Page 185 of 185

Re: Cycle Analyst V3 preview and first beta release

Posted: Sep 26 2020 7:46am
by ZeroEm
by cnrd » Aug 04 2020 2:38am

Does anyone know if it's possible to set different speed limits for PAS and throttle?

I would like to set a lower speed limit on PAS, but still have extra speed available on the throttle.
Yes you can have different speed limits for PAS and the Throttle!
Here in Texas PAS can be up to 28 mph but Throttle only the speed limit is 20 mph. Different than what you want. Make sure regen is off or you will be pedaling and charging your batter when you hit the limits.

Re: Cycle Analyst V3 preview and first beta release

Posted: Sep 27 2020 11:45am
by alexayala.ut
Djhandz wrote:
Sep 26 2020 7:02am
Hi all. Reading a few horror stories about people blowing there cycle analyst
Please can someone tell me if it's safe to use this pas sensor

https://www.ebikes.ca/shop/electric-bic ... p-chr.html

Plugged in to the pas plug on 10v running a 22s battery fully charged 92v

Will this setup be ok or do I need to reposition the 10v to the 5v inside the cycle analyst?

Thanks
At 92V, you have about 8mA for accessories, per the unofficial user guide. Typical hall throttle takes 5mA, typical PAS takes 10mA. With just these two, you're asking almost double the current the CA can provide. You need a different power source for the PAS. You could pull from the controller throttle +5V, or run a dc converter off the DC power output jack on the CA to the PAS.

Re: Cycle Analyst V3 preview and first beta release

Posted: Sep 27 2020 11:58am
by Djhandz
Thankyou very much alexayala. Very understanding answer
I will get power from my controller throttle

Thanks again dude

Re: Cycle Analyst V3 preview and first beta release

Posted: Sep 29 2020 2:03am
by HrKlev
justin_le wrote:
Sep 23 2020 6:09pm
HrKlev wrote:
Aug 03 2020 3:30pm
Is there any reason why the "electric freewheel" feature that is inlcuded in the Phaserunner controllers is not implemented in the Cycle Analyst?
This is a great question and was actually on the CA3 roadmap for a while, but we decided to pull it from the list. As NCC1941 pointed out, the CA3 only sees your battery current, not your phase current. At constant phase current the battery current is almost 0A at low speeds and then linearly increases until reaching your actual phase amps once you get near the unloaded cruising speed. It is possible for the CA3 to emulate this behavior and have a setting that is a fixed watts/kph, so as you go faster then it will inject proportionally higher watts even with the throttle off. But even with that algorithm in place it there were also the following two problems:
a) Many motor controllers do not behave well at all when you try to command a very low wattage from them in a constant watts by modulating the throttle. and
b) Many setups have lights, DC-DC converters, and other accessories powered by the CA in a way that the CA also shows the watts of these accessory components. So then if you flip on your headlights or blast a horn or whatever it could add some 10's of watts to the CA's power reading which it couldn't distinguish from watts going into the motor which would mess up this behavior.
Thanks a lot for the in-depth explanation. It makes a lot of sense. I am enjoying the feature on a phaserunner controller now :)

I lnow the CA is a little limited now, but Im just throwing an idea out here; Would it be possible to implement a EMTB mode/ torque mode? It is a mode for tecnical terrain where the output relies on the torque input from the BB torque sensor, and not torque and rpm combined. Im getting pretty much the behavior as I want with setting the avg pole to 1, and fast PAS ramps, but a torque mode would be icing on the cake. Assist from zero rpm would also be very useful when not using a throttle (not allowed here, unfortunately). Maybe in a new version in the future if its not possible with todays HW?

Edit: So I just realized after I wrote this.... Since Im not using throttle, maybe I could just split the 0-5v signal from the torque sensor to the throttle input and set up one of the presets to be throttle input only. That should get me torque mode with assistance from 0rpm, and I could adjust the throttle input ramping to get the desired behaviour.... hmm....

Re: Cycle Analyst V3 preview and first beta release

Posted: Sep 29 2020 3:23am
by justin_le
HrKlev wrote:
Sep 29 2020 2:03am
It is a mode for tecnical terrain where the output relies on the torque input from the BB torque sensor, and not torque and rpm combined.
Yes, actually more to the point I'm strongly considering in the next update that we just drop the watts/Hwatt scaling factor and have it instead be Watts/Nm. This is actually the original implimentation of the CA3 way way in the early beta/alpha firmwares. It was switched over to being W/HW because we thought that would be a more straightforward parameter for people to understand (oh, it just doubles my power), but I think most people would prefer if the motor power scaled linearly with pedal torque regardless of RPM. Then the RPM is just an on/off gate, and it makes it easier to have torque assist at 0 rpm, and if after like half a second ifthere is still no pedal cadence detected then it would shut off the PAS. Probably not in 3.15, but likely in 3.2.
Edit: So I just realized after I wrote this.... Since Im not using throttle, maybe I could just split the 0-5v signal from the torque sensor to the throttle input and set up one of the presets to be throttle input only. That should get me torque mode with assistance from 0rpm, and I could adjust the throttle input ramping to get the desired behaviour.... hmm....

You can absolutely do this. And you can actually do it even with a throttle too by just parallel connecting the analog signals together. The throttle can't pull down but it can pull up which means you don't need any diodes for this. Just put like a 500 Ohm resistor between the torque signal and the throttle input while the throttle signal connects directly. If you don't use the throttle, the signal will come from the torque sensor, and if you do use the throttle then it will generally read the higher of the torque or the throttle signal.

Re: Cycle Analyst V3 preview and first beta release

Posted: Sep 29 2020 6:53am
by HrKlev
justin_le wrote:
Sep 29 2020 3:23am

Yes, actually more to the point I'm strongly considering in the next update that we just drop the watts/Hwatt scaling factor and have it instead be Watts/Nm. This is actually the original implimentation of the CA3 way way in the early beta/alpha firmwares. It was switched over to being W/HW because we thought that would be a more straightforward parameter for people to understand (oh, it just doubles my power), but I think most people would prefer if the motor power scaled linearly with pedal torque regardless of RPM. Then the RPM is just an on/off gate, and it makes it easier to have torque assist at 0 rpm, and if after like half a second ifthere is still no pedal cadence detected then it would shut off the PAS. Probably not in 3.15, but likely in 3.2.
Very, very interesting! Just a note from a tsdz2 equipped bike I had where I could choose pure torque lode; It was very hard to maintain a high cadence when I was cycling on a normal road. It might be something with the torque sensor or the motor, so maybe it is not applicabke at all.... But it was a lot easier with power mode. So maybe have a way to switch between the modes?

justin_le wrote:
Sep 29 2020 3:23am


You can absolutely do this. And you can actually do it even with a throttle too by just parallel connecting the analog signals together. The throttle can't pull down but it can pull up which means you don't need any diodes for this. Just put like a 500 Ohm resistor between the torque signal and the throttle input while the throttle signal connects directly. If you don't use the throttle, the signal will come from the torque sensor, and if you do use the throttle then it will generally read the higher of the torque or the throttle signal.
Awesome! Sorry for being such a electronics noob (I have no idea what that resistor does :lol: )... do you mean like this?
IMG_29092020_134006_(1000_x_922_pixel).jpg
IMG_29092020_134006_(1000_x_922_pixel).jpg (179.21 KiB) Viewed 782 times
Im really excited how this will turn out. The electronic freewheel feature makes the mid drive system so responsive, and with a torque mode in addition I think it will be even better for the technical stuff.

Re: Cycle Analyst V3 preview and first beta release

Posted: Oct 03 2020 3:40pm
by DtiK
Is there possibility to get somewhere Grin tech stickers? 8)

Re: Cycle Analyst V3 preview and first beta release

Posted: Oct 03 2020 10:26pm
by ZeroEm
I want one, throw it in my next order :wink:

Re: Cycle Analyst V3 preview and first beta release

Posted: Oct 07 2020 10:43pm
by michaelkramer11
Please Help?!

Sorry if this has been mentioned or discussed before, but i've been dealing with this problem for years and i've finally decided to look for answers.

I got my cav3 dp back in 2014 and for as long as I can remember I've had sporadic screen cutouts/random characters flash on the screen or the screen will just go blank. It continues to operate when the screen cuts, so it hasn't been a total loss, but I know these products don't typically have these issues. The backlight also stays on the whole time so I know its not losing power. I did wire this unit into a controller purchased off of ebay but i'm quite certain the problem is in the CA.

I'm running 16s lipo's and i feel like the handful of times that I've powered up without the pre-charge circuit could be to blame? Is there anywhere to look for diagnosis/fix tips?

Re: Cycle Analyst V3 preview and first beta release

Posted: Nov 16 2020 2:20pm
by grizzlymann
It looks like the attachment for the Unofficial User Guide might have been removed. Does anyone have a copy or could restore the file?

Re: Cycle Analyst V3 preview and first beta release

Posted: Nov 16 2020 11:01pm
by amberwolf
Just be aware that Teklektik is likely to ream a new asshole in anyone that actually posts it, as he did me some years back when I attempted to help everyone by copying it to the archival Tech forum here, and insisted that *only* links to his single post of the file (that apparently now no longer exists) were allowed. He was quite strongly and angrily rude about it.

In a quick search, I couldn't find another post of his that contains the file, so unless he's not going to light into people for reposting it now that it is gone from there, I dont' see any other way to get it back, as there's no file to link to now.

Re: Cycle Analyst V3 preview and first beta release

Posted: Nov 16 2020 11:06pm
by john61ct
Wayback Machine at the Internet Archive?

Re: Cycle Analyst V3 preview and first beta release

Posted: Nov 17 2020 5:01am
by Santacruz
Both the official and unofficial are available on Grins site under Cycle Analyst

Re: Cycle Analyst V3 preview and first beta release

Posted: Nov 17 2020 11:50pm
by amberwolf
I only see a link to this thread for the UUG on the CA info pages at Grin's site.

Re: Cycle Analyst V3 preview and first beta release

Posted: Nov 18 2020 3:02am
by Santacruz
amberwolf wrote:
Nov 17 2020 11:50pm
I only see a link to this thread for the UUG on the CA info pages at Grin's site.
My appolgies.
It was only a few weeks ago that I downloaded the UUG. I see now that it is not available.
So, disregard my post, it is obviously wrong.
But, Grin should remove it from their site, I think.

Re: Cycle Analyst V3 preview and first beta release

Posted: Nov 18 2020 10:43am
by john61ct
5sec google found the B22 version, is that latest?

https://www.google.com/search?q=Grins+C ... unofficial


Re: Cycle Analyst V3 preview and first beta release

Posted: Nov 18 2020 10:57am
by john61ct
Looks from Wayback, latest version was

document version 3.0-e
2016.02.16
teklektik

Found this at a German site fasterbikes.eu
CaV3-Anleitung-Endless-SphereojLDyj5SH3bt5PfU69hV2rlPoH.pdf

https://www.google.com/search?q=CaV3-An ... 2rlPoH.pdf

Made a backup https://drive.google.com/file/d/1635vQM ... p=drivesdk

Re: Cycle Analyst V3 preview and first beta release

Posted: Nov 23 2020 8:48am
by EddySPalm
I have, in two different ways (which I don't intend to reveal as I don't want to destroy my reputation on the forum :lol: ) burnt the PCB/main PIC of my CA's.

I am now currently on my third PCB. I choose to desolder the pads as this is quicker than rerouting the wires through my old Vespa frame again. The newest CA PCB I have which I am now soldering to, is missing a whole bunch of solder pads, including the thermistor pads.

being able to monitor temperature was actually one of the key features I wanted in my CA. Now I'm "stuck" with two loose wires inside the CA which don't have anywhere to go. or do they? Did Grin do something to their PCB that allows me to solder them elsewhere?

I cannot find any info that refers to the "sudden" lack of solder pads on the newer PCB's. Wherever I find info on how to convert CA's from one type to another, it says that all PCB's are the same, so you can convert to and from anything you'd like.

I have asked Grin in an e-mail also, but half of my e-mails remain unanswered, and I am really eager to get going on this project...

Re: Cycle Analyst V3 preview and first beta release

Posted: Jan 04 2021 3:15pm
by HrKlev
justin_le wrote:
Jun 03 2020 3:25am

Agreed. We will aim to support inputting higher PAS pole capabilties in future releases.
Any news on this, or other smart workaround? I have a 36 pole torque sensor on one of my bikes. Im using 18 pole setting for now, and just use half the assist numbers and divide all the statistics in 2. It works kind of OK, but one of my favorite things with the CA is all the data you get instantly, and when the numbers are not what they really are it is quite annoying :p