Speed settings on Infineon controller

[quuxman]

I recently ordered a 6 fet IRFB4110 Infineon controller from em3ev.com (along with USB plug), and have a question about programming it. I found this page describing the settable parameters:

https://code.google.com/p/xpd-ebike/wiki/ParameterDescription

I'm guessing the 3 speed switch controls the maximum frequency of the motor phase power, so controls the RPM. This is kind of useless to me, because I'm a lot more concerned with power consumption than speed. Is there any way to make the 3 speed toggle control the current limit?

UPDATE: On further searching, I found these two posts, but there are no conclusions. Seems to still be an open question. Sounds like this is possible with a CA (which I don't yet have for this bike), but I'd like to know how it's done with the CA, and if it's reasonably possible without one.

https://endless-sphere.com/forums/viewtopic.php?f=2&t=18076
https://endless-sphere.com/forums/viewtopic.php?f=2&t=24555

 

[amberwolf]

I don't know if any of this info is helpful, but here it is:


There aren't any actual Infineon controllers anymore AFAIK. The various companies that make controllers have generally moved on from Infineon MCUs to other brands (though I don't recall the company names at the moment, there are at least a couple of different ones I've seen on various boards).


As for what the 3-speed switch actually causes the controller to do, that probably depends on which company's MCU is used, and what firmware has actually been installed in it and what that firmware was written to do.

Back when the actual MCUs *were* Infineon (several years back, AFAIK), I think it was determined more or less that it was acting as a throttle modifier, thus basically acting to control the maximum PWM % (on-time, thus maximum voltage across the phases at full throttle).

When not altered from factory settings (or on non-programmable controllers), it was typically (I think) 80% (or 70%), 100%, and 120% of the basic default value of the setting itself, for the 3 positions of the switch. Since the center position of the switch is the same as not having a switch, the 100% setting there means it would simply operate as "normal".



Some controllers don't do anything at all when a 3 speed switch is used, regardless of switch position. Some respond as expected.


For instance, I have two XC8M06-based 12FETs, which are similar to each other but not identical. Both are on my trike, one on the left wheel and one on the right. One does respond to a 3-speed switch, but seems to be changing the current limit rather than speed. The other does not appear to cause any noticeable change in response. Also, one has two ways to switch modes (the other only has the usual one): a rocker/toggle switch input with two pads and a ground, and a single momentary SPST-switch input with one pad and a ground, which toggles thru the same 3 modes one at a time, starting at whichever one it is at before you press it.
https://endless-sphere.com/forums/viewtopic.php?f=2&t=68501
https://endless-sphere.com/forums/viewtopic.php?f=2&t=67833&p=1037371

So...I don't know how your particular controller will behave with a 3-speed switch in different modes, or what the programming of it will actually do (meaning, what the parameter actually controlled by the switch is, or how it interacts with other parameters you can change).

 

[quuxman]

Well that is helpful. Should I still expect the "Infineon" controller I ordered from em3ev.com to work with the xpd-ebike parameter setting software?

 

[amberwolf]

I don't know if it will or not. I've never tried XPD or any other programming software on my recent controllers.

You might see if searches of ES for "EM3EV" "controller*" "program*" come up with any useful info from anyone that has already tried it.

If it's one of the "EBxxx" series of controllers (as marked on the circuit board) it may work--but depending on the MCU in there and the firmware version and setup (including how and where they chose to store the bits of data representing the settings), the changes made via any of these programming softwares may or may not do anything at all (may be ignored by the firmware, or may not be valid settings for that version, with unknown results).



I wish I had more helpful info for you.

 

[DanGT86]

quuxman,

I think you might need a CA to accomplish true variable current limiting. I use a CA3 with a 3 speed power switch and a 3 speed switch on my controller for speed. During hard accelerations it appears to pull the same current on all 3 positions of the speed switch. I can't answer your question for sure since the CA is manipulating the throttle signal but it would appear that the current is unchanged between speeds. The cycle analyst is only "fooling" the controller to make it act as if its a current based throttle but I believe internally all of the "infineon" controllers are speed based.

I think the 3 speed switch directly to the controller will re-map the throttle to use its full sweep range in all 3 positions of the switch. So if for instance your speed was limited to 10mph than half throttle would be 5mph. That's a lot of hand movement spread over only 5mph so you would probably find yourself using less amps in the lower speed switch positions because the throttle wont be as jerky. So indirectly you are using less current in the lower speeds if you are reasonable with the throttle.

Hope that makes sense. Basically if you have a bike that can go 40mph and you crack the throttle open to try and only go 10mph, the lack of resolution in your throttle will create way more high current spikes since each little wiggle of throttle represents a lot of speed change.

When I get a chance I will go set my CA3 throttle in direct pass through mode and see how the current compares in the different speed switch positions. One test is worth 1000 opinions.

 

[quuxman]

A test would be great information DanGT86. I will experiment with mine as soon as it arrives, but nothing I can do will get very good information until I get a current meter in addition to the volt meter I ordered. According to amberwolf above, there are many different builds of currently labeled Infineon controllers and behavior varies even between ones with the same label on the controller logic.

 

[whatever]

variable resistor in parallel to the shunt should do the job

 

[DanGT86]

Ok, did my testing. Results are only partially conclusive.
Test was on an "infineon" Lyen 9 fet controller. EB209 board. Not sure which chip.
Battery amp max at 20
3 speed switch at 40% 70% 99%
Full throttles from a dead stop yielded the following results.

switch at 40% speed was 9.5 amps
switch at 70% speed was 21 amps
switch at 99% speed was 21 amps

I also tried full throttle up hill and dragging the brake and got the same current load results.

So it would appear at some point the speed percentage is so low that it does also limit the amp draw but there was no difference in the amp draw between 70% and 99%. There may be other ways to isolate factors for a more definitive test but this is the best I can do with out a data logger. I need to try it with a larger current setting range and possibly my other hub motor to rule out some factors. Hope that helps shed a little light on things.

 

[quuxman]

Interesting that the amperage was only different at 40%. I plan on testing out a range from 10% to 50% when mine finally shows up.

 

[DanGT86]

I thought that was strange as well. like I said there are plenty of variables to isolate before making a definative call on exactly how it works.

Perhaps there would be a difference in the current draw between 70 and 99% if the block time was set differently in the program. Maybe it wanted to limit more but couldnt respond fast enough.

I have another bike going together soon with a lyen 12fet running an EB312 board. I think there were some changes in the way the 2xx and 3xx controllers calculate current so I will do the same testing with that one. That bike has a higher speed motor so I should have more time to keep it loaded with high current.

The motor winding I was testing with is very slow so the current starts dropping off really quickly because it hits top speed right away.

It will be interesting to see how your testing works out.