Cycle Analyst V3 preview and first beta release

[John in CR]

Some of us will soon be using the CA3 on dual controller rigs, and I want to make I'm not overlooking potential issues. The two controllers run a single motor, actually on alternating teeth of the same stator, so there aren't any sinc issues with exactly a 1 tooth (15°) difference in the commutation sequence between the controllers. The supply is from the same battery with main power wires together to close to the controller to avoid voltage differentials trying to flow through the common throttle, and we've had no issues running a single throttle wired to both controllers. We have the Ebikes.ca high current stand alone shunts, so the CA3 will be able to sense the current fed to both controllers.

My only question is about the yellow wire on the CA connector the Sp wire. We can only wire that to 1 controller. As long as it only senses the rpm based on the hall pulses and pole count, then we're fine. If it is used somehow to sense and regulate current, especially in the torque throttle mode, then it may be problematic, since the other controller "sees" a next magnet and get a hall pulse in between hall pulses of the 1st controller. Is this a potential problem, or does that Sp wire only sense rpm?

Anyone see any gotchas in this plan?

Also, I didn't see mention of tying an inclinometer to the CA3. I know quite a few people interested in that for anti-flip protection along with power wheelie control. I just wanted to post a reminder about that, since it's been several months since Justin and I discussed it.

John

 

[izeman]

i have another concern: i use regen braking with a directly connected brake switch to the controller brake pads. so i'm not using the brake's connector of the ca, which can imho not be used for regen, but for cutting the throttle only. correct?
if i don't release the throttle and pull the brake, the throttle is cut and the wheel is in generator mode. one could "forget" to release the throttle completely when braking, and as soon as the brake is released, full throttle is applied from 0-100% in 0 seconds.
how can this be avoided? i'm talking about the most obvious thing as releasing the throttle before applying the brake. i'm talking about some safety feature: eg. one has to release the throttle completely after brake is pulled, before throttle is engaged again. how could that be acchieved?

 

[teklektik]

The Unofficial User Guide for B22(final) is posted here.

Firmware B22 Prelim3 became B22 so there are actually no feature changes to the previous Prelim3 Guide. This version upgrades images, expands/clarifies existing content, and adds details for using LM35 temperature sensors (cell_man MACs). The sections on throttle ramping and power oscillations have been reorganized and new material on tuning the Speed Throttle has been added. The Speed Throttle stuff is incomplete (continuing bad weather for testing) but hopefully it will be finished up in the B23 UUG version.

Enjoy!

 

[Merlin]

Great Job Tekle....

...i have a new solution for my "speed"/PAS Problem....
i never tried cos i was still setting up my CA.....

today i tried it..............THE CRUISE CONTROL

now i can go on all Presets to every Speed i want. Just hold throttle 2 seconds on maybe 15kph and i can cruise minimum 15kph.
everything i pedal on top is saving accu-power and my girlfriend dont kill me when i do every 100 meters acceleration runs :lol:

I hope that was not a trick...i hope thats a feature that the cruise control runs all time including me is pedal. It is not the End of my Test(will try a 3 mode switch on AUX), but iam really glad that the CC works in this way....SO PLEASE DONT CHANGE IT THAT STARTING PEDALING ENDS THE CRUISE CONTROL

 

[hillzofvalp]

John, the current throttle function takes the shunt reading only. If I'm not mistaken the yellow speed wire does not play a role. When my sp wire was disconnected I still had current throttle function. I think all will be fine but the best experience will be with external shunt. Likewise, I think that it will be best if there is a middle man board between the CA throttle output and the controllers so that we can fine tune the thresholds for each controller. SPeaking of which, have you tried measuring the throttle voltage when the controller turns "on" and also when it switches to "full pwm"? btween the same controllers I feel like the difference will be negligible. The important part is making sure the shunts are the same. But even still once you figure that one controller is ramping up higher than the other, you can tune the "warm" controller a fraction lower to balance out the power sharing.

 

[pendragon8000]

Interesting point there... I was thinking it might be nice to have speed influence the algorithm for throttle output so power isn't wasted off the line. Just an idea.

 

[John in CR]

I don't think that would mess me up. To throw a monkey wrench in my plan is if the CA was matching current to each commutation change, because it will only see half of them.

HOV,
Controller matching down to the amp isn't a requirement. My controllers are 20A different, and that's battery side so 40A or more difference in peak phase current. The only time I am able to notice it, is occasionally during a launch where I can feel a pulse or two or three of different force for a portion of the first rotation. I believe it's only the low pole count that enables me to feel that. At higher rpm the BEMF should force the 2 controllers to identical output, despite the difference in current limit.

 

[justin_le]

The Unofficial User Guide for B22(final) is posted here.
Firmware B22 Prelim3 became B22 so there are actually no feature changes to the previous Prelim3 Guide. This version upgrades images, expands/clarifies existing content, and adds details for using LM35 temperature sensors (cell_man MACs). The sections on throttle ramping and power oscillations have been reorganized and new material on tuning the Speed Throttle has been added. The Speed Throttle stuff is incomplete (continuing bad weather for testing) but hopefully it will be finished up in the B23 UUG version.
Enjoy!

Hey Teklektik, once again an over-the-top contribution here that is of massive help to all those needing some kind of reference in setting up the CA-V3! It's a huge job to put together a clear and detailed set of documentation like this.

To everyone else, Teklektik has no financial stake in this but I'd like to thank him by offering a CA3, a Thun torque sensor, and a whole bunch of ebikes.ca store credit for whatever EV project parts he needs in the future too, and just want to mention that here in the interest of transparency. If anything it's just a small token for the unsolicited help he's given in making this rollout happen.

I'm just going to do some very minor tweaks to the B22 code and then it will 'go live' as the Cycle Analyst Version 3.0. This thread started on Mar 23rd, 2013, and we'll remove the beta after exactly 1 year, on Mar 23rd 2013.

-Justin

 

[teklektik]

...I'd like to thank...

Wow! Very much appreciated!

Actually this has just been a way to flog you into coding cool new features instead of writing documentation!
My 2WD build was designed around the CA V2 from the outset and ran into some shortfalls - the V3 came along last year at the perfect time and has really made the build work as originally envisioned. Thanks from all of us for a great product!

BTW - I've often posted recommendations here about emailing Grin Tech for pressing or complicated problems. Although much of the Guide is a compilation of information posted in this thread, many of the technical details came from just such email communications with Justin. Outstanding product support...

 

[teklektik]

I was building a cable today and noticed that the connector on the CA-DP cable in the Guide was drawn with the wrong sex (oops!). The CA-TRQ/PAS connector also had the pin numbering inverted in both illustrations (happened when the connector images were flipped to face left).

Anyhow - fixed the Guide and the small Connector/PCB 'shop PDF' as well. Reposted.

 

[Architectonic]

Yeah, big thanks to teklektik. The v3 can be quite daunting to set up initially (due to the amount of options).

 

[Kepler]

Big thanks to teklektik from me also.

And of course a huge thanks to Justin for bringing the CA3 out. I love this device and have them on both my ebikes now. 8)

 

[2handy]

My first post;- I had to finally register (after lurking over this thread for some time), as I finally had something to say.

I've been following this thread for some time; read the whole thing & check on it everyday.

At this point, just prior to the birth of the non-beta CA3, I'd like to add my sincere thanks to Justin & his team for the creation of this amazing piece of technology & continual expert refinement that I've seen happening to the CA3 over the course of this entire thread.

& to teklektik for the enormous amount of good natured & expert support that you have offered consistently, virtually since the beginning of this loooong thread nearly a year ago. All of the forum readers have benefited from your input teklektik, & your amazing unofficial manual would seem to be the current CA3 bible, & all of this you have volunteered! Wow!

Take a bow man?

I think that Justin has honoured you teklektik in a well deserved fashion, you have made his life so much easier & more productive re. the CA3 (at least). 8) This has benefited everyone concerned with the project on the inside & the outside of its development.


So who am I & what am I up to?

I'm just embarking on the ebike world, having recently received my simple (compared to some of the amazing bikes that lots of you guys here have!) 1kw rear wheel mp3 hub kit & a 48v 10a battery. I'll get it sorted over the coming week or so, before I too look at becoming a proud CA3 owner.

Can't wait to get the old 2006 GT DHi race, downhiller electrified! 8)
I'll post a picture when I do.

 

[mrbill]

I, too, would like to add my thanks to Justin and "teklektik" for their work on the CAV3. I've been enjoying using and testing the different options over the last several weeks. I think I may be warming up to the power throttle after reducing the wattage limit and reducing the WGain parameter. This is an incredibly useful and flexible device for electric vehicle home-builders, making a host of features available even to users of relatively "dumb" motor controllers.

But, my posting would not be complete without including a small complaint.

Today I rode for the first time since I updated to Beta23 on a ride longer than 99.9 miles. What happened to the tenths in the display when trip distance units roll past 99.9? I am fairly certain that tenths were displayed for mileages over 99.9 in Beta22, or I would have noticed their absence earlier. Are we already hard-up for memory that only 3 significant digits can be displayed for mileage (or "kilometerage")? I hope not.

 

[Kepler]

I have now had a chance to try quite a few combinations of settings and get a good feel for the CA3.

A couple minor items that may require addressing.

LiPo battery charge level indication on the battery gauge I think still needs some tweaking. I am finding that it still shows 1/3 full when well and truly empty. As a stop gap, I have entered 19 cells rather then 18 cells. This gives a better indication of remaining battery with just one pixel left on the gauge when I hit my 62V cutout. with an 18S LiPo pack, 62V to 63V is the correct cutout point with the packs typically sitting at 5% to 10% remaining charge as measured with a Hyperion EOS Sentry battery checker (a must have tool if running LiPo).



This gives a no load V of approx 3.45V per cell. My observations with running LiPo exclusively on ebikes for 4 years now is that 3.45V is the sweet spot as far as minimum voltage per cell is concerned.

Perhaps there is a setting I have missed in relation battery setup but I think either the minimum battery gauge voltage for LiPo needs to be adjusted or perhaps a settable minimum voltage for the battery gauge be added to the software. The stop gap of entering an extra cell is ok for larger cell counts but it will throw things way off on lower cell count LiPo setups.

The other thing that I am having trouble with is setting a PAS option that allows you to use the throttle to 6kph but after that, only a PAS input will power the bike (throttle inactive after 6kph). My understanding is that this is the allowable config for AUST (together with 250W and 25kph max). Its not a major issue but I would really like to have one of my presets completely AUST compliant.

I know there is an option that allows the throttle to work only if you are pedaling. This I have tested and works fine however I cant see how to setup "throttle inactive after 6kph".

 

[Diamondback]

I have now had a chance to try quite a few combinations of settings and get a good feel for the CA3.

A couple minor items that may require addressing.

LiPo battery charge level indication on the battery gauge I think still needs some tweaking. I am finding that it still shows 1/3 full when well and truly empty. As a stop gap, I have entered 19 cells rather then 18 cells. This gives a better indication of remaining battery with just one pixel left on the gauge when I hit my 62V cutout. with an 18S LiPo pack, 62V to 63V is the correct cutout point with the packs typically sitting at 5% to 10% remaining charge as measured with a Hyperion EOS Sentry battery checker (a must have tool if running LiPo).



This gives a no load V of approx 3.45V per cell. My observations with running LiPo exclusively on ebikes for 4 years now is that 3.45V is the sweet spot as far as minimum voltage per cell is concerned.

Perhaps there is a setting I have missed in relation battery setup but I think either the minimum battery gauge voltage for LiPo needs to be adjusted or perhaps a settable minimum voltage for the battery gauge be added to the software. The stop gap of entering an extra cell is ok for larger cell counts but it will throw things way off on lower cell count LiPo setups.

The other thing that I am having trouble with is setting a PAS option that allows you to use the throttle to 6kph but after that, only a PAS input will power the bike (throttle inactive after 6kph). My understanding is that this is the allowable config for AUST (together with 250W and 25kph max). Its not a major issue but I would really like to have one of my presets completely AUST compliant.

I know there is an option that allows the throttle to work only if you are pedaling. This I have tested and works fine however I cant see how to setup "throttle inactive after 6kph".

a few posts back, Justin said he could implement a system where the end user could decide the cut off for lipo packs.
as it currently stands, he is using 2.9V per cell, and this is why the remaining change is highly over estimated.

as for the aust law, i believe throttle is allowed beyond 6kph, but you must be pedaling.
if you are not pedaling, throttle is only allowed to 6kph.

i use 3.5v per cell as my safety, with a 42v LVC (for 12s)

Jason

 

[Merlin]

a few posts back, Justin said he could implement a system where the end user could decide the cut off for lipo packs.
as it currently stands, he is using 2.9V per cell, and this is why the remaining change is highly over estimated.

as for the aust law, i believe throttle is allowed beyond 6kph, but you must be pedaling.
if you are not pedaling, throttle is only allowed to 6kph.

i use 3.5v per cell as my safety, with a 42v LVC (for 12s)

Jason

i use 3,5v/cell too. i made a lot of tests with turnigy lipos and the start to drift around 3,6v discharge.
i also charge only 4,17v.
after 3,6v lipos looses hsi power too. so no one need to discharge below 3,5v

i have to set another cell count too. on full discharge it shows 2/5 "fuel"

the Throttle Problem exists for German guys.
with RPM | Thro i can give throttle till 25kph
with RPM & Thro i have to give throttle to get PAS Assist.

setting throttle Mode from pass thru to "Speed" will work for max throttle speed (6kph) but the bike only stops throttle (once) short. second later KPH is blinking and after that i can give throttle till max speed (25kph) (kph still blinking)

Both are not legal in Germany.

 

[Diamondback]

im using the TRQ PAS mode on my CA V3, haven't updated to the latest FW yet.

i too would like to see an option for a legal Australian (and im guessing by definition EU legal too) PAS mode.

where i can throttle (with no pedal) up to 6kph and then if im pedaling, i can have 25kph with throttle,
or i can continue with no throttle and use the TRQ mode only.

so basically, i want Trq | Thro but have it work as above.

in the above example, if you were doing 25km/h pedaling with the throttle open, then you stop pedaling, the motor should stop until
speed is reduced to 6kmh, but i shouldn't have to open the throttle for the PAS to work either.....

Jason.

 

[Kepler]

a few posts back, Justin said he could implement a system where the end user could decide the cut off for lipo packs.
as it currently stands, he is using 2.9V per cell, and this is why the remaining change is highly over estimated.\\

Jason

Yes I did see that but with the non beta release imminent, I was hoping this had in fact been adressed in release 23.

as for the aust law, i believe throttle is allowed beyond 6kph, but you must be pedaling.
if you are not pedaling, throttle is only allowed to 6kph.

I need to read this once more then as the way I understood it was that the throttle was there to get you off the line without pedaling then the PAS works like normal once you got past 6kph with the throttle no longer operational. I may have got this wrong though. wouldn't be the first time

i use 3.5v per cell as my safety, with a 42v LVC (for 12s)

Jason

I did use 3.5V (63V on 18S) I found I was getting 10% to 12% measured remaining. I push my packs a little harder now with the 3.45V setting. I think the higher voltage and corresponding lower amps tends to put less strain on the packs hence the slightly lower setting.

 

[d8veh]

Just got my V3 with Torque sensor. It looks impressive.
First question: there's a short wire coming out of it with a black blob thing on the end. What's that? It's not shown in the pdf manual.

Next. i have a 500w geared motor, is there anywhere a description of ideal settings to use it with the torque sensor, or have I got to go through the manual and do it by trial and error?

 

[izeman]

I have to confirm that too. I have 24s lifepo4 and it shows FULL @85v. But it shows EMPTY @76v which is 3.2v per cell and surly not empty.
If I could add a feature request as well I'd like to see a configurable low and high voltage. Adjusting the discharge curve and translating this into a gauge depending on empty and full village values tough is very difficult.

 

[Merlin]

hmm today i take a 30 miles ride.
the caV3 tell me that i used ~20WH/km + 870WH + 12,7Ah capacity used.

my powerlab puts 41Ah in 9x6S5000 Lipos. Thats each ~4500mAh ...on 75volts it means 13,5Ah used.

Is there a way to "calibrate" the CA that i can clear and correct see what i get from my accus?!

 

[justin_le]

i use 3,5v/cell too. i made a lot of tests with turnigy lipos and the start to drift around 3,6v discharge.
i also charge only 4,17v.
after 3,6v lipos looses hsi power too. so no one need to discharge below 3,5v

Sweet, thank you for that discharge graph, it's exactly what I needed. It's quite interesting how much more there is to a battery chemistry than the nominal battery chemistry, since this curve from the Turnigy looks absolutely nothing like the LiPo discharge profile of most ebike packs. Anyways, I have updated the setup so that there are two LiPo options now, one with a ~3.2V end of charge matching the ebike packs, and a 2nd that terminates at about 3.5V based on the graph above. The only additional thing that could be useful is if you could estimate by how much the voltage in that curve had dropped from internal resistance? I'm guessing not very much, like 50mV?

-Justin

 

[justin_le]

the caV3 tell me that i used ~20WH/km + 870WH + 12,7Ah capacity used.
my powerlab puts 41Ah in 9x6S5000 Lipos. Thats each ~4500mAh ...on 75volts it means 13,5Ah used.
Is there a way to "calibrate" the CA that i can clear and correct see what i get from my accus?!

Yes, that is what RShunt is for. However, between 12.7Ah and 13.5Ah you may be getting close to the accuracy limit of your measuring devices so you should make sure you are tweaking it to a known good reference. A lot of amp-meters are only in the 3-5% accuracy range.

 

[Golden Motor Canada]

Just got my V3 with Torque sensor. It looks impressive.
First question: there's a short wire coming out of it with a black blob thing on the end. What's that? It's not shown in the pdf manual.

Next. i have a 500w geared motor, is there anywhere a description of ideal settings to use it with the torque sensor, or have I got to go through the manual and do it by trial and error?

The green wire is for the throttle output when you are not using the direct plugin on a controller. I'm not 100 percent sure but I think the black is a ground and the yellow goes inside the CA enclosure but is not attached to anything. You can use it for a temp sensor or whatever you like. I only used the green. I made a diagram to help people with magic pie controllers. Maybe this will help you picture what it does.

The settings really just depend on your bike and what you want from it so it is really just trial and error.

Gary