Cycle Analyst V3 preview and first beta release

Get real world experience and user feedback on the electric bicycle products.
chilltrout   10 W

10 W
Posts: 69
Joined: Oct 30 2016 8:08am

Re: Cycle Analyst V3 preview and first beta release

Post by chilltrout » May 07 2018 4:51am

Ok so it turns out that it was the gnd cable in the CA plug on the controller side!!!!
It had blown in half! wasn't even a trace!
But now the controller is doing strange things and I think one of the groups of fets are dead :(

I forgot to mention that when the CA died it went wide open throttle and ended up running into a wall and the motor locked up, it still spun this time when I got the CA working again but probably spun for 30 sec and then controller failure. (wheel cogging etc, only on 1 phase in 1 direction)

User avatar
amberwolf   100 GW

100 GW
Posts: 25406
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Cycle Analyst V3 preview and first beta release

Post by amberwolf » May 07 2018 8:28am

fwiw, if ground opens up on a hall throttle connection, it tends to create a WOT condition. If the ground that failed was the only ground between CA and cotnroller, and throttle went to CA first and then to controller, then when the ground blew it couldve caused the WOT runaway.

As for the controller, if the motor was unable to turn long enough and the controller failed to limit current or shutdown from detecting no wheel spin, it could definitely have blown FETs. You can test them via the info on ebikes.ca learn tab, troubleshooting pages.

kevinscargobike   100 mW

100 mW
Posts: 48
Joined: Nov 01 2017 6:47pm

Re: Cycle Analyst V3 preview and first beta release

Post by kevinscargobike » May 10 2018 4:49pm

I finally got my Sempu installed on my daily commuter, and I noticed some odd PAS behavior, and so I looked through the manual and saw:

"In all modes, if throttle is applied even a small amount while pedaling, PAS assist is ignored and the throttle alone determines the output."

It would be nice to have the option for the maximum of the two to always be chosen. Having the PAS assist drop out whenever the throttle is returning to it's default/0 position feels pretty clunky - like the bike is dropping and surging for no good reason. My wife's exact quote when giving the bike a try was "Bionx this is not."

-- since nobody has replied, a quick edit to specify the use case: If I'm crossing an intersection and want to scoot across a little faster, I might hit the throttle while pedaling. I continue pedaling after crossing the light, release the throttle, and I'd like the assist to remain proportional to my pedal effort, and not drop below and then jump back up as I release the throttle.

User avatar
teklektik   10 GW

10 GW
Posts: 4020
Joined: Mar 26 2011 1:15pm
Location: CT, USA

Re: Cycle Analyst V3 preview and first beta release

Post by teklektik » May 10 2018 8:10pm

Having throttle deactivate PAS is very handy when navigating peds, handling intersections, etc where precise control is necessary and an extra push on the pedals might otherwise cause an unwanted power application in tight or dangerous quarters. The alternative would be to force the user to manually deactivate PAS in these situations which would be an invitation to mishap.

The problem you are having is due to the PAS system relying on the Power PID controller which gets suppressed when you go to PassThru throttle mode. This makes it necessary for the power PID controller to wind up again when you release the throttle. If you have an aggressive DownRate, then the applied power due to throttle PassThru mode may drop before the PAS Power PID can wind up to reapply PAS power. You can go after this a couple of ways:
  1. Reduce DownRate so that power is removed more slowly. This will give the PAS system a bit more time to come up to speed.
  2. Switch to using Power Throttle. This will make the PAS and throttle use the same PID controller and the power will transition smoothly when switching between the two. Throttle response will also be much more linear and controllable than with PassThru.
Presumably you have WGain set pretty low anyway so PAS power is applied smoothly without a lot of jerking around. PasRate can help take the edge off power application, but general smoothness is strongly affected by WGain which controls how fast the PID controller tracks changes in pedal effort.

Anyhow a couple of things to try...
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

kevinscargobike   100 mW

100 mW
Posts: 48
Joined: Nov 01 2017 6:47pm

Re: Cycle Analyst V3 preview and first beta release

Post by kevinscargobike » May 11 2018 1:17pm

Thanks! Those settings did help quite a bit. I moved my PAS up rate from 4 v/s to 2 v/s, and the down rate from 4 v/s to 2 - but to be honest, I find it pretty unnerving when the motor pushes for more than a second after I've stopped asking it to. There's probably some combination of a middle ground value and just time getting used to it.
Having throttle deactivate PAS is very handy when navigating peds, handling intersections, etc where precise control is necessary and an extra push on the pedals might otherwise cause an unwanted power application in tight or dangerous quarters. The alternative would be to force the user to manually deactivate PAS in these situations which would be an invitation to mishap.
If in those same tight quarters you stop activating the throttle and bump the pedals, isn't it just as dangerous? It seems like there's several solutions available to mitigate that already in the PAS settings. Especially on a bike with a torque sensor.

Do you think anyone who is producing a steady Hwt would ever have the expectation that touching the throttle would reduce the power output?

User avatar
teklektik   10 GW

10 GW
Posts: 4020
Joined: Mar 26 2011 1:15pm
Location: CT, USA

Re: Cycle Analyst V3 preview and first beta release

Post by teklektik » May 11 2018 2:26pm

kevinscargobike wrote:
May 11 2018 1:17pm
Thanks! Those settings did help quite a bit. I moved my PAS up rate from 4 v/s to 2 v/s, and the down rate from 4 v/s to 2 - but to be honest, I find it pretty unnerving when the motor pushes for more than a second after I've stopped asking it to. There's probably some combination of a middle ground value and just time getting used to it.
2v/s is very slow for a DownRate. Although you may have achieved one goal, I think you need to explore other setup strategies that don't so markedly interfere with the stop-pedal delay.
kevinscargobike wrote:
May 11 2018 1:17pm
If in those same tight quarters you stop activating the throttle and bump the pedals, isn't it just as dangerous?
PAS will not engage unless you actually pedal - 'bumping' the pedals does nothing.
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

kevinscargobike   100 mW

100 mW
Posts: 48
Joined: Nov 01 2017 6:47pm

Re: Cycle Analyst V3 preview and first beta release

Post by kevinscargobike » May 11 2018 3:29pm

2v/s is very slow for a DownRate. Although you may have achieved one goal, I think you need to explore other setup strategies that don't so markedly interfere with the stop-pedal delay.
That's helpful! I changed that and the throttle mode at the same time. I'll try going back to 4 v/s down and seeing if the throttle mode alone does the trick. It's less clear to me what I'd do with wgain - there's no weird oscillating if I'm all throttle or all PAS, so it seems like it's probably fine.
PAS will not engage unless you actually pedal - 'bumping' the pedals does nothing.
My point was that if your protection from accidental PAS acceleration is that you have the throttle activated, that safety goes away as soon as you let to of the throttle, which may well still be in a crowded situation.

I don't think that's really an argument for why it shouldn't be an option that can be chosen, but I'd totally buy it if you told me that it's baked into the code and not particularly practical to make it optional.

kevinscargobike   100 mW

100 mW
Posts: 48
Joined: Nov 01 2017 6:47pm

Re: Cycle Analyst V3 preview and first beta release

Post by kevinscargobike » May 15 2018 4:23pm

I ended up tweaking my PLL settings for the Phaserunner, as well as going to a very low wgain, and it's definitely going in the right direction. I'm still trying to figure out the right size dead band to have on the aux pot so that I can set it to not assist at all, and the other morning it was sitting at .2x rather than going down to 0, and initially I only noticed because of the CA display.

I've noticed a couple of things that seem odd.

After using the throttle, sometimes all it takes is a weight change on the pedals and the motor will jump up to ~200 watts for a moment, even if I'm on a fairly low assist multiplier. The Hwt I'm producing during that weight change are below my "Assist Start Level" of 35 watts - so I'm surprised that the motor is turning on at all, but then it also seems to be applying a power at something like an 8x multiplier, even if I'm set at something more like 2x.

I've also noticed that the relationship between my human wattage and motor wattage being mismatched in the other direction under more normal usage, like, when I'm set for 7x and producing 100 Hwt, I might only see 500 watts as the power reading. Does that mean I have something configured incorrectly, or is it not necessarily intended to be a perfectly 1-1 kind of thing?

User avatar
teklektik   10 GW

10 GW
Posts: 4020
Joined: Mar 26 2011 1:15pm
Location: CT, USA

Adjusting CA3 AuxA Deadbands for POT input

Post by teklektik » May 15 2018 6:16pm

kevinscargobike wrote:
May 15 2018 4:23pm
I'm still trying to figure out the right size dead band to have on the aux pot so that I can set it to not assist at all...
The default values should work fine. However, if the % shown on the AuxA Preview screen does not go from 0-99% as the pot is turned min to max then follow one or both of the procedures below:

Setting the Lo Deadband:
  1. Turn the pot to the minimum
  2. Set the LoDeadband to 0.00V
  3. go to the AuxA prefix screen and check the %Limit.
    If non-0 then increase deadband in 0.01V steps until the %Limit goes to 0.
  4. Increase the deadband by 0.1V
Setting the Hi Deadband:
  1. Turn the pot to the maximum
  2. Set the HiDeadband to 0.00V
  3. go to the AuxA prefix screen and check the %Limit.
    If not 99% then increase deadband in 0.01V steps until the %Limit goes to 99%.
  4. Increase the deadband by 0.1V
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

brickwall   10 W

10 W
Posts: 85
Joined: Jun 27 2015 3:34am
Location: Stockholm, Sweden

Re: Cycle Analyst V3 preview and first beta release

Post by brickwall » May 16 2018 4:20am

I've finally managed to configure the CAv3 with firmware 3.1b to behave like I want to, except for the throttle response when riding. I had to employ strict throttle ramping to prevent the bike from flipping backwards when accelerating from a standstill so it's no longer dangerous and takes off smoothly, but that also means a sluggish throttle response when riding. I think I should use UpRate and FastRate for this, but I'm not sure how to set them. The UUG seems to suggest they are used for the opposite purpose:
• FastRate is similar to UpRate and is typically set to a faster rate. Values of 4-8V/s are typical.
• FastThrsh sets current at which the CA switches from FastRate to UpRate.
• If FastThrsh = 0.00A, FastRate is ignored and UpRate is always employed, otherwise:
• When current is less than FastThrsh, the CA uses FastRate to quickly advance Throttle OUT to the point where the motor is bearing load. When already underway, this shortens the ramping delay that would otherwise occur if using the slower UpRate ramping.
• When current is equal to or greater than FastThrsh, the motor is bearing load and the CA switches to UpRate to apply slower ramping.
FastThrsh should be set higher than the current necessary to accelerate the unloaded wheel to maximum rpm. This is always greater than the maximum unloaded current. Setting the current too low will cause the CA to prematurely switch to UpRate. Typical DD values are 0.5-1.0A, gear motor values are typically 2-4A.
How can I use these features to get a conservative throttle ramping from a standstill but a quick one while riding?

User avatar
teklektik   10 GW

10 GW
Posts: 4020
Joined: Mar 26 2011 1:15pm
Location: CT, USA

Re: Cycle Analyst V3 preview and first beta release

Post by teklektik » May 16 2018 10:34am

brickwall wrote:
May 16 2018 4:20am
I've finally managed to configure the CAv3 with firmware 3.1b to behave like...
Although you don't specify which beta version you are using, you should really upgrade to the production 3.1 release which has the issues repaired that were discovered in the beta versions.
brickwall wrote:
May 16 2018 4:20am
I had to employ strict throttle ramping to prevent the bike from flipping backwards when accelerating from a standstill so it's no longer dangerous and takes off smoothly, but that also means a sluggish throttle response when riding. I think I should use UpRate and FastRate for this, but I'm not sure how to set them. The UUG seems to suggest they are used for the opposite purpose:
...
How can I use these features to get a conservative throttle ramping from a standstill but a quick one while riding?
FastRate is really designed to be of assistance with drivetrains employing a freewheel clutch so the motor could spin up to speed rapidly and then regular ramping would take over after the clutch engages. There is no mechanism in the CA to alter UpRate at speed.

I suggest that you switch to Current Throttle so the throttle control is more disciplined at low speed which should not require such strong UpRate limiting. You will need to adjust AGain which will again have effect over the entire speed range - low values can make the throttle a little laggy so it's a balancing act.

A complimentary strategy is to reduce the phase current limit of your controller which will only have effect at very low speeds. By reducing the aggressive motor behavior at low speed, you won't need such aggressive settings in the CA and CA operation won't have such an adverse effect at higher speeds. Although reducing the available motor torque off the line (max phase amps) may seem Bad, if you are suppressing use of it through CA settings, then the high phase current limit isn't really being used anyway - you're just limiting it indirectly via the CA and paying a price for making the CA do the controller's job.
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

brickwall   10 W

10 W
Posts: 85
Joined: Jun 27 2015 3:34am
Location: Stockholm, Sweden

Re: Cycle Analyst V3 preview and first beta release

Post by brickwall » May 17 2018 2:08am

teklektik wrote:
May 16 2018 10:34am
Although you don't specify which beta version you are using, you should really upgrade to the production 3.1 release which has the issues repaired that were discovered in the beta versions.
Done and done! I just brought out my bike after a long Swedish winter.
teklektik wrote:
May 16 2018 10:34am
I suggest that you switch to Current Throttle so the throttle control is more disciplined at low speed which should not require such strong UpRate limiting. You will need to adjust AGain which will again have effect over the entire speed range - low values can make the throttle a little laggy so it's a balancing act.
I use power throttle in one of my three modes, for smoother acceleration. I will experiment with current throttle to see if that's even better. The really dangerous bit is passthrough throttle with no power limiting. Fun, but scary.
teklektik wrote:
May 16 2018 10:34am
A complimentary strategy is to reduce the phase current limit of your controller which will only have effect at very low speeds. By reducing the aggressive motor behavior at low speed, you won't need such aggressive settings in the CA and CA operation won't have such an adverse effect at higher speeds. Although reducing the available motor torque off the line (max phase amps) may seem Bad, if you are suppressing use of it through CA settings, then the high phase current limit isn't really being used anyway - you're just limiting it indirectly via the CA and paying a price for making the CA do the controller's job.
This is probably what I should do, though I need to open up the controller and do some soldering in order to be able to program it. I love the simplicity of the CA.

As always, thanks for your suggestions, teklektik. Very helpful!

EDIT: Yup, changing to current throttle was a much better solution. I could remove the throttle uprate limiting altogether after limiting current instead. Great!

User avatar
teklektik   10 GW

10 GW
Posts: 4020
Joined: Mar 26 2011 1:15pm
Location: CT, USA

Re: Cycle Analyst V3 preview and first beta release

Post by teklektik » May 17 2018 9:34am

brickwall wrote:
May 17 2018 2:08am
I use power throttle in one of my three modes, for smoother acceleration. I will experiment with current throttle to see if that's even better. The really dangerous bit is passthrough throttle with no power limiting. Fun, but scary.
brickwall wrote:
May 17 2018 2:08am
Yup, changing to current throttle was a much better solution. I could remove the throttle uprate limiting altogether after limiting current instead.
Great this worked out. That said, I'm a little surprised that Current Throttle worked better than Power Throttle - they are essentially the same but Power throttle can work better for saggy batteries to keep throttle performance more constant over the discharge cycle.

I suspect that in your case you are seeing the differences in Current/Power Throttle behavior because of your AGain (Current Throttle) and WGain (Power Throttle and PAS) settings - perhaps with the AGain giving somewhat softer response...

Anyhow - no need to turn this into a science project - sounds like it's working fine so better to just be riding and enjoying it. :D
Thanks for posting back with results!
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

brickwall   10 W

10 W
Posts: 85
Joined: Jun 27 2015 3:34am
Location: Stockholm, Sweden

Re: Cycle Analyst V3 preview and first beta release

Post by brickwall » May 17 2018 9:42am

teklektik wrote:
May 17 2018 9:34am
I suspect that in your case you are seeing the differences in Current/Power Throttle behavior because of your AGain (Current Throttle) and WGain (Power Throttle and PAS) settings - perhaps with the AGain giving somewhat softer response...
You're right. I've been tweaking WGain quite a lot but I don't remember even touching AGain since I haven't enabled current limiting or throttle until now. In other words, I should be able to achieve the same results by going back to power throttle/limiting and reducing WGain a bit. Maybe to the same level as AGain is currently at?

Tuning this thing to perfection has been so exhilarating, enjoyable and educational. Science project or not, I like to learn as much as possible. :)

User avatar
teklektik   10 GW

10 GW
Posts: 4020
Joined: Mar 26 2011 1:15pm
Location: CT, USA

Re: Cycle Analyst V3 preview and first beta release

Post by teklektik » May 17 2018 10:37am

brickwall wrote:
May 17 2018 9:42am
In other words, I should be able to achieve the same results by going back to power throttle/limiting and reducing WGain a bit. Maybe to the same level as AGain is currently at?
Yep - that's the ticket.
I'm not sure about the relative values - I haven't visited that blob of code in a while...

That said, it can sometimes be useful to use Current Throttle if you have PAS so WGain controls the motor for PAS (soft response) and AGain is separate and controls things for the throttle (brisk response). This business is pretty subjective - but, more options... :)
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

User avatar
izeman   10 GW

10 GW
Posts: 4731
Joined: Jun 21 2011 8:25am
Location: vienna, austria
Contact:

Re: Cycle Analyst V3 preview and first beta release

Post by izeman » May 24 2018 10:25am

teklektik wrote:
May 16 2018 10:34am
FastRate is really designed to be of assistance with drivetrains employing a freewheel clutch so the motor could spin up to speed rapidly and then regular ramping would take over after the clutch engages. There is no mechanism in the CA to alter UpRate at speed.
while reading that: any update on "slow rate" feature? you remember? applying only a little throttle for really powerful middrives until all the drive train slag is gone, and then apply full throttle.

User avatar
teklektik   10 GW

10 GW
Posts: 4020
Joined: Mar 26 2011 1:15pm
Location: CT, USA

Re: Cycle Analyst V3 preview and first beta release

Post by teklektik » May 24 2018 11:51am

Hey izeman!
To be honest we've been swamped so far this year: big push to get out 3.1, push to get out 3.11, push to build the SolarAnalyst for Sun Trip 2018, and a couple of other custom firmwares. If we decide to move forward with the 'slow rate' feature, it won't be until later in the year at the earliest due to other things in the wind before then. I appreciate this is something that you'd like to have in hand, but it's a resource issue...

That said, 3.11 should be out very soon with a couple of fixes for some annoying bugs as well as a nice PAS improvement. :D

Justin should be posting up some stuff about the Sun Trip Solar stuff soon as well...

SolarAnalyst.png
SolarAnalyst.png (185.53 KiB) Viewed 748 times
Last edited by teklektik on May 25 2018 4:45am, edited 1 time in total.
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

User avatar
ABritInNY   100 W

100 W
Posts: 114
Joined: Jul 12 2015 1:55pm

Re: Cycle Analyst V3 preview and first beta release

Post by ABritInNY » May 24 2018 11:42pm

I updated my CAv3 firmware to enable the aux dig and aug analog seperate settings and I was running the Grin switch for PAS and a 3 speed to change the amps. Today out of nowhere the screen alternates between 55a 65a and 75a by itself then 3 speed cut out and stopped working and the PAS only works to change up but down does not function? Later today I had issues with the throttle cutting in and out and only the PAS input is constantly working? what the heck could be up with my CA?
EEB frame set, 1500w Leafmotor 4t winding, Powervelocity 18fet/Lyen 18 fet 4110 ESC, 24s 30ah prismatic cell Lifepo4 batt w/60a-120a BMS, CA V3, Dropper Seat post.
Fusion 48v AGM (4x 12v 15ah) 1000w nom DD rear hub, 50a max ESC, 21sp FS Genesis overkill 650b conversion
Kayman Flash+ 350w Mxus geared hub, dual 36v 5.2ah batts 10.4ah 13a-15a ESC
Freway Vr01, 27sp Pedelec 250w Bafang geared hub, (2x) 36v 5.2ah batt, FVS ESC limited to 10a!! (yup sloow on roads but great on trails ;)

brickwall   10 W

10 W
Posts: 85
Joined: Jun 27 2015 3:34am
Location: Stockholm, Sweden

Re: Cycle Analyst V3 preview and first beta release

Post by brickwall » May 25 2018 3:46am

teklektik wrote:
May 24 2018 11:51am
SolarAnalyst.png
Ooooohhh what is this?

User avatar
teklektik   10 GW

10 GW
Posts: 4020
Joined: Mar 26 2011 1:15pm
Location: CT, USA

Re: Cycle Analyst V3 preview and first beta release

Post by teklektik » May 25 2018 4:23am

brickwall wrote:
May 25 2018 3:46am
Ooooohhh what is this?
http://thesuntrip.forumactif.org/t510-p ... ar-ca-code (images load very slowly)
viewtopic.php?f=6&t=93482

All part of:
https://www.thesuntrip.com/presentation ... tation-en/

Details to follow....
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

User avatar
teklektik   10 GW

10 GW
Posts: 4020
Joined: Mar 26 2011 1:15pm
Location: CT, USA

Re: Cycle Analyst V3 preview and first beta release

Post by teklektik » May 25 2018 4:43am

ABritInNY wrote:
May 24 2018 11:42pm
I updated my CAv3 firmware
...
what the heck could be up with my CA?
Hey ABINY-
I'm guessing that you have run into a bug in the 3.1 firmware that is addressed in the coming 3.11 release. My advice would be to roll back the firmware to a pre-3.1 release and then upgrade with 3.11 (next week?). PM me if you want to try an interim firmware solution instead.
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

redline2097   100 W

100 W
Posts: 162
Joined: Jun 18 2015 1:43am

Re: Cycle Analyst V3 preview and first beta release

Post by redline2097 » May 26 2018 2:52pm

Just a quick confirmation from you before I make anything dumb. :D

Does the Cycle Analyst supply battery pack voltage out from that 3.5mm headphone jack looking adapter?

So I could connect step down converter there and supply 12v to PAS, about 40mA.

It would be awesome that PAS would not use any sleep current. It would power on immediately when CA turns on.

User avatar
teklektik   10 GW

10 GW
Posts: 4020
Joined: Mar 26 2011 1:15pm
Location: CT, USA

Re: Cycle Analyst V3 preview and first beta release

Post by teklektik » May 26 2018 5:53pm

redline2097 wrote:
May 26 2018 2:52pm
Does the Cycle Analyst supply battery pack voltage out from that 3.5mm headphone jack looking adapter?
Not on the 3.5mm TRS stereo jack but on the barrel connector.
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

redline2097   100 W

100 W
Posts: 162
Joined: Jun 18 2015 1:43am

Re: Cycle Analyst V3 preview and first beta release

Post by redline2097 » May 27 2018 1:11pm

What are the best shunt settings for 150A 75mv 0. 5mOhm shunt?

User avatar
amberwolf   100 GW

100 GW
Posts: 25406
Joined: Aug 17 2009 6:43am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Cycle Analyst V3 preview and first beta release

Post by amberwolf » May 28 2018 2:09am

I don't think I understand the question.

There is only one shunt setting for a shunt, and that's the actual shunt value.

(unless the value is too low for the low range and isn't accepted, in which case you have to switch to high range to enter it, basically moving the decimal point over one place in the various associated readouts).

Post Reply