Search all of my sites with Google

Tuesday, June 16, 2009

Death To All Controllers....

Ok, not ALL of them, but someone apparently sentenced mine. :(

Since it is made from a many-times-repaired Scoot'N'Go controller that started out it's life with me as already cooked *and* corroded, I'm not all that surprised. The issue this time is not in the power section, which I had upgraded last year to handle over 100 amps at 100 volts, using offboard-mounted MOSFETs and JFETs driving them (plus a diode pack for freewheeling currents). It is apparently the LM339 op-amp, probably internally shorted, as it drags down the supply voltage on it's side of the board to less than 3V.

All the other components are fine, except for one of the power resistors used in a voltage divider that provides the power to the LM339 section--it is burned open, probably from excessive current draw by the LM339.

I can't remove the LM339 and replace it without destroying the PCB, due to it's condition. It was not well-made to begin with, and what it went thru before I got it, causing burns and corrosion on it, would probably ensure that the copper will come off with the chip.

I *might* be able to Dremel off the chip by cutting it's legs, then carefully soldering a new LM339 in it's place by bridging from the legs to the PCB, but I am not confident it would work for long if it worked at all.

I am instead working on a version of the application note circuit for the Unitrode UC3637 PWM motor control chip, which is fairly simple. The big problem is that I have to do it on perfboard, at least at first, and I'm not sure how well that will work no matter how I beef up wiring, etc. If it all works out ok, I'll see if I can find the materials to etch a PCB for it if that would fix any issues I have with it, but I don't want to waste that kind of material if it doesn't work right to start with.

There are other motor control chips, such as the MC33035, which can also be adapted to brushed motor use (it's made for BLDC control), and I am slowly working on a design for that one that will let me use either kind of motor, with jumpers or switches for the configuration. It's not a "now" kind of project, though, and I need something right now to use with the motor on the bike.

Another thing still being worked on, hopefully integrated into this new controller, would be a version of that autothrottle controlled by pedal chain tension.

Yet one more thing to test is a converter circuit that would let me use a hall-based throttle on any controller, even one that requires a pot, by basically using the hall output voltage to control a digital pot. Then the digital pot wiper and terminals would connect in place of the pot throttle, and the hall throttle would connect to the controls of the digital pot (via some intermediate circuitry). It's actually pretty simple, but would probably cost more than just replacing whatever hall throttle someone has with a pot throttle. Still, if a non-replaceable hall throttle had to be used with a pot-based controller, like my autothrottle that has to be a hall sensor for environmental and mechanical-simplicity reasons, it would at least allow it to work!

Until I finish the controller and can test it, I may pull the motor and batteries off the bike so I can at least ride it around. I did patch the tubes sufficiently that at the moment it appears to be holding pressure, though I won't be confident of them until I can replace them with new ones (or at least used ones that don't have any holes in them!).


  1. Hi Mike. Since you have a brushed system, you can go simpler than converting a brushless controller. I have a simpler one I constructed on the free simulator, LTspice, in the files of my group. Ah, post took!

  2. Mike, I thought of a possible way to make a drive assy that uses a ring gear that attaches to the underside of the wheel rim. Then the motor has a meshing drive gear. The material of the ring gear could be a problem. --Joe

    PS. no need to post this Off Topic message.

  3. Re: controller; I have a brushed controller chip (UC3637) I already started building into a controller prototype, based on it's appnote. I thought I would finish it today, but the flat tire problem came back again, so I just yanked all the motor and batteries off the bike (around 60 pounds) and repatched it, which took the rest of my time today.

    Re: ring gear on rim, I've got a post sometime back (probably 2007 or early 2008) that I thought of something like that, to have a kind of hub motor but with a planetary gear arrangement. After some discussion and thinking, it's simply too much mass to deal with in the wheel, as frequently as I'd be spinning it up and down in traffic.

    Braking the extra mass would wear the brakes more, and the rim, and might even require disc brakes instead of rim brakes to do it quickly enough.

    Accelerating the extra mass would use up extra power every time, and require higher currents during that time.

    Engineering the ring gear itself is also beyond my machining capabilites, and probably would be for some time.

    It's not necessarily a bad idea, but it has a lot of gotchas that have to be dealt with. It's probably why Neodymics didn't end up building their design of motor and batteries all inside a wheel, to be driven similarly. :-)

  4. Indeed about the ring gear. It would probably have to be a material other than metal. But even that could get a bit heavy. The upside is that it should keep the bike balanced better do to gyroscopic action.

    I have some flats that simply seemed resistant to the patch staying sealed. It seemed to be related to the particular rubber composition of the inner tube. The last flat that I fixed with the Slime brand patch kit seemed to stick really well, though the inner tube is not that problematic one. --Joe

  5. Even if the ring gear were something other than metal, there are still mass issues with most of the durable enough materials. Those that are light enough may have either durability or fastenability problems.

    Gyroscopic action is undesirable (to me) in a bike, as when you need to turn it can prevent you from doing so as quickly or easily as you might need to. It's one reason I don't much like hub motors in front wheels, especially large-diameter ones (though they're not as bad as it could be, since the mass isn't at the rim).

    Of course, if the rim is really so heavy that gyroscopic action is significantly more than typical, it's probably so heavy that power requirements for starting up and stopping are high enough to put my worries elsewhere. ;-)

  6. 20, 2009 at 9:48 AM

    The ring gear should probably be made of nylon. It's light and durable. It can be fastened at the nut of each bike spoke.

    I wonder if the Zeta friction drive was a failure. It spreads the contact area over a length of flat belt. Even though it might be better than a roller in contact with the wheel, it seems kind of hard to implement still.

    I wonder if a bike can even be ridden by steering balance alone or if the gyroscopic action makes it possible at all. That's really got me wondering about something that I've wondered about other times in the past. --Joe


Alternate suggestions or improvements to anything that's been posted is very welcome, and extreme detail is preferred to brevity.

Keep in mind that unless you leave an email address in your comment, I haven't any way to reply to you except to reply to your comment here. That means if you want a reply, you'll have to come back to *this* blog entry and it's comments to see my reply to you, unless you leave some method of contact within your comment.