Running 2 controllers on one bike alternately - EE question..

Where i ride, i intermittently have to go bicycle speeds, and car traffic speeds. It's difficult because there is no such thing as one tune that works best for both. I don't know of any modern controller that has multiple programmable tunes that you can switch on the fly with a physical button. So i'm thinking about running two controllers.

Also, when going fast, i believe in the old adage that loud pipes save lives. I want motor to scream when you throw thousands of watts into it, but be completely silent when cruising along a bike path with pedestrians or other cyclists.

So here's the idea.
Hook something like an 18FET infineon/xie chang clone and 6FET FOC controller like a phaserunner in parallel to the motor.
The 6FET FOC gets tuned so that it only produces a top speed of 20mph and has much weaker acceleration.
The 18FET xie chang is tuned for 'balls to the wall', but still has a 3 speed switch so that you can very between 30, 40, 45mph.

There is one throttle, and the ground/negative is paralleled between both controllers, but the other two wires are switched between the little 6FET and the big 18FET using a double pole switch.

Are there any possible holes in this plan? I wonder if 5000w peak flowing through the phase wires may leak some current back into the smaller controller and make it go kaboom.

The Powervelocity controllers can be programmed with 3 maps using the 3-speed switch. It works great as you can program each setting for current limit and speed limit. I use one on my A2B. Low setting is "bike legal". Medium setting is "moped legal", and the high setting is "off road only" but you can dial these in anywhere you want.

Two controllers in parallel on the same motor can work. It would risky to run both at the same time, but I've heard of people doing it before successfully. You would also want the negative battery wires from both controllers to be joined very securely and as close as possible to each other.

Oh! that's an interesting solution. Is it possible to switch sinewave mode on and off, and have that set in the profile?

Also, does it do torque based throttle or any fancy stuff like that?

neptronix said:

Oh! that's an interesting solution. Is it possible to switch sinewave mode on and off, and have that set in the profile?

Also, does it do torque based throttle or any fancy stuff like that?

Click to expand...

Sine wave mode is not selectable. It works well and is very quiet.

It doesn't have a true torque mode either. It seems to be a blended torque/speed. Whatever it is, I like the response. It's very controllable at low speeds.

There is an optional Bluetooth phone app that allows you to reprogram things on the road. Regen level is fixed but you can set it over a wide range. The app has a dashboard you can use as a speedometer, etc.

It has something like field weakening so it can go faster than a trapezoidal unit for the same voltage.

If you got the larger one, you could program one of the speeds for something really weak, like 5A and have the option for also limiting the speed.

I'm a bit confused. What is the problem with the infineon 18 fet with the 3 speed switch?

neptronix said:

Is it possible to switch sinewave mode on and off, and have that set in the profile?

Click to expand...

Newest PV versions have sensorless/fault mode, where it switches to trapezoidal. Just switch off/unplug one hall.

district9prawn said:

I'm a bit confused. What is the problem with the infineon 18 fet with the 3 speed switch?

Click to expand...

Ah.. have you not done a high power build with an infineon clone?

Due to the 'speed control' methodology of the controller, a 3 speed switch is basically just a speed limiter which causes the controller to deliver full power until it tapers off into the top speed specified. In my case, i will have >100ft-lbs of torque available. This is a huge problem. You don't want an accidental power wheelie to happen on a sidewalk with pedestrians around.

The only way around the low speed throttle control issue to use torque control or variable tunes.
Controllers with proper torque control ( usually FOC types ) that i have seen do not have 3 speed switches and such. No selectable variable tunes that limit the peak torque. Your only option is to dig through menus in a cycle analyst frequently while riding. I want a physical switch because i have to switch between modes frequently due to the inconsistent design of our bicycling infrastructure, which changes from block to block..

My bike needs to crawl along at 15-20mph in the city and combine pedal power with motor power reasonably, but also be able to blast up a 7% grade for 13 miles at 35mph, and do 45mph on the flats.

There other reasons i want two distinctly different tunes/modes:
1) The bike looks outlandish and i expect the risk of legal trouble to be higher. The high power tune should be enabled using a key switch, allowing me to remove the key and disable the high power mode.
2) A lower powered mode would be ideal so that i could hand my bike to a first time ebike rider for a spin. I often demonstrate my ebikes at local events. Even 1.5kw on a high torque, high efficiency setup is enough to spook most people.
3) I want a completely silent motor at lower speeds and the most motor noise possible at high speeds.

neptronix said:

Your only option is to dig through menus in a cycle analyst frequently while riding. I want a physical switch because i have to switch between modes frequently due to the inconsistent design of our bicycling infrastructure, which changes from block to block..

Click to expand...

What about using the CA's presets to change modes on various things, which can be changed via one of the Aux inputs (and the other Aux could be used to change a single parameter, separately)?

There are even handy switches like these already premade, or you can wire up your own:

https://endless-sphere.com/forums/viewtopic.php?f=31&t=98145

This combination 2 button digital and 3 position auxiliary input switch can be configured to control two of the following Cycle Analyst V3 parameters:

Mode presets
Amps limit
Speed limit
Power limit
ThrO limit
PAS level

Click to expand...

I don't really like the use of piggyback controllers.
I have not figured out how to program the bit of jitter and delay out of the CA V3 it on higher torque setups...
Would rather this sort of thing happen at lower power levels.
If you do get it dialed in, limiting battery amps but keeping the 250 phase amps i'll have set for max power probs gives questionable throttle action too.

The powervelocity idea suggested by fechter is closer to feasible, but would rather not run in sensorless mode at high speeds to meet the noise on/off requirement.

I don't know hte PV controllers, but there are sensorless controllers that work well at high currents (and high speeds, but I haven't tested beyond 20-25MPH personally).

For example, Incememed's SFOC5 is sensorless-only, and torque (current) throttle control, and works really well once you get up to speed...smooth, silent, but very powerful and rapid acceleration, when you throttle up, but holds a particular torque well when you hold the throttle, and not jumpy, etc. (Using the MXUS 4503).

Other people with more-normal-mass bikes and DD hubmotors (like the Stealth), it also works well from a stop, according to testing posted about in his thread.

But with DD hubmotors and a lot of mass to accelerate like my SB Cruiser trike, it hadn't done well from a stop, and didn't apply much power at speeds below several to 10-15MPH. But he's done work on the firmware since then, and I'm still testing and working on the settings for my MXUS 4504 to see how it performs now that it has some specific startup-situation updates.


There's also the Lebowski controllers, which do sensored at low speeds and switch to sensorless once moving well enough, but those are kind of a lot of DIY (and still not cheap), so might be more work than you want to do. (I wish his new version/algorithm for full-power startups was available, though). I haven't built mine yet (lots of parts to buy), so no direct experience with them.


FWIW, there doesn't seem to be an issue with running sensorless once a motor has reached a certain speed; it's generally only at lower speeds where there is not enough BEMF for positional feedback that sensorless controllers can have difficulty.

neptronix said:

The powervelocity idea suggested by fechter is closer to feasible, but would rather not run in sensorless mode at high speeds to meet the noise on/off requirement.

Click to expand...

It stays sine wave at all speeds.

Ah..
At second glance, the aux controls on the CA V3 look like they can be used to change parameters on the fly, not select different profiles. I can't be constantly futzing around with the CA screen as i bounce between high and low speed areas.

So back to the original question.. of whether the big controller pops the small one or not when the phase leads of both are paralleled and 4kw is traveling through them.

If i can't get an answer, then i need a cheap 6fet 72v capable sine controller to experiment with, that i can afford to lose. Would a kunteng fit the bill?

Sorry I don't have an answer on the dual controller parallel question...my guess is it would probably work as long as the controller not in use was "off" (ignition/keyswitch wire without power) so the MCU and FET drivers wouldn't ahve power, so they couldn't cuase any potential shorts of the one that was active.

Theoretically, a simple two pole on-off-on switch would work, with power wired to teh center contact of both poles, and then wire the ignition from each controller to the diagonally opposing end contact of it's respective pole. All other pins of the switch would be left unused. That way you'd switch from "everything off" center position to either end, which would power on one controller, and to switch to the other you'd switch to the other end, whcih would turn everything off whiel you're on the way over. Switch doesn't need to carry much current, just enough to run the MCU / etc.

In theory you could wire the throttle, ebrakes, halls, phases, etc., of both controllers completely in parallel.

But you'd have to test it to see if it would actually work. :/

neptronix said:

At second glance, the aux controls on the CA V3 look like they can be used to change parameters on the fly, not select different profiles. I can't be constantly futzing around with the CA screen as i bounce between high and low speed areas.

Click to expand...

Sure they can be used to select different profiles (called presets in the CA setup).

I do this all the time with the "3 speed rocker switch" type from ebikes.ca (so exactly the same as using any other 3-speed switch). AFAICR it uses the AnalogAux mode to do this.

switching between three different profiles for parking lots (low) vs streets (high) vs assorted other uses like school zones, etc (middle), mostly to limit my max speed if using PAS, but the profiles can be used to change a number of different things all at teh same time.

If using the switch for presets, you can use the DigiAux mode (buttons on that thing I linked) to change any other single parameter that's in the list quoted with the link, IIRC. (but you don't have to use it, so you can use just the 3-speed switch if you like, if all you need is two or three "levels" of operation).


But I expect a single controller with a good torque (current) throttle would be even better, letting you use as much or as little power as you want, without changing modes or anything. Once I got used to the one on Incememed's SFOC5, for instance, I could easily control exactly what I wanted to do with it (within it's capabilities), and wished every controller had that available.

neptronix said:

So back to the original question.. of whether the big controller pops the small one or not when the phase leads of both are paralleled and 4kw is traveling through them.

If i can't get an answer, then i need a cheap 6fet 72v capable sine controller to experiment with, that i can afford to lose. Would a kunteng fit the bill?

Click to expand...

No reason a powered off controller can't be attached at the phase wires. Problem is if both controllers are on.