Power in Flux
Likes Likes:  23
Page 16 of 20 FirstFirst ... 61415161718 ... LastLast
Results 151 to 160 of 193

Thread: Transmissions. yeah.

              
   
   
  1. #151
    Senior Member
    Join Date
    Feb 2011
    Posts
    520
    Post Thanks / Like
    Quote Originally Posted by Brutus View Post
    Controlling the motor with no load has not been an issue at all because it is hooked up to a trans/primary and even when it is in neutral or clutch in at a stop there is drag on the primary drive and trans internally, more so than when rolling in a selected gear. Downshifting the trans at max rpm has not induced "lock-up" or "over rev" conditions(part of the reason to have 4-6 speeds instead of 2). Also I still say the clutch is useless for an electric other than another part to wear out or break.
    You're working with a brushed motor, I think it's series wound too right? So when you have rapid changes in RPM your controller doesn't know, care, or need to know or care. Your controller is equally happy running a hot water heater as it is a motor.

    In a brushless motor (way more so with a PM motor than induction) your controller needs to precisely always know rotor position to know which banks of FET's to switch on an off, getting the sequence off by a bit, even by 10deg of rotor position (which could be 100deg electrical off-set in a 10 pole motor) can result in instantly exploding the controller into plasma (by trying to clamp BEMF). Having large dRPM/dt makes it very difficult to resolve accurate rotor position in a timely manor to control the commutation correctly. Brushes actually rock for this, they just float and skip a bit when things get hairy, end up back where they belong, and can't really get out of time unless the whole assembly melts into charred goo. This takes all the difficult burden of controlling away from the controller.

  2. #152
    Not to be taken seriously DaveAK's Avatar
    Join Date
    Aug 2010
    Posts
    2,240
    Post Thanks / Like
    Quote Originally Posted by liveforphysics View Post
    In a brushless motor (way more so with a PM motor than induction) your controller needs to precisely always know rotor position to know which banks of FET's to switch on an off, getting the sequence off by a bit, even by 10deg of rotor position (which could be 100deg electrical off-set in a 10 pole motor) can result in instantly exploding the controller into plasma (by trying to clamp BEMF). Having large dRPM/dt makes it very difficult to resolve accurate rotor position in a timely manor to control the commutation correctly. Brushes actually rock for this, they just float and skip a bit when things get hairy, end up back where they belong, and can't really get out of time unless the whole assembly melts into charred goo. This takes all the difficult burden of controlling away from the controller.
    Interesting. So how does the controller cope with just normal slowing down of the bike? I assume that the bikes momentum will cause the rear wheel to drive the motor to some extent. Is this handled normally, and you're just raising the issue of a more extreme case caused by down shifting?

  3. #153
    Senior Member Brutus's Avatar
    Join Date
    Mar 2011
    Location
    Henderson Nevada
    Posts
    211
    Post Thanks / Like

    My Social Networking


    Visit Brutus's Youtube Channel
    Quote Originally Posted by liveforphysics View Post
    You're working with a brushed motor, I think it's series wound too right? So when you have rapid changes in RPM your controller doesn't know, care, or need to know or care. Your controller is equally happy running a hot water heater as it is a motor.
    Yup Series wound dc, because it is cheaper, rebuild-able, upgradeable, and doesn't have the issues you speak of in the rest of your post. I am happy to know I can pull stumps and re-purpose the bike into a water heater though. Can I classify it as a tri-purpose machine now?

  4. #154
    Seņor Member podolefsky's Avatar
    Join Date
    Oct 2010
    Location
    Boulder, CO
    Posts
    3,890
    Post Thanks / Like
    Quote Originally Posted by liveforphysics View Post
    In a brushless motor (way more so with a PM motor than induction) your controller needs to precisely always know rotor position
    Induction doesn't need to know position at all, just RPM.
    - Noah Podolefsky -
    The GSX-E

  5. #155
    Ghost Rider Allen_okc's Avatar
    Join Date
    Mar 2011
    Location
    Oklahoma City, Oklahoma
    Posts
    1,812
    Post Thanks / Like
    There's Tesla again - i wish i could have worked with him.

    as everyone stands in line to slaps me upside the head...
    The Greatest thing about a EV, is the Song of the Terrain...

  6. #156
    Senior Member Hugues's Avatar
    Join Date
    Jan 2012
    Location
    Switzerland
    Posts
    1,080
    Post Thanks / Like
    Quote Originally Posted by podolefsky View Post
    Induction doesn't need to know position at all, just RPM.
    ok,
    but then what's the purpose of the motor encoder on the AC-20 ?
    Regards from Switzerland
    My 2.5 Upgrade Thread

  7. #157
    Senior Member
    Join Date
    Feb 2011
    Posts
    520
    Post Thanks / Like
    Quote Originally Posted by DaveAK View Post
    Interesting. So how does the controller cope with just normal slowing down of the bike? I assume that the bikes momentum will cause the rear wheel to drive the motor to some extent. Is this handled normally, and you're just raising the issue of a more extreme case caused by down shifting?

    PMBLDC motors are equipped with systems to resolve rotor position. You can do it sensorlessly, by measuring the change in inductance of the phases, as its influenced by the magnets position in relation to the teeth. This involves taking precise measurements of inductance very quickly a thousand times a second. This is known as flux vector drive, or vector drive, or various other trade names. It's pretty awesome. Processor and ADC use heavy, $$$, can only work at low-ish rotor speeds.

    You can do it like an RC controller, have a pre-determined algorithm sequence it plays into the motor to attempt to get any rotor motion. It is only possible to have 1/3rd of available torque on the motor, while the other 2 phases are purely canceling their torque in the form of heating the motor. It works fine for RC propellers and very high power motors in light vehicles because in the first second you get enough rotor motion. Once the motor has rotor motion, it has high speed compairators watching between all the phases, and they send a signal everytime they have a zero crossing point on the BEMF. This gives the controller processor the input it needs to know magnet position, and once it has motion it can continue to run effectively. It's limitations are only a maximum of 1/3rd of stall torque or very low RPM torque torque while making massive amounts of heat to do it. This makes its use model OK in a bicycle (pedaling to get a little speed before hitting the throttle helps it work perfectly), but you're not going to be starting on a hill in a motorcycle or car with it, or trying to roll slowly under high loads etc.

    Hall sensors latch when the magnet passes them. When embedded at the right points in the stator, they pulse at the right times to provide the controller processor the info of where the poles are at in relation to the stator, so it knows which of the 6 fet banks to turn on. You can actually make processor-less analog controllers with hall sensors. This style has some weaknesses in getting any precision in rotor position, because it only can tell you a range of the magnet location. This makes them have a high torque ripple and be jerky at low RPMs (unless you do things like skew the stator lams, but this comes at a pretty big torque and efficiency penalty). This is the industry standard method for ebike hubmotors. They are kinda jerky and noisy and rough at low RPMs, you feel each pulse clearly almost like an engine chugging. They do handle high RPM's pretty well though, but they are never super accurate for any speed.


    Optical encoders have a disk with tons of little slits that flash by in a pattern that indicates position. It works well, it's high-speed capable, but if you have the speed rate change very rapidly, it only knows it's location by measuring the time between pulses, so having very rapid changes in the rate it gets pulses can fault them. Most are pretty damn good though, and do well.

    Sine-Cosine encoders read a spinning magnetic field very precisely, and output a sine wave and it's 180deg shifted buddy, the co-sine wave. The controller reads the potential between them and knows the exact rotor position at all times. It's high speed capable, and you're only limited by your ability to accurately sample and read this voltage. Good controllers sample it thousands of times a second, but it's still not fast enough to handle super rapid changes in rotor speed sometimes...

  8. Likes Nuts & Volts, protomech liked this post
  9. #158
    Senior Member
    Join Date
    Feb 2011
    Posts
    520
    Post Thanks / Like
    You can also combine any of these you like. Very often the top one gets combined with the second one, as the top one is great at very low RPMs, and the second one is great at high rpms.

    Combo's of #3 to start and #2 to run at high speeds are also done quite a bit.

  10. #159
    Seņor Member podolefsky's Avatar
    Join Date
    Oct 2010
    Location
    Boulder, CO
    Posts
    3,890
    Post Thanks / Like
    Quote Originally Posted by Hugues View Post
    ok,
    but then what's the purpose of the motor encoder on the AC-20 ?
    The encoder tells the controller the actual RPM of the rotor. Then the controller varies the field rotation frequency accordingly.

    On an induction motor, the torque is proportional to the difference between field rotation speed and the rotor speed (also known as "slip").

    http://www.engineeringtoolbox.com/el...lip-d_652.html
    - Noah Podolefsky -
    The GSX-E

  11. #160
    Senior Member T Rush's Avatar
    Join Date
    Mar 2018
    Location
    Iowa City, IA. USA
    Posts
    103
    Post Thanks / Like
    I've been reading thru a bunch of these threads, and Ted's blog
    ....and when I read on Ted's blog the TRANSMISSIONS VS. DUAL MOTORS he used to start this thread....it gave me an idea

    see, I think the basic idea behind the EV transmission is to change the compromise from "acceleration OR speed" to "acceleration THEN speed"....ok, but don't we really want "acceleration AND speed" ?


    and while some people enter this debate from the stand point of 'its fun to change gears and slip a clutch, gives my other hand and foot something to do' ...well, I guess you could leave those leftover controls from the ICE on the bike after the conversion, maybe hook them up to a noise maker, and those people would be happy just pretending without actually having a trans.....but thats not important
    ...and yeah yeah yeah, you still ultimately use the same amount of energy to go fast, slow, quick, long whatever no matter what gear you are in or if you got a big motor or little ones....or however that goes...to go just as quick or fast the same as another setup....but I think we are trying to use more power at slow speed, then change the drive ratio to be able to use more at fast speed...I guess like an ICE does, changing the drive ratios to allow the engine to run where it can suck the most air-fuel into power; thus use more energy and go quicker then faster


    so this 'dual motor' thing...sure, I get Ted's point that if you are going to add weight with a trans, you could maybe get the same result and weight with adding a second motor(doubling torque at lower speed, like a lower gear; while still maintaining the same gearing for highest rpm)
    ...but it seemed everyone(even he did?) take the 'dual motor' thing as only theoretical, and just assumed that in practice it would be easier to upgrade to one single larger motor thats twice the size....however either way that kinda puts you right back where you started with the single fixed speed(tho sure, now its more powerful) so you didn't really fix the problem, just the symptom, and only moved the problem to a different place(where you will find it again once you want more acceleration OR speed, and go back to still thinking about that trans to be able do the one THEN the other thing since you still cant have both)

    - and this is where I got my idea -

    what about looking at this the other way? say instead you start with one big motor with a big nice hp hump, and you split that motor into two smaller motors(and say their power adds together to give you the same hp hump as the one big motor on that dyno graph)
    ... so while you got the complexity of chaining all this stuff together with dual drive sprockets to the final drive on the wheel, here is where you have some fun using the idea of what a transmission does! say you are after something like a two speed trans with a higher and lower gears....see where I'm going with this?
    ...you put a different drive sprocket on each motor, say one has a 13 tooth and the other has a 17 tooth...but both motors are still being driven at the same time together, just at different RPMs
    ....now you don't quite get as high of a peak in the middle of your hp hump as with your single large motor, but the two smaller motors are still adding together for the mid-range overlap and averaging out to a nice flat plateau of hp...but one motor is now more efficient for pulling wheelies, while the other motor isn't out of breath for top speed
    ...you are running two gears at the same time to get that 'and' in the "acceleration AND speed" thing....maybe even optimized further with dual controllers running different mapping?
    ...sure, the two motors working best at different ends of the RPM range aren't going to be as strong as the one big motor geared to run slow OR fast; but thats only at a peak, and now you don't have to change gear ratios....and your pretty much going to be doing most of your riding in the mid range anyway, right where you made that band wider(and maybe actually taller vs slopping off quickly on either side of that one peak)...you have just flattened out your power band, you compromise max power at a single peak RPM to spread that out more to the low and high speeds...maybe even did something weird to the thermal issues and soaking, so that each separate motor can work harder if they are taking turns at different RPM, each able to hit their peak output at different times, for a longer total time together; ultimately being able to use more power, and longer, than a single bigger motor? idk

    here, look at this picture
    Quote Originally Posted by podolefsky View Post
    blue line is one big motor; yellow and red lines are two small motors with different gear ratios but say we have these two motors on the same bike turning the same chain at the same time, working together(now in that graph, the ratios are likely set at a normal compromise so that the bike operates fairly with a single motor; but with dual motors you could spread the ratios farther apart)

    ...the red motor is showing more power than the yellow under 35mph
    ...say at 10mph the red motor is transferring 9hp + yellow motor just doing 6hp = 15hp total
    ...next at 25mph the red might be pulling hard at 22hp, but the yellow is still adding 15hp for 37hp total
    ...and at 38mph they both are capable of 25hp each, for a total of 50hp
    ...then at 50mph the red motor is running out at 16hp + the yellow motor now up to 28hp = 44hp total
    ...but by 80mph the red motor has nothing left, while the yellow motor is still going
    now they might only combine to have 50hp max with the power peaks offset(vs closer to 60hp if they both focused on one speed, like the blue motor is doing only at 50mph) but together they average that power from 30mph to 45mph

    ok, you can go back to the other forum threads not about transmissions now, sorry, hope I didn't break anything
    Last edited by T Rush; 3 Weeks Ago at 0922. Reason: yells down the rabbit hole: hellloo, anyone still in there?

  12. Likes Richard230 liked this post
Page 16 of 20 FirstFirst ... 61415161718 ... LastLast

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •