be of any benefit for long hills?

John

Post brought over from the Subject: INFINEON CONTROLLERS FOR DUMMIES thread

John in CR said:

ZapPat,
I'm not sure what you mean by synchronized, since the wheels travel different surface distances it's not possible.

Click to expand...

I'm thinking of a similar principal to subaru's AWD, but much easier to do since it's with a two wheel base. I don't quite understand why you say that the wheels travel at different speeds on a bike though? On a car I understand that anything other than straight driving (no turning) requires different wheel speeds, but on a two wheeled bicycle I don't see how this problem would come into play.

John in CR said:

Your ice e-bike sounds like loads of fun, and I look forward to being able to use all this new controller stuff instead of just read about it.
John

Click to expand...

I guess you won't be needing ice driving improvements where you are! However, I think the active 2WD concept would be great also for off-trail use, were a single wheel drive could spin in dirt/mud while the non-driven wheel sits and does nothing. Also, dual wheel regen will make for better braking in these conditions, and I'm hoping too for slightly increased efficiency here too from sharing the current between two motors and controllers (twice less current in each should result in two times less losses since P=I^2*R). All done automaticaly by the controller pair of course.


As for you actually using this kind of feature instead of just reading about it, well it'll take a bit of time since there are many many things to think about when taking one's own gadget and making it into a commercial product for others to use! Many small but very important details to make it more "fool proof" for the end user. Anyways, I'm starting with a basic controller (although it will be more advanced than regular ebike controllers anyways), and will add the more optional features afterwards. Docbass is going to be one of my testers, and I'll be looking for one or two others when the time comes - likely in about two to three months. I'm very empassioned about this project, and really want to help the ebike universe by actually listening to user feedback and making these features come to life!

John in CR said:

I want to do some torture testing of this geared hub, so yes I'll have an on/off switch on the DD throttle. My thought is in normal operation to just have the DD on full time, since it doesn't make sense to me to drag the DD around just adding more load on the geared motor. Is there a reason why I shouldn't use the DD full time? Could a variable pot on the DD throttle line be of any benefit for long hills?
John

Click to expand...

Personnaly I think you will get better efficiency by using both motors instead of only one almost all the time, exept maybe when you are riding along at high speeds on flats and when the hub motor would not be using much current even alone. With two direct drives I would use them both full time and evenly for sure.
To get real-world answers though we'll have to do tests ourselves I guess, since this is kind of "guinee pig" territory with very little data collected so far. I wish I could do a 2WD setup myself very soon, but I'll have to wait until I can afford to get two small DD hubs from golden motors or elsewhere, or maybe trade something with someone here for a second larger hub to use with the one I already have. Having two different hubs would likely make it more of a challenge and take longer to do, but should be possible too.

I use 2 (different) motors and 2 (identical) controllers and 1 throttle and 1 battery ... http://endless-sphere.com/forums/viewtopic.php?f=2&t=8610&start=15#p131469
The controllers share a common GND (obviously) and also a common 5V throttle input and throttle hall signal output.
Other than that the controllers are INDEPENDENT to drive their individual motors respectively.

The motors CAN BE DIFFERENT ratings. They will share the total load applied to the "system" (each motor only draws amps as it needs to).

ZapPat,

The motors truly can't be synchronized in their timing. The front wheel travels further, because the back wheel follows it around in a straighter line. eg to turn left, first you turn the front wheel to the right to get the wheels out from under you to create the lean enabling you to start the curve to the left. The rear wheel follow to the right some too, but in a shallower arc. Go make some tire tracks and see. Even riding as straight as you can, the wheel tracks can't coincide exactly. I bring this up because by synchronization I take it that the phase timing of the motors will exactly coincide, making one controller able to run 2 motors. It might work for a train, but not a bike.

My 2wd ambitions got thwarted temporarily by a slippage in some kind of clutch mechanism in the geared motor, so I'm modifying a big direct drive motor to fit it on the front. The motors will be identical, but I will run a slightly larger tire on the one in front. I don't see this as an issue at all with 2 controllers. They'll both be full time on, so no matter what I'll get less amperage dump on the two 25%+ grade hills I can't avoid. It's going to be a real speedster too, since I've had the single motor up to 50mph with a tail wind.

John

hi
if this 2x2 is going on ice alot a set of studded tyres might be a good idea.
the Infineon controller has the ability to alter the throttle to rpm ratio, I'm sure others do as well I just work with the Infineon.
On the infineon the ratio is set as a percentage of maximum in incriments of 1 percent from 30% to 100 plus some overdrive settings of 110%, 115% and 120%. If this type of thing was used then if you set the slower motor to say 100% and then set the second motor at a lower percentage speed setting this might well get the two wheels to travel at the same speed mph so causing minimum loss of grip.
This is theoretical as yet though I hope to make a 2x2 bike up myself.

Geoff

John in CR said:

The motors truly can't be synchronized in their timing. The front wheel travels further, because the back wheel follows it around in a straighter line.

Click to expand...

Duh! Why didn't I think of this, being so obvious? Anyways, thanks for pointing this out. I've now spent a bit of time thinking of how two controllers will act with two motors when the motors are forced to turn at different speeds...

geoff57 said:

if this 2x2 is going on ice alot a set of studded tyres might be a good idea.

Click to expand...

My two ice tires are now ready. About 130 screws in one, and over 300 in the other. Now it feels safe even poping wheelies on ice! Now to get some motors setup to push them, and I'll be one happy winter biker.

geoff57 said:

The Infineon controller has the ability to alter the throttle to rpm ratio, I'm sure others do as well I just work with the Infineon. On the infineon the ratio is set as a percentage of maximum in incriments of 1 percent from 30% to 100 plus some overdrive settings of 110%, 115% and 120%.

Click to expand...

Knuckles said:

The motors CAN BE DIFFERENT ratings. They will share the total load applied to the "system" (each motor only draws amps as it needs to).

Click to expand...

OK. So we have a rear wheel that slows down compared to the front one the more we are turning in a tight circle. If using one throttle and two controllers I can see different behavior depending on how the chosen controllers work. Simple controllers work by changing their output voltage in direct proportion to throttle demand, which is equivalent to motor RPMs, as geoff points out. Setting each controller's throttle to RPM ratio differently would make up for using motors having different RPMs/Volt, yes. But the question is: How will each motor react when we are turning? There are two answers; one for simpler throttle/RPM controllers, and a different one when using controllers that work in torque control mode (motor current control if you rather).

Situation 1: Regular throttle/RPM controllers (most ebike controllers)
As the back wheel slows down, the motor RPM goes down, and thus the produced motor voltage (BEMF) drops also. Since both controllers are still outputing the same duty cycle (same output voltage), this situation will force more current into the back motor wheel. The final effect will be that the bike will 'pushed' by the back wheel more as we turn, since it's torque will increase as we are turning.
Situation 2: Torque-mode controller (kelly,...)
In a similar setup but this time using two motor current controlled contollers (ie torque mode control), the outcome is not the same. This time as the back wheel slows down, the back controller will automaticaly lower it's voltage output to keep the same current flowing as in the front motor. This will have the effect of keeping the thrust distribution equal between both motors no matter how much the back wheel is slower than the front one. And I think this is what I'm after, as I don't want handling to be affected as I am turning corners.

Any opinions or ideas about this, guys?

ZapPat said:

Any opinions or ideas about this, guys?

Click to expand...

Yes ...

Build it.

Test it.

Ride it.

ZapPat said:

Situation 1: ...........
Situation 2: ............
Any opinions or ideas about this, guys?

Click to expand...

Yeah, you know what your likelyness of ever getting this tuned out without full computer controll is?
you're describing a need for stability controll, and on a bike, it would havwe to react diffrent for each wheel, based on load, torque, speed, and wheel RPM. If one wheel broke traction, it would have to know what to do, and when.

lose traction on the rear wheel in a straight line, you would want to cut power to the rear and add power to the front. But do that in a turn, if the rear wheel was slipping sideways, the sudden regaining of traction can cause the bike to flip over, and it's best to let the back wheel keep slipping as long as the angle can be maintained.

Lose traction on the front wheel, and you would want to kill all power to the rear smoothly, as you lose all ability to steer and balance when the front slips. But if you're trying to climb a hill, you might want it to lock the rpm of both wheels and let them slip

It would take a computer to be able to manage all the input data and regulate both wheels like that.
Or you can go the simple route, and skip all that.

I'd say just build the simplest one and ride it. Hard and Fast.

Drunkskunk said:

ZapPat said:

Situation 1: ...........
Situation 2: ............
Any opinions or ideas about this, guys?

Click to expand...

Yeah, you know what your likelyness of ever getting this tuned out without full computer controll is?
you're describing a need for stability controll, and on a bike, it would havwe to react diffrent for each wheel, based on load, torque, speed, and wheel RPM. If one wheel broke traction, it would have to know what to do, and when.

lose traction on the rear wheel in a straight line, you would want to cut power to the rear and add power to the front. But do that in a turn, if the rear wheel was slipping sideways, the sudden regaining of traction can cause the bike to flip over, and it's best to let the back wheel keep slipping as long as the angle can be maintained.

Lose traction on the front wheel, and you would want to kill all power to the rear smoothly, as you lose all ability to steer and balance when the front slips. But if you're trying to climb a hill, you might want it to lock the rpm of both wheels and let them slip

It would take a computer to be able to manage all the input data and regulate both wheels like that.
Or you can go the simple route, and skip all that.

I'd say just build the simplest one and ride it. Hard and Fast.

Click to expand...

That sounds like something a tiny microcontroller could easily do. No dedicated computers need apply!

You guys THINK way too much!
Two controllers work FINE together!
2x2 ebike is SO EASY and WORKS GREAT!
Just nice to have some SERIOUS BATS to drive all that POWER. :wink:
Build it. Test it. Ride it. THEN POST about it.
That's EXACTLY what I have done. F'n'A.



THINK
FINE
SO EASY
WORKS GREAT
SERIOUS BATS
POWER
THEN POST
EXACTLY
F'n'A

Knuckles said:

You guys THINK way too much!

Click to expand...

It's a time-money trade-off. It's often cheaper to think about and work out the kinks before finding out those two controllers didn't work out as well as one hoped.

It's this kind of planning that has helped me avoid some nasty logic errors in programming(Experience goes a long way, as well.).

Having LOTs of controllers to play with also helps.

A side benefit of "traffik'n" tech ... 8)


"Wanna Dance ... Gotta Take a Chance"

-KNUCKLES

http://endless-sphere.com/forums/viewtopic.php?f=2&t=8610&p=131469#p131469

Knuckles said:

You guys THINK way too much!

Click to expand...

I thought engineers LIKED to reflect on technical things? :wink:

Two controllers work FINE together!
2x2 ebike is SO EASY and WORKS GREAT]!

Click to expand...

I have no doubt that your setup works "great"... I'm just looking to see if there may be even better solutions possible. Your 2X2 ebike uses the "situation 1" as described above, but I am mostly wondering what using two different controllers than yours would yield. You can appreciate that there may be a big difference between using two regular controllers vs two torque mode controllers in such a setup?

That's EXACTLY what I have done. F'n'A.

Click to expand...

Send over two Kellys and two hubs and I'll test it out right away! swbluto makes a good point about $/time. Don't worry though, in a couple months I'll hopefully have my own torque mode controller on the road for tests, so practical tests will come eventually!


I think this thread was made to hash out ideas and theories and practical info about 2WD setups, and ideas and theories are important here since there are very few practical 2WD setups so far. Why not use our brains to aim for something better? I know with cars, having a well thought out AWD system helps much for handling in difficult conditions, so why wouldn't a well thought out ebike AWD system also help?

Enough people have already tried running two motors that I have no doubt that it will work fine.

Consider that even in voltage control mode that the no-load speed is always quite a bit higher than the loaded speed, so if the motors are anywhere near matched, they will share the load evenly.

If the motors are geared very differently, then as long as the slower one has a freewheel, things will be fine, but load will not be shared evenly.

Drunkskunk said:

Yeah, you know what your likelyness of ever getting this tuned out without full computer controll is?

Click to expand...

Situation 2 just uses two regular torque mode controllers, such as I think kellies could be setup to work (this just means that your throttle controls motor torque instead of simply outputing a fixed voltage as regular controllers in situation 1).

you're describing a need for stability controll, and on a bike, it would havwe to react diffrent for each wheel, based on load, torque, speed, and wheel RPM. If one wheel broke traction, it would have to know what to do, and when.
lose traction on the rear wheel in a straight line, you would want to cut power to the rear and add power to the front. But do that in a turn, if the rear wheel was slipping sideways, the sudden regaining of traction can cause the bike to flip over, and it's best to let the back wheel keep slipping as long as the angle can be maintained.
Lose traction on the front wheel, and you would want to kill all power to the rear smoothly, as you lose all ability to steer and balance when the front slips. But if you're trying to climb a hill, you might want it to lock the rpm of both wheels and let them slip

Click to expand...

This is exactly the kind of info I want people to post, thanks man! I know that there are many hardcore bikers here on ES that instinctively know how to use both brakes and pedaling combined in certain ways that helps them keep control of their bike during more extreme driving. Now, instead of only pedals and brakes, we have two active wheels that can offer both acceleration and braking through electronic means. I like biking for sure, but I mostly know electronics and microcontroller programming, and as such don't know how a bike should react during different riding situations as much as some savvy bike guys here. If people can explain a bike's ideal driving methods in a logical way, chances are that we can translate these action/reaction relationships into C language structures (or assembler if you're masochistic). Once you've build the whole base program for a brushless controller, it's only a bit more code to achieve some more intelligent behavior. And it usually doesn't take much processing power, just thought mostly...

I'd say just build the simplest one and ride it. Hard and Fast.

Click to expand...

That's what I plan on doing, don't worry about that. I just bought 5 lipo packs a few days ago, can't wait to try them with these destructive looking screw tires! So I am going to order two hubs shortly, direct drive or maybe geared without freewheel if I can get these somewhere not too expensively. I think solarbbq has the direct geared hubs, but in australia and I'm in canada ($hipping). Or maybe two nine continents would be nice?