Please download the [strike]Un[/strike]official Guide and refer to section "4.5.1 Configure Current Range and Shunt".redline2097 said:
Cycle Analyst V3 preview and first beta release
- Thread starter justin_le
- Start date
ABritInNY
100 W
- Joined
- Jul 12, 2015
- Messages
- 118
Good guess Tek, that's what I thought, so I grabbed a spare CAV3DPS I had in Stock (with the old firmware duh!) and went to test it but obviously it needed upgrading to even work, so after doing that I plugged in and same issue?? So more confused, thinking I'd exported the bug with my, read from (old CA) and fw update and write to new CAdps, on a hunch I grabbed another grin up down (pas) double aux input switch and tested it... Lo and behold it all works, eureka! I think I've solved it. I unplug the dps and switch all inputs back to the original CA and with the new up/down switch, the 3speed and PAS all work... I'm happy, so I closed up the bike panels... I tested it on the stand and... no throttle...PHUUUUUUCK... open up panels again! switch units...working throttle, vica versa and now the throttle works! Yay, I'm done...teklektik said:
I close up...test it and again 2 mins later I get no throttle signal Well to cut the story (and frustration) short, it turns out the throttle jst connector, or wires have an issue, some kind of loose or bad connection? When held in a pinch grip the throttle works, (or changing plugs over,) but let the wires go and 2 min later the throttle 'dies.' So I changed to the dps permanently and removed the old CA and bad switch. I've no idea what caused this, but the fw update is now far down the list of culprits. But hey with electronics you never really know?
Now however, I've no speedo, because I've not used a dps before and naively thought they could work both ways, either like a DP, or with a spoke mag. But one magnet is way too slow plus low speed tracking would be awful but it's all I have, so now I've been advised to hack the speedo cable and attach a link from the white wire (or black,) to any of my halls, is that correct?
Not sure why you say 1 magnet is 'slow'... reset #poles =1 and all should be fine.ABritInNY said:
Otherwise, please download the [strike]Un[/strike]official Guide and refer to section "Appendix B. Add/Remove Wheel Speed Pickup Sensor". No need to hack any cables.
ABritInNY
100 W
- Joined
- Jul 12, 2015
- Messages
- 118
You're right, I shouldn't say slow it's just after being spoiled by being accustomed to 23 pole pairs reporting their pulse signal Vs 1 magnet reporting, (I guess I could fit more spoke mags to 8 equidistant spokes and have a similar response to what I'm used to and change the settings to 8.teklektik said:
But I think a strategic (and easily reversible,) splice into the sheath of the speed sensor wire, to reveal the white wire and just patch that with a wire direct to the yellow (or any colour) hall signal in the 6 pin connector. To me that is easier to do now and also opening up the case needlessly, (with my luck) would probably make my unit lose it's waterproofing or something else lol.
I also want to know, of I should upgrade to the latest firmware version as I'm running the original update that enabled the split functionality, I did a clean template copy for that, do I need to do anything different to upgrade to b22? Is it? Or can I just updated and use the new slick looking uploader program?
Magnets do not need to be equidistant. I use three - it's really just a matter of better speed sensing at really really low speed which is normally important only to those using the speed PID controller. I have to have 3 to test firmware otherwise I wouldn't bother...ABritInNY said:
Since you didn't say what you are running, it's hard to tell the implications of upgrading - except to note that 3.1b22 is old - there is a 3.1 release out now and a 3.11 beta release in imminent. I would wait a few days for 3.11b2 (b1 got rejected at the last minute...). Only certain old versions can be upgraded directly using the new Setup Utility. You can try and if SU1.54 has a hissy fit, your best bet will be to just rebuild your config from scratch.ABritInNY said:
Do you happen to remember which beta version had the option to start from torque alone, without cadence input? (I've been reading thru this thread from the beginning as I have had time and concentration but haven't run across it yet, unless I just missed it; I'm not very far in yet).
EDIT: still haven't found it, but found a changelog here
https://endless-sphere.com/forums/viewtopic.php?f=4&t=37964&p=700677&#p700677
however it doesn't seem to list a feature like that; perhaps I'm looking for the wrong terminology?
I'd like downgrade my CA3 to that version to make my trike do what I need it to (activate the full motor power based on torque alone, because I can't pedal hard enough quickly enough, without the motor power, to get it going soon enough to get the motor power to start....).
I've been pondering simple electronics to do this, but it seems to get complicated more quickly than I'd like.
ATM, even setting the CA3.1 to it's most aggressive, it can't startup from a stop unless I use the throttle to do it. And with these settings it is much more like on/off PAS rather than having any real degree of control over the system.
(well, if I'm in the very lowest pedal gear, I can eventually get it to get going, but not until cars are honking behind me and getting mad cuz I didn't instantly get going out of their way--that's really the primary reason I have so much power on tap, just to very quickly get out of the way of traffic waiting behind me at a light, regardless of how much of a load I'm pulling).
Presently it has a TDCM bb sensor, but I still have a THUN I can put in there if I have to (though I would much rather use the one that reacts to either crank rather than the one that only reacts to one side).
EDIT: still haven't found it, but found a changelog here
https://endless-sphere.com/forums/viewtopic.php?f=4&t=37964&p=700677&#p700677
however it doesn't seem to list a feature like that; perhaps I'm looking for the wrong terminology?
I'd like downgrade my CA3 to that version to make my trike do what I need it to (activate the full motor power based on torque alone, because I can't pedal hard enough quickly enough, without the motor power, to get it going soon enough to get the motor power to start....).
I've been pondering simple electronics to do this, but it seems to get complicated more quickly than I'd like.
ATM, even setting the CA3.1 to it's most aggressive, it can't startup from a stop unless I use the throttle to do it. And with these settings it is much more like on/off PAS rather than having any real degree of control over the system.
(well, if I'm in the very lowest pedal gear, I can eventually get it to get going, but not until cars are honking behind me and getting mad cuz I didn't instantly get going out of their way--that's really the primary reason I have so much power on tap, just to very quickly get out of the way of traffic waiting behind me at a light, regardless of how much of a load I'm pulling).
Presently it has a TDCM bb sensor, but I still have a THUN I can put in there if I have to (though I would much rather use the one that reacts to either crank rather than the one that only reacts to one side).
Oh, ok--sorry. Somehow I'd got it into my head that it went into a beta but didn't get the bugs worked out to make it into release, from this:
https://endless-sphere.com/forums/viewtopic.php?f=4&t=37964&p=1362554#p1362554
My bad.
Were there any alpha versions with it?
Am I the only one that needs this feature? (probably)
https://endless-sphere.com/forums/viewtopic.php?f=4&t=37964&p=1362554#p1362554
My bad.
Were there any alpha versions with it?
Am I the only one that needs this feature? (probably)
The project never got beyond a preliminary alpha release (in-house test).
As Justin pointed out, a general implementation for all sensors with different torque characteristics is problematic. Attempting to address these issues with timer, etc to accommodate different torque lapses in various units led to unexpected power applications that turned common riding situations very startling and dangerous.
While this can probably be made to work for some particular well behaved sensor on a modestly powered bike, the CA3 is about general solutions without operational caveats to mitigate danger.
That said, since you like to build things, you might try putting a 555 clock permanently on the RPM input so the CA assist level will be gauged entirely by torque. I don't think this will ride super well, but it will eliminate the no-RPM issue. Since the CA assumes a baseline 50rpm as soon as it detects any pedaling at all, this operation will not be dissimilar from a pure torque-based getaway and never exceeding 50 rpm cadence.
As Justin pointed out, a general implementation for all sensors with different torque characteristics is problematic. Attempting to address these issues with timer, etc to accommodate different torque lapses in various units led to unexpected power applications that turned common riding situations very startling and dangerous.
While this can probably be made to work for some particular well behaved sensor on a modestly powered bike, the CA3 is about general solutions without operational caveats to mitigate danger.
That said, since you like to build things, you might try putting a 555 clock permanently on the RPM input so the CA assist level will be gauged entirely by torque. I don't think this will ride super well, but it will eliminate the no-RPM issue. Since the CA assumes a baseline 50rpm as soon as it detects any pedaling at all, this operation will not be dissimilar from a pure torque-based getaway and never exceeding 50 rpm cadence.
I'm not sure if this is practical but have you guys thought about having a sort of log step mode for the digital aux? It might just be my situation and not something that would be useful widely but in a torque sensor setup, small assist at the bottom end and larger assist at the top end seems like it would be helpful. For example, my wife wants minimum or no assist on flats but help on hills (possibly with the min setting being a very low background assist to compensate for the added weight of the motor and battery). The idea being that if you have pas max assist at 5.0 or something for very steep situations, you could have steps that start off at .25 but not have to go through so many steps to get max assist. Also as you get up higher in the assist levels if 3.0 isn't enough it would make sense for the next step to be 4.0 and not require going through multiple pushes. For example a set of steps might look like:
1 0.25
2 0.5
3 0.75
4 1.0
5 1.5
6 2.0
7 2.5
8 3.0
9 4.0
10 5.0
With the current software if I want 5.0 max multiplier and 0.25 steps it would take 20 steps to get there, and in a steep hill situation take a lot of pushes to get the boost up, at which point the rider might have lost forward momentum. I've thought about putting a throttle on the bike, but not sure she will be comfortable with that so I'm trying to work out the best balance of settings.
1 0.25
2 0.5
3 0.75
4 1.0
5 1.5
6 2.0
7 2.5
8 3.0
9 4.0
10 5.0
With the current software if I want 5.0 max multiplier and 0.25 steps it would take 20 steps to get there, and in a steep hill situation take a lot of pushes to get the boost up, at which point the rider might have lost forward momentum. I've thought about putting a throttle on the bike, but not sure she will be comfortable with that so I'm trying to work out the best balance of settings.
kevinscargobike
1 W
- Joined
- Nov 1, 2017
- Messages
- 64
I’d want a log scale option for the aux pot too. Precision is nice on the flats for normal riding with others, and not nearly as important when climbing a hill and it’s up at 20x or whatever.
Right now I have to use throttle on my hill because I can’t get both the top end range and the precision at the bottom.
Right now I have to use throttle on my hill because I can’t get both the top end range and the precision at the bottom.
vmbray said:
Hmmm - no, actually it never came up, but I'll write a ticket to get it in the queue for feature discussion. The CA PAS system is essentially torque-based (uses the Power PID) and so the situation is different than advancing the throttle to climb a hill using an ordinary controller with PWM-based throttle. This might need a proof-of-concept experiment....
FWIW: I have the digital Aux on PAS on one bike and have it set for 21 steps . I don't mind tapping the button until the bike does what I want (I seldom look at the display) and a long press goes to max right away.
For the case of a pot on the AuxA input, you could replace the present linear pot with an ordinary 5K audio log taper pot. Done.kevinscargobike said:
Since I posted I did notice that a long hold goes to max, that is actually pretty helpful. I think the other comment about precision vs max power sums it up very well. On the low end you want small steps so you’re at a happy level of work vs assist but on the high end if you’re already at 3x there’s no point going up to 3.25 because the change is so small relative to the level.
Actually just go ahead and do Alexa integration so the rider can say “moah powah!”
Actually just go ahead and do Alexa integration so the rider can say “moah powah!”
Actually, I often use the learning feature that lets you say "Alexa - play something I'd like."vmbray said:Actually just go ahead and do Alexa integration so the rider can say “moah powah!”
So maybe a better Elon-esque integration would allow "Alexa - help me ride the way I want!".
kevinscargobike
1 W
- Joined
- Nov 1, 2017
- Messages
- 64
teklektik said:
Is there a similar minimum assumed for torque?
I regularly find that the motor spins up when I'm just swapping which pedal is down for a turn, or when I'm preparing to come to a stop, even when I turn the pedals so slowly and gently that the freewheel doesn't engage.
I think it's more pronounced when I've used a lot of power recently (maybe some kind of smoothing / averaging?). As an example, I crest the top of my hill, and I've probably been using the (power) throttle, and I'm just changing which foot is down so that I can stop at my mailbox, and the bike jumps as if I'd stomped on the pedals, even though the assist multiplier is low, the pedal RPM was low, and almost no torque was applied.
—edit
Here’s a quick video I managed to get while riding home tonight. I pedaled with a moderate effort, paused, and then moved the pedals just a tiny bit (an inch or two?). Hwt (accurately) stays at zero, but W goes to 75.
https://www.dropbox.com/s/a0r8qewfcgliqlo/Video%20Jun%2005%2C%206%2004%2026%20PM.mov?dl=0
Sorry for the messed up orientation.
Raphael303
100 mW
- Joined
- May 31, 2016
- Messages
- 38
I have tried to update my CA3 v3.0x with 1.54
It connects but fails on reading/writing/updating
I can read and write with setup utility v1.31 OS high sierra and the Grintech TTL cable (I downloaded the newest USB Drivers)
It seems that my CA3 firmware is too old to work with 1.54.
Is there another way to update to 3.1?
It connects but fails on reading/writing/updating
I can read and write with setup utility v1.31 OS high sierra and the Grintech TTL cable (I downloaded the newest USB Drivers)
It seems that my CA3 firmware is too old to work with 1.54.
Is there another way to update to 3.1?
The core question here is whether it simply does not communicate or gives some other error?Raphael303 said:
You need to email Grin directly about this issue so the Setup Utility folks can get on board with your issue. The Mac-iness of this may at the root of the problem...
Yes - exclusively for torque. The Basic Power setting applies the rpm scaling only on rpms above 50rpm (e.g. at 40rpm there is to rpm component while at 60 rpm there is 10rpm scaling.kevinscargobike said:
kevinscargobike said:
Not sure what's going on there. I'll write a trouble ticket on it.
Raphael303
100 mW
- Joined
- May 31, 2016
- Messages
- 38
teklektik said:
I believe it can connect, but then fails. The older version can read and write so if it is maciness at work it is only in the new version an issue.
I have written to grin. They told me I have to select "reset to default settings".
I get the following messages then nothing, (and Grin has not responded to this yet
---------------
Failed to open temporary file
Copying Firmware to temporary file
Loading Firmware File
Please Cycle power to the device
-----------------
By the way your 2WD bike looks super nice!!
Looks to me like you may have a permissions problem and things are hanging because the temp file is never created or loaded with download data.Raphael303 said:I get the following messages then nothing, (and Grin has not responded to this yet
---------------
Failed to open temporary file
Copying Firmware to temporary file
Loading Firmware File
Please Cycle power to the device
-----------------
What did it say after you power-cycled the CA?
(Glad you like the bike - got about 15000 miles on it now
. These days I mostly ride a Paratrooper that I can fold into the rear of the car and ferry to more distant bike trails...)v3.11b2 Released
A beta of the 3.11 production release is now available via the CA Setup Utility. Press 'Get New Firmware' on the toolbar and select the 'Download beta firmware' checkbox. (And yes, there was a 3.11b1, but it never made it to release...)
There are different Help and Release Notes for each firmware version so you just need to get to a Setup Utility tab with this new version. You can get to the Release Notes by installing Setup Utility 1.54 if you don't already have it then updating your CA or selecting 'New Setup File' and choosing 3.11b2. Next select Help/Help and click on 'Release Notes' in the Nav Bar.
Beyond addressing some smaller display and documentation issues, 3.11b2 corrects a significant bug that causes settings to become corrupted. The corruption causes a variety of 'red herring' symptoms that appear to point to other issues. Very vexing.
A second important bug fix remedies a problem with wonky speed readings over about 45mph. This is not an issue for many, but it can confuse speed limiting, etc as the CA suddenly reports the speed to be one half the true speed.
On the 'Feature' side of the ledger, the PAS pedal detection logic has been reworked and gives a noticeable improvement to pedal motion detection. This gives brisker response regardless of the number of magnets (poles) in the PAS sensor or the particular CA PAS mode.
Due largely to repair of the setup corruption bug, we strongly recommend upgrading to this new version.
So - 'Download away!' and post up here with any issues you find.
As always, many thanks for the efforts and support of our loyal beta testers!
A beta of the 3.11 production release is now available via the CA Setup Utility. Press 'Get New Firmware' on the toolbar and select the 'Download beta firmware' checkbox. (And yes, there was a 3.11b1, but it never made it to release...)
There are different Help and Release Notes for each firmware version so you just need to get to a Setup Utility tab with this new version. You can get to the Release Notes by installing Setup Utility 1.54 if you don't already have it then updating your CA or selecting 'New Setup File' and choosing 3.11b2. Next select Help/Help and click on 'Release Notes' in the Nav Bar.
Beyond addressing some smaller display and documentation issues, 3.11b2 corrects a significant bug that causes settings to become corrupted. The corruption causes a variety of 'red herring' symptoms that appear to point to other issues. Very vexing.
A second important bug fix remedies a problem with wonky speed readings over about 45mph. This is not an issue for many, but it can confuse speed limiting, etc as the CA suddenly reports the speed to be one half the true speed.
On the 'Feature' side of the ledger, the PAS pedal detection logic has been reworked and gives a noticeable improvement to pedal motion detection. This gives brisker response regardless of the number of magnets (poles) in the PAS sensor or the particular CA PAS mode.
Due largely to repair of the setup corruption bug, we strongly recommend upgrading to this new version.
So - 'Download away!' and post up here with any issues you find.
As always, many thanks for the efforts and support of our loyal beta testers!
You can buy direct from Grin Tech at ebikes.ca but they are carried by OZ shops as well. Grin can certainly give you the names of some retailers. As always - shipping to the other side of the world is NG no matter how the part gets there...
Member Samd comes to mind at https://ballaratebikes.com/.
Also member Hyena at http://hyenaelectricbikes.com/.
Member Samd comes to mind at https://ballaratebikes.com/.
Also member Hyena at http://hyenaelectricbikes.com/.
redline2097
100 W
- Joined
- Jun 17, 2015
- Messages
- 166
I have weird problem with self wired Aux potentiometer. When turning the pot display brightness goes bright and beyond (text disappears).
It does the same even I disconnect Digital Aux Button. I tried dozen of deadbnd-settings but display does adjust every time so bright that text disappears. I'm running 20s lipo. Everything else working fine. Software version is CA3-1. There is not other devices connected to CA at the moment than Domino Throttle.
Type of potentiometer is b5k "Single Mono 5K linear". With multimeter I get linear range between 0 to ~5000ohm.
Video here:
https://www.youtube.com/watch?v=d762RJlrFk8
It does the same even I disconnect Digital Aux Button. I tried dozen of deadbnd-settings but display does adjust every time so bright that text disappears. I'm running 20s lipo. Everything else working fine. Software version is CA3-1. There is not other devices connected to CA at the moment than Domino Throttle.
Type of potentiometer is b5k "Single Mono 5K linear". With multimeter I get linear range between 0 to ~5000ohm.
Video here:
https://www.youtube.com/watch?v=d762RJlrFk8
The LED backlight is driven in series with the CA electronics and gets brighter with increased current draw.
This strongly suggests that your pot is wired improperly and going to the high setting is shorting out the +5V supply. Recheck your wiring - I'm pretty sure you have the wiper and one end interchanged.
This strongly suggests that your pot is wired improperly and going to the high setting is shorting out the +5V supply. Recheck your wiring - I'm pretty sure you have the wiper and one end interchanged.
Similar threads
- Question
