Cycle Analyst V3 preview and first beta release

Get real world experience and user feedback on the electric bicycle products.
User avatar
cruzxia   100 W

100 W
Posts: 101
Joined: May 18 2012 5:27am

Re: Cycle Analyst V3 preview and first beta release

Post by cruzxia » Aug 16 2012 2:12am

Justin

My setup is as follows:
Battery's - 18s lipo 75-61 volt Low voltage cutout is at 61v on the controller and 61.5v on the CA
Controller - 6 fet Lyen/Infineion direct plug in CA , shunt 2.000mohm
Throttle - Legacy (I plan to change this once the hang is resolved)
Motor - BPM Code 12

Problem 1 I have not had occur

Problem 2
The throttle dose this on every startup. I cant pinpoint what corrects the issue, I just wind the throttle on and off several times and suddenly it goes. I was wandering if it was time related, but I don't think so.
note:You do not have to go full throttle, to clear this, I just tested 0-1/4 throttle and it cleared.

Problem 3
This occurred yesterday again, as I was nearly home so I did not stop. after about 20 seconds it started working again.
What I noticed though was that after the event, the human power input was jumping up and down on the left of the main screen, I dont use the torque input. I then flicked through the screens, and the RPM was rolling numbers.

I think that there is the possibility that with no inputs on the torque throttle in and aux connectors, that spikes may be induced to these inputs. If the function is not being used (disabled in settings) you should disable the inputs and tie the ports high or low, so that there is no chance of induced voltages causing issues.
Once this issue is resolved, you should enable the brownout and watchdog functions in the PIC if you haven't already done so.

There is also the possibility of glitches through the main input, I have not studded the CA PCB, If you need help with this issue, you can PM me and I can try to help you out. I have been programing PIC Micros and designing PCB's for the auto and other industries for over 20 years.

cruzxia
Last edited by cruzxia on Aug 16 2012 2:43am, edited 1 time in total.

User avatar
justin_le   10 MW

10 MW
Posts: 2152
Joined: May 28 2007 3:27am
Location: Vancouver
Contact:

Re: Cycle Analyst V3 preview and first beta release

Post by justin_le » Aug 16 2012 2:30am

cruzxia wrote: Problem 3
This occurred yesterday again, as I was nearly home so I did not stop. after about 20 seconds it started working again.
What I noticed though was that after the event, the human power input was jumping up and down on the left of the main screen, I dont use the torque input. I then flicked through the screens, and the RPM was rolling numbers.
cruzxia
Thanks, that's exactly what Teklektik observed as well when he had the same thing happen. If it's possible for you to ground the "RPM" input from the 5-pin connector and see if it still occurs or not that could go a long way towards cluing in on the origins. We definitely had issues in earlier board revisions where noise generated during regenerative braking would cause signals to show up on the RPM line for false pedal cadence readings, and had a cap added to solve that. But the pattern of it showing up here after a freeze seems unrelated.
Once this issue is resolved, you should enable the brownout and watchdog functions in the PIC if you haven't already done so.
Indeed another great point. The B17 firmware attempted to use the watchdog so these dangerous hang situations would cause a chip reset right away, but unfortunately the bootloader can't change the configuration words in the PIC to actually enable the watchdog function at the hardware level. So for the beta units on the field it won't be easy to get this protection unless flashed with an in circuit programmer.
There is also the possibility of glitches through the main input, I have not studded the CA PCB, If you need help with this issue, you can PM me and I can try to help you out. I have been programing PIC Micros and designing PCB's for the auto and other industries for over 20 years.
Thanks, we'll see what I can sort over the next couple days and may take you up on that! -Justin
Last edited by justin_le on Aug 16 2012 2:33am, edited 1 time in total.
Reason: better english
Currently recovering from the Suntrip race on a back to back tandem solar powered row/cycle trike. 550 watt solar roof, dual Grin All Axle hub motors, dual Phaserunner controllers, 12 LiGo batteries, and a whole wack of gear.

Now back in Vancouver with my Big Dummy Frame (yes This One, thanks ES!) with Grin all-axle front hub, Phaserunner controller, and 52V 19Ah Cellman triangle pack
My website: http://www.ebikes.ca
Please contact via email, info@ebikes.ca, rather than PMs, which are disabled

User avatar
Diamondback   10 kW

10 kW
Posts: 525
Joined: Dec 15 2008 3:52am
Location: Sydney, Australia

Re: Cycle Analyst V3 preview and first beta release

Post by Diamondback » Aug 16 2012 3:14am

Hi Justin

Can you please confirm if the CA can power the torque sensor if I am running 12s lipo ?
That's 50.4v hot off the charger.

I think I'll wait until the final versions of the CA are finished.
I don't really like the idea of having a glitchy unit until the final FW
Is released. Last thing I need is a WOT situation !

I have really liked the two previous CA incarnations I've owned.
Looking forward to getting my hands on V3 !

I think I'll need to swap out my controller when I get the V3 and tq
Sensor though. My controller seems to have a delay in ramping up.
I'm sure this would not play nice with the way I intend to run it.

Thanks again for all the work you and the team have put into
Making this awesome device.

Jason.
Ignorance can be solved, stupid is forever

User avatar
justin_le   10 MW

10 MW
Posts: 2152
Joined: May 28 2007 3:27am
Location: Vancouver
Contact:

Re: Cycle Analyst V3 preview and first beta release

Post by justin_le » Aug 16 2012 3:24am

Diamondback wrote:Hi Justin

Can you please confirm if the CA can power the torque sensor if I am running 12s lipo ?
That's 50.4v hot off the charger.
We've been running them off 13S LiMn without issue yet, which is about 54V off the charger, but the climate isn't as hot here as other parts of the world. It's better if you don't also have a hall throttle attached at the same time (which draws another 5mA).
I think I'll wait until the final versions of the CA are finished.
I don't really like the idea of having a glitchy unit until the final FW
Is released. Last thing I need is a WOT situation !
That's smart for sure, it's Beta for a reason and may be for a while yet.
I think I'll need to swap out my controller when I get the V3 and tq
Sensor though. My controller seems to have a delay in ramping up.
I'm sure this would not play nice with the way I intend to run it.
Our experience is that if the controller's ramp delay time is more than 1 second it can be really problematic to control, but less than 0.5 sec tends to regulate no problem. So if you pick the wheel off the ground and time how long it takes to reach full RPM after you apply full throttle, you should get a good idea. You can always try with your existing controller first, and only if it seems impossible to tune it switch to another model.

-Justin
Currently recovering from the Suntrip race on a back to back tandem solar powered row/cycle trike. 550 watt solar roof, dual Grin All Axle hub motors, dual Phaserunner controllers, 12 LiGo batteries, and a whole wack of gear.

Now back in Vancouver with my Big Dummy Frame (yes This One, thanks ES!) with Grin all-axle front hub, Phaserunner controller, and 52V 19Ah Cellman triangle pack
My website: http://www.ebikes.ca
Please contact via email, info@ebikes.ca, rather than PMs, which are disabled

User avatar
Diamondback   10 kW

10 kW
Posts: 525
Joined: Dec 15 2008 3:52am
Location: Sydney, Australia

Re: Cycle Analyst V3 preview and first beta release

Post by Diamondback » Aug 16 2012 3:34am

Thanks for the info.

My current ezee controller seems like its about 2 sec.
If you nail the throttle, it seems to be about 2 sec before anything happens.

I won't be running any other throttle. I want it set up solely as a human / electric
Hybrid. So the motor actively rewards you for pedalling. No pedal = no go.

Yeah, temps here can soar into the low 40's (c) here in summer.
In fact where I work it got up to 47c last summer one day.


I can wait as long as it takes to get this done.

I'll just keep watching this thread with great interest.

Jason.
Ignorance can be solved, stupid is forever

crusoe   100 W

100 W
Posts: 191
Joined: Apr 03 2011 12:52pm
Location: Ottawa, Ontario

Re: Cycle Analyst V3 preview and first beta release

Post by crusoe » Aug 16 2012 4:31pm

Not sure if it's already been covered, but any chance of a digital clock on the CA?

Architectonic   1 kW

1 kW
Posts: 491
Joined: Nov 05 2011 8:46pm
Location: Adelaide, Australia

Re: Cycle Analyst V3 preview and first beta release

Post by Architectonic » Aug 16 2012 9:10pm

crusoe wrote:Not sure if it's already been covered, but any chance of a digital clock on the CA?
Or a stopwatch? :wink:

User avatar
cruzxia   100 W

100 W
Posts: 101
Joined: May 18 2012 5:27am

Re: Cycle Analyst V3 preview and first beta release

Post by cruzxia » Aug 17 2012 12:26am

Justin

More on the throttle issue. I removed the throttle diode from my controller and replaced it with a 1K resistor, in preparation for moving to pass through throttle. As there is a diode in the CA, all should be ok.

Still connected as legacy, when I turned the power on, the motor started up for a few seconds and then turned off, then a few seconds later it did it again. It went on and off about 3 times and then it stayed on.

I turned down the throttle out max to 1 volt and it stopped. Increased to 2 volts and it started up again.

I think that the slow running at startup is linked to the output on the TH0 for pass through mode. like wise I think that the pass through is still outputting when in legacy mode. I have disabled the throttle in and set the values as 0.00 and 0.02 as instructed.

This should help you with this issue, to try and pinpoint the problem.

I put the diode back for now.

Cruzxia

User avatar
shorza   1 kW

1 kW
Posts: 488
Joined: Oct 13 2010 11:29pm
Location: Melbourne, Australia

Re: Cycle Analyst V3 preview and first beta release

Post by shorza » Aug 17 2012 2:44am

crusoe wrote:Not sure if it's already been covered, but any chance of a digital clock on the CA?
http://endless-sphere.com/forums/viewto ... ck#p564552
http://endless-sphere.com/forums/viewto ... ck#p564862

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 » Aug 17 2012 4:07am

Summary of Throttle (Diode) Pad Connections by PCB Revision
cruzxia wrote:I removed the throttle diode from my controller and replaced it with a 1K resistor, in preparation for moving to pass through throttle. As there is a diode in the CA, all should be ok.

Still connected as legacy, when I turned the power on, the motor started up for a few seconds and then turned off, then a few seconds later it did it again. It went on and off about 3 times and then it stayed on.

I turned down the throttle out max to 1 volt and it stopped. Increased to 2 volts and it started up again.
There is a fine point in Justin's explanation here that relates to the differences in throttle output pin labeling between the various CA PCB revisions. Here is a table that may help clarify the situation:
CA-Throttle-Output-By-Rev.gif
CA-Throttle-Output-By-Rev.gif (4.63 KiB) Viewed 3415 times
The labeling change between v2 and v3 is a bit confusing and seems to be what derailed your efforts. In your case since neither the controller nor the CA had a diode and your CA had no throttle input (legacy mode), when you raised ThrO->MaxOutput to 2v, ThO exceeded the controller minimum throttle voltage causing the motor to run.

The v3 Thd output in the 2nd photo in this post was not discussed in Justin's post but allows operation in legacy mode with controllers that lack an internal diode. Relocating the CA V3 DP throttle connection (green) from ThO to Thd would have introduced a diode as you expected and made your modifications operate properly.
Last edited by teklektik on Jan 27 2013 1:54pm, edited 2 times in total.
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

User avatar
cruzxia   100 W

100 W
Posts: 101
Joined: May 18 2012 5:27am

Re: Cycle Analyst V3 preview and first beta release

Post by cruzxia » Aug 17 2012 5:15am

Thanks for clarifying that, I did scan through, and all I could find on THd was that it was for future use.

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 » Aug 17 2012 9:28am

cruzxia wrote:Thanks for clarifying that, I did scan through, and all I could find on THd was that it was for future use.
Thd was explained to me in the context of the new CA DP interface that Grin Tech will be proposing and that Justin described in his post. By removing the diode in the controller, you created a situation similar to the coming interface standard where that 'future use' scenario would be in play to allow use of a legacy throttle.

Your point is a good one - my annotation on that photo should probably omit the 'future use' phrase as unhelpful and perhaps misleading. Thanks!

EDIT - Photo revised. Also added note about ThO w/o diode to help minimize confusion.

cruzxia - Your phased implementation strategy was a good one and it's unfortunate the information to support your efforts was not clear. But thanks for the detailed post - maybe the doc is a little more squared away now so others won't have the same difficulties.
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 » Aug 21 2012 11:16am

justin_le wrote:I've attached here the B17 firmware here, ... The only significant difference from the B16 is that the leading zero's in the setup menu are implemented in a way that they are hidden when showing the value, but all show up the moment you go to edit a number. Hopefully that does make it the best of both worlds.
This is a somewhat belated post, but the numeric reformatting worked out great. It also fixed the invisible zero values that displayed with just a decimal point ('_._') and added a leading zero before fractional values so '.99' becomes the more familiar '0.99'.
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

User avatar
justin_le   10 MW

10 MW
Posts: 2152
Joined: May 28 2007 3:27am
Location: Vancouver
Contact:

Re: Cycle Analyst V3 preview and first beta release

Post by justin_le » Aug 21 2012 2:59pm

cruzxia wrote:I have been testing the limited power on startup again.

After using the throttle a few times, suddenly it goes back to unrestricted full power, and is fine until you restart the CA

cruzxia
Hey guys, just so you know know Beta18 version of the firmware is being field tested a bit and will go live soon. The issue with it sometimes failing to boot with noisy on/off switches has been fixed. However, for the above problem people have had with the throttle initially cutting low a few times during starting, can those who've experienced this try setting the differential gain term to zero? So in setup->Speed Limes-> DSGain, change it to 0 rather than the default of 2. I'd like to see for sure if this was the source.

What I think was happening is that the moment that the CA first started to detect an actual speed after the 2nd magnet pass, its calculated vehicle acceleration could be very high or ill-defined. Even if the current speed is well below the speed limit, if the CA thinks it is accellerating fast enough towards that limit then the throttle output could drop low without triggering the overspeed flag.

-Justin
Currently recovering from the Suntrip race on a back to back tandem solar powered row/cycle trike. 550 watt solar roof, dual Grin All Axle hub motors, dual Phaserunner controllers, 12 LiGo batteries, and a whole wack of gear.

Now back in Vancouver with my Big Dummy Frame (yes This One, thanks ES!) with Grin all-axle front hub, Phaserunner controller, and 52V 19Ah Cellman triangle pack
My website: http://www.ebikes.ca
Please contact via email, info@ebikes.ca, rather than PMs, which are disabled

User avatar
cruzxia   100 W

100 W
Posts: 101
Joined: May 18 2012 5:27am

Re: Cycle Analyst V3 preview and first beta release

Post by cruzxia » Aug 21 2012 7:35pm

Hi Justin

Yes that was the problem, I set the DSGain to 0 and it starts up unrestricted. To confirm I returned the setting to 2 and it restricted again. I did notice that the DSGain when set to 0 does not show any number on the display.

I have not had the unit cut out on me during the last few rides, however I did drop the current settings in my controller from peaks of 49A down to 34A. So it is possible that the high current was causing enough noise to affect the CA. It also could be the crystal falling over from the noise and then restarting, You can monitor the crystal in the firmware, and if it stops for any reason, you can switch to the internal oscillator to keep the micro running until the crystal reboots. I implement this feature in most designs and it works well, it switches over in ms.

Re Watchdog - I have a programmer, so I can reflash the whole unit including the config data, to allow use of the watchdog and brownout features, if you decide to enable them as part of your firmware.

I will be connecting the through CA throttle as soon as the JST connectors arrive in the mail. I will advise if i run into any issues.

Cruzxia

User avatar
hillzofvalp   1 MW

1 MW
Posts: 1861
Joined: Dec 26 2010 2:09am
Location: Los Angeles, CA

Re: Cycle Analyst V3 preview and first beta release

Post by hillzofvalp » Aug 21 2012 10:05pm

shorza wrote:
crusoe wrote:Not sure if it's already been covered, but any chance of a digital clock on the CA?
http://endless-sphere.com/forums/viewto ... ck#p564552
http://endless-sphere.com/forums/viewto ... ck#p564862
BUT! For a hobbyist extreme edition I would love to see a 0-XX speed stopwatch so that we can get accurate accerleration data... Maybe this would be more manageable in the existing code structure

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 » Aug 22 2012 6:19pm

hillzofvalp wrote:
shorza wrote:BUT! For a hobbyist extreme edition I would love to see a 0-XX speed stopwatch so that we can get accurate accerleration data...
Buy an Analogger and run the data dump at 5Hz for 0.2s data timing interval.
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 » Aug 22 2012 6:31pm

justin_le wrote:... for the above problem people have had with the throttle initially cutting low a few times during starting, can those who've experienced this try setting the differential gain term to zero? So in setup->Speed Limes-> DSGain, change it to 0 rather than the default of 2. I'd like to see for sure if this was the source.
cruzxia wrote:... I set the DSGain to 0 and it starts up unrestricted.
I concur. SLim->DSGain=0 seems to alleviate the cutout-on-startup problem.

Since I do not normally run with speed limiting where a non-0 DLim setting would otherwise play a role in proper operation, this is a completely effective workaround until the firmware can be revised.
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

lizardboy   10 mW

10 mW
Posts: 34
Joined: May 31 2011 6:19pm

Re: Cycle Analyst V3 preview and first beta release

Post by lizardboy » Aug 26 2012 10:37am

The voltage change is about .3 volts it displays this higher value after power up but drops after a reset. I notice because sometimes the CA shows my lipoly 10s string as 41.6 volts off the charger. Which is the value i got before I calibrated the voltage input. After reset it drops to 41.3 which is what the charger and my voltmeter say. Sometimes the reset does not change the value. If this happens then the random cutouts and instability happen.

I too experience the random cutouts usually for only a second or so and only at max power such as climbing something at 15 km/hr or so. Always a suprise but it usually comes back if i reapply the throttle. I also get a whole different cutout that I isolated months ago. It's caused by using a switch on the low power red wire on the controller. If I switch the low power red wire (not the main power red wire) then the throttle cutout will occur everytime it's taken at time 50 to 60 power cycles of the main power to break it out of this mode. I've simply given up on using the power switch as no fix seems to be coming for this issue. This weekend I forgot and turned it off using the button and couldn't get it to work again without cutting out for almost an hour.
Last edited by lizardboy on Sep 04 2012 11:19pm, edited 5 times in total.

User avatar
cruzxia   100 W

100 W
Posts: 101
Joined: May 18 2012 5:27am

Re: Cycle Analyst V3 preview and first beta release

Post by cruzxia » Aug 27 2012 1:38am

Just to let you know B17 is still hanging.

I have removed the Power out connector and wire and the torque sensor connector and wire from the CA PCB, as I do not use them.
Today I was riding at a steady pace with low current, and the CA screen froze for about 30 seconds. Just before it froze I noticed the Torque sensor input display line on the main page turning on and off. I rode on for about 500m and stopped, and by then it had frozen. I got my phone out to take a photo, but it had turned off, Just as I got it powered up the CA started working again.

Cruzxia

User avatar
The Riddler   1 mW

1 mW
Posts: 13
Joined: Oct 30 2010 5:16pm

Re: Cycle Analyst V3 preview and first beta release

Post by The Riddler » Aug 31 2012 10:55am

Would that work?
CA3.jpg
CA3.jpg (107.09 KiB) Viewed 489 times

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 » Aug 31 2012 11:13am

I cannot speak to the PAS portion (no PAS experience) but the throttle setup needs a small tweak.
ca3_annotated.gif
ca3_annotated.gif (24.16 KiB) Viewed 3058 times
Regarding the controller connection: It looks like the controller has two choices for throttle input (SP, SLA). In either case the throttle input will get pulled to zero in case of a broken throttle connection. I would personally use the same sex controller connection as my throttle so I had the option of re-plugging the throttle in the field to bypass the CA in the event of an issue or CA crash damage.
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

User avatar
shorza   1 kW

1 kW
Posts: 488
Joined: Oct 13 2010 11:29pm
Location: Melbourne, Australia

Re: Cycle Analyst V3 preview and first beta release

Post by shorza » Sep 04 2012 2:33am

I have found a few bugs/quirks:

ver.17

Most times when i first switch the bike/CA on and try to take off, I get about 2 seconds of throttle before it cuts out for a couple of seconds and back on for good.

I had the CA freeze on me the other day. Luckily I didn't i wasn't throttling on at the time. I noticed as i was trying to change pages. Hasn't happened again.

Amps drawn is always up to 0.3a off zero when starting the bike. It could be positive or negative, and tends to drifts during the day.
Zeroing the amps in the calibration menu fixes it for a while, but doing that also zeros the A/H reading i had at the time.

Dingo2024   100 W

100 W
Posts: 243
Joined: Dec 30 2009 12:57pm
Location: Yorkshire UK
Contact:

Re: Cycle Analyst V3 preview and first beta release

Post by Dingo2024 » Sep 04 2012 7:54am

Carzy wrote:
We've also received our full order of Thun sensors now so for those looking to get up and running with one you can purchase through the CA3 page: http://www.ebikes.ca/store/store_CAV3.php
Hi Justin,

had a look at the pdf of the Thun sensor.................can this be used on a 72mm bottom bracket frame?

thanks,

Ian :D
started as a boy, grew into a man....not sure where I am headed now
build thread; http://endless-sphere.com/forums/viewto ... =6&t=18130
top speed now stands at 43.9 mph

Gab   100 W

100 W
Posts: 282
Joined: Aug 05 2012 7:18am

Re: Cycle Analyst V3 preview and first beta release

Post by Gab » Sep 10 2012 4:19pm

is it possible to implement a motor rpm counter and rev limiter ? to protect overvolted unloaded motors

Post Reply