Cycle Analyst V3 preview and first beta release

[amberwolf]

I have a lame question (that won't help you out any) but can you describe how many e-brakes/regen systems you have active? I don't quite understand your statement about

when I brake with ebrake levers + throttle

The CA's original (pre3.2x) braking modes for variable braking work by reading your ebrake lever to turn on braking mode, which then outputs a different throttle voltage range (set in it's own field vs teh one for normal thorttle output) as the rider then manipulates the thorttle to control braking amount. This only works for controllers such as Phaserunner (and other *runners) and some Grinfineons, which accept a separate voltage range outside the normal throttle range for braking control--typically this is the range below 0.8v where a hall throttle normally never goes.

Since for me (and probably most typical bicycle riders) it's not a natural instinctive/trained reaction (necessary in emergency situations) to use a throttle to control braking amount while *also* pulling your brake lever only just enough to trigger the braking but *not* use the lever to *modulate* braking, I came up with this:
https://endless-sphere.com/forums/viewtopic.php?f=2&t=105460
to switch to a cable-pulled throttle on that specific brake lever so the lever itself controls the amount of braking exactly like a regular brake lever would.


I don't yet know enough about the new braking modes available in 3.2x to describe any of them or their limitations yet.

 

[amberwolf]

As I note above, I don't yet know enough about the new modes to give suggestions, but could you post your specific complete CA settings? (screenshots of the program, or attaching the saved file from it, etc) At least let us know which specific mode of regen you've chosen and the setttings in that section would probably be helpful.

I posted a question about the CA 3.2b firmware in its own thread, but as it went mostly unnoticed I thought I'd try here where the discussion about this firmware seems to happen.

So my question: when I brake with ebrake levers + throttle, I can brake to a full stop without using the mechanical brakes (more precisely: I can brake down to 2kph, which is set as the regen brake speed minimum in my baserunner). But if I brake with backpedal regen, it suddenly stops braking around 6kph and I cannot reach a full stop - I have to use ebrakes or good ol' brakes.

I went over all the parameters in the CA setup application (and the phaserunner too, just to be sure), but I didn't see anything that looked like "packpedal regen min speed".

Did I miss something? is it a limitation for some reason? a bug?

 

[SRB22]

I have a lame question (that won't help you out any) but can you describe how many e-brakes/regen systems you have active? I don't quite understand your statement about

when I brake with ebrake levers + throttle

If you want to know everything about regen, you should have a look at the great video Justin made about it: https://ebikes.ca/blog/post/next-level-regen.html

Then everything will be clear!

 

[SRB22]

As I note above, I don't yet know enough about the new modes to give suggestions, but could you post your specific complete CA settings? (screenshots of the program, or attaching the saved file from it, etc) At least let us know which specific mode of regen you've chosen and the setttings in that section would probably be helpful.

Attached is my settings file. This is for a bike with a 52V battery, baserunner, and Grin All-Axle hub motor (and a CA, obviously!). Tell me if you need to know more, I'm not sure what information is relevant in that case...

 

[amberwolf]

Ok, based on the tooltip hints the setup program gives me for the parameters you've got set, some first thoughts. I haven't learned all the interactions in the new stuff yet, so I'll post more if I think of anything.

Backpedal regen is set to Position, and will take 6 steps (6 pole positions, meaning 6 magnets passing the sensor) before regen will reach it's max amount after it starts to engage, so you have that much control over it. It will take passing 4 magnets before it will even start this, per the engage steps setting.

Your PAS device poles says it only has 8 magnets in it, so it will take more than a full reverse revolution of the cranks (almost 1.5) to reach full regen strength.


Probably not relevant, but:

It says it's a "BB TRQ" sensor, which not being one of hte preset brands doesn't say for sure if it is capable of actually detecting pedal direction (it says it's 2-wire, so it should be, if it's setup correctly for your sensor). Also, 8 poles is very low resolution for a cadence sensor (very coarse response, slow response time, etc).

Note taht if for some reason the torque sensor you have doesn't support pedal direction detection, you can buy a separate crank-mounted PAS sensor to wire in place of the cadence portion of the torque sensor, that does do pedal direction detection and has higher resolution (faster finer response).

 

[SRB22]

Thanks amberwolf for your thoughts.

Backpedal regen definitely works for me, so I can be sure my hardware supports it. I still have to fiddle with the number of steps, finding the sweet spot between "reacts fast" and "can be finely controlled". But it already works quite well and I like it very much.

Still, this doesn't answer my original question: why does backpedal regen stops braking when reaching 6kph, when ebrake regen is able to brake down to a full stop...

As @dunlop suggested, this might be a question only @justin_le can answer... Justin, are you around?

 

[amberwolf]

I've been poking around since my first reply, and haven't yet found anything to tell me why there is a different lower speed limit for backpedal regen vs brakelever/throttle regen.

Looking htru the documentation for the 3.2b3 html file (attached) i found two references to a setting MaxRegenSpeed that does not actually exist anywhere else within the documentation, so I don't know how you can change this setting...but it might be the cause of the problem if it means something different than it sounds. It is associated with another term *also* only found in the same lines with it, RgnSpd.

But skimming the files, I found no references to anything that seems like it could limit the backpedal regen speed, or a note about a known bug for it.

 

[SRB22]

Thanks for the files, I'll also have a look at them

Looking htru the documentation for the 3.2b3 html file (attached) i found two references to a setting MaxRegenSpeed that does not actually exist anywhere else within the documentation, so I don't know how you can change this setting...but it might be the cause of the problem if it means something different than it sounds. It is associated with another term *also* only found in the same lines with it, RgnSpd.

3.2bX distinguishes between MaxSpeed and MaxRegenSpeed, that is between the max speed at which you can assistance and the speed where regen engages. That's one of the great new features of this version. I'm quite sure that's what is referred to in the documentation.

So there is a difference between these *max* speeds. But I couldn't find any information about the *min* speeds to be different (and cannot really imagine a use case where they should be different). So the mystery remains...

 

[Dunlop]

by SRB22 » Nov 21 2022 3:07am

If you want to know everything about regen, you should have a look at the great video Justin made about it: https://ebikes.ca/blog/post/next-level-regen.html

Then everything will be clear!

That is definitely a great video!! I had started to watch that some time ago, and never returned to watch the end....
I like the set-up you are running, and will consider the ebrake/throttle method as well when I get to set mine up... BUT, what really interests me is the back-pedal method you are having problems with... That is what I think I want to use as my primary method of Regen/braking.
Have you made any progress with your issue?? I really hope you will come back and share, when you conquer it!!
Keith

 

[SRB22]

I like the set-up you are running, and will consider the ebrake/throttle method as well when I get to set mine up... BUT, what really interests me is the back-pedal method you are having problems with... That is what I think I want to use as my primary method of Regen/braking.

I use three regen braking modes: ebrake/throttle, backpedal and RegenSpeed via an analog aux pot. All three work great, I recommend them all. My bike is also set up to use "negative power assist levels" via digital aux, but honestly I never use it.

One interesting side effect: almost 7'000km on this bike, and still using the original brake pads!

I would recommend that you setup all methods of regen braking and see what works best for you in what situation. For instance, I tend to use backpedal much less in busy city traffic, because it is slightly less reactive, as you have to "unbrake" by pedaling forward before you can power your bike again. But when going downhill for many kilometers, as is quite common where I live, backpedal regen is a real blessing!

Have you made any progress with your issue?? I really hope you will come back and share, when you conquer it!!

Not yet. I emailed Grin but did not get any answer yet. Of course I will post a reply here if I get any more information on the subject!

And don't let this slight problem stop you: backpedal regen is still very useful even if it does not allow to fully stop (you'll need mechanical brakes anyway!)

 

[SRB22]

Sorry I got an answer from Justin quite a while ago, but first I was busy and then I forgot to post it here. Here is what he said:

This is the intended and programmed behavior. When I had it at a lower cutout speed (like 2 or 3 kph) then if you came to a quick stop and then started to pedal again the regen would still stay active as there is a bit of a latency for the CA speed reading to updpate, and it was found by experimentation that 6 kph was about the sweet spot to still get you almost to a full stop while at the same time always allowing you to pedal and accelerate if you come to a quick stop and then start again right away.

Unfortunately there is no spare eeprom room left to make this be a user programmable setting as I'd otherwise wish. But with some better logic looking at the deceleration value I can have a more effective predictive algorithm to get closer to zero, and I may tweak with it a bit in CA3.2b4 release.

 

[miro13car]

since it is very comprehensive thread about CA V3.
I hope I post in right thread.
about troubleshooting
after some wet snow riding my CA V3 displays RESET mode and switches are dead.
after disconnecting screen with 2 switches CA powers up normally , it works normally with no display.
it powers up and show WATTS screen just normal.
I connected display from my spare CA V4 to this bad CA and all works OK
I checked display connctor thoruoghly, just 2 tactile switches , ribbon cable and connector - what could be simpler that this!
no shorts on connector, continuity from ribbon to on board connector is here, no shorts beween pins.
those 2 displays seems identical.
HELP please, what could be wrong?

 

[miro13car]

it would be the shame to throw away this
CA

 

[amberwolf]

Is probably moisture intrusion into the membrane switches. AFAIK Grin still carries the front panel with the switches; at one time some years back there was a known problem with them so they were sending out replacements as problems were found and reported.

 

[miro13car]

thank you Amberwolf for answer
maybe my mulitimer ohmmeter cannot measure this but must be some resistance of Megaohms on this tactile switches.
simply there are not really opened
and there is reason that they sell those front panels with switches.

 

[Albeegood]

Ca3-wp on a sabvoton 72150. Is it possible

 

[Albeegood]

Sabvaton
Is it possible to connect the CA3-WP to my sabvaton 72150 Bluetooth? Thinking on replacing my current UCK1 display with a CA3-wp.. i have from a previous build.
From what i read here
https://ebikes.ca/documents/CA3-WP_Plug_Guide.pdf
It looks possible. These are my plans.
Get the W8P female plug, connect V+, Gnd, Spd, Th0 to my controller and reroute the Thi (see photo) and instead of using the Stand Alone Shunt im planning on using the Inline Shunt. Now assuming that’s possible i have a concern. My battery cables are parallel, i have 2 on each side of battery. Will i have to put one in each Negative?

[/img][/img]

 

[miro13car]

my CA V3 display is repaired.
I got down to tactile switches under all glued adhesive tapes and what?
one of the traces to switch was worn out and bearly contacted .

NO ANY water damage, no humidity.

how could it happened??
membrane ot tactile switch cut thought the trace on the board - pretty bad quality
impossible to repair?
possible, possible
I replaced original switches with water proof real tactile switches - cut openings for switches, epoxy them, connected 3 wires to black connector to curcuit board.
check picture

 

[Ingo]

Hi everybody,

since version 3-2b3 i have some problems with the PID control to my CA.
with the previous version 3-15b3 there were no problems with my settings.
I attached the files.

With the previous version and these settings (dsgain=0 , PSgain = 4V/sec , Isgain = 150) the speed limit worked great. but from version 3-2b3 something must have changed completely I myself have not changed my settings. my components have also remained the same : 100kg ebike, qs250 motor, kelly KLS7275D controller and 100ah 12s lifepo pack.

i dont use prop regen because the controller have not this feature.
if i step on the gas quickly, the km/h display starts to flash and the CA regulates the gas output back. If I give it gas slowly, it works without hesitation. That's very strange I haven't changed the files.
also the speedlimiting is very bad

it fluctuates +- 4km/h is never constant. and the regulation is very slow. Hope someone can help me! Greetings Ingo

 

[amberwolf]

First, are you just uploading the existing setttings files to the CA, or did you create new settings files for the new FW version with the existing settings?

Sometimes a different firmware version may hold the values differently, and writing the old FW version settings file may not put the right values in the right place.

Also, sometimes a new FW update may require resetting the CA completely after the update, and then setting every setting to what you need it to be.

If it isn't any of those things, I would guess it is a bug in the new firmware, and you may want to ask Grin about it. I recall at least one recent post about a speed issue with a recent FW version, and having to revert to old FW to fix it until Grin comes up with a bug fix, but I don't remember what FW they went to and from.

 

[Ingo]

thanks for the answer. usually justin always writes that when you have to create a completely new setup file. this time it went like this. I will try to reset the app completely and create a new setup file. Yes, I have already written to grin but they have no solution for this.

 

[Ingo]

now i have create a new setting + resetting the CA to standart...
still no change. the speed limiting is very poor, overshoot, hesitate. not good

someone have an idear ? grin ?

 

[amberwolf]

If it's still not working correctly, and the speed sensor is providing smooth reliable data (you don't get any jitter or spikes of wierd readings in the speed display while riding at various speeds), and the system works normally if you disable the speed limiting, then there is probably a bug in the firmware, and you'd have to either revert to the last previous version that doesn't do this, or wait for a new version from Grin that fixes it.

(and I"d recommend submitting a bug report about it to Grin; there used to be a page for that on their site but I don't see it now, so you'd have to email it to them. For a good bug report you would give the system specifics as a list of items, describing any nonstandard interconnections or diagramming it, then provide a settings file for the CA along with the firmware version that you know it will happen in, and a recipe of specific steps under conditions that you know will cause it to happen, so they can reproduce it--if they can't reproduce it, they will have a harder time fixing it. If you can also specify the last previous fimrware that *doesn't* have the problem under those same conditions, they can code-compare and find it even easier.)

 

[Ingo]

I wrote grin a long time ago. justin gave me a short answer and then he handed it over to an employee. this employee said it was my controller... my controller has been running trouble-free for several years and I haven't changed anything.

 

[jk1]

Can the cav3 put in a maintenance alarm so after every certain amount of km certain maintenance alarms come up ?