Cycle Analyst V3 preview and first beta release

Stealth_Rider said:
teklektik said:
See:
"6.10 Monitoring Charge Current"
"Appendix A. Calibrating the Cycle Analyst RShunt Value"
6.10.1 requires that the controller be modified whereas comparing Satiator AH in to CA AH used doesn't require modification. You do, however, need to have a Satiator :lol:
Where I was going was to merge the two hook-ups in those sections so that you could use the Satiator to do the calibration.

Instead of the normal business of adding the extra wire to the controller so the CA can monitor the charging current -- like so:

chargeDetailed2.png
You might instead temporarily hook up the charger like this just to calibrate the CA -- leave the battery connected normally, unplug the phase wires from the motor, and use one phase wire to connect to the charger:
(This is unsuitable for normal charging because of the altered termination voltage.)

chargerCalibration2.png
The only issue here is the Satiator current display which only goes to 0.1A. This means you are not going to get more than about 1-1.5% accuracy in your calibration because of the display precision. You can likely do better using the Ah technique because the Satiator is much more precise about Ah's that the available precision of the Amps display. But this is pretty quick and the results will give serviceable results.

That said, if you have a DMM with more digits of display precision, this technique should yield quite good results with any charger as a variation of the standard constant current calibration technique called out in the Guide Appendix.
 
justin_le said:
Stealth_Rider said:
I'll continue trying out my AH comparison method and see how rapidly I converge on a value of Rshunt that give approximate AH equality.

You don't need to converge Jim, you can get this spot on in your very first iteration.

RShunt(new) = RShunt(old)*[CA_Ah]/[Satiator_Ah]

Tweaking it up or down a bit at a time and repeating until it converges is just silly.

Justin,

Well the positive part of tweaking is I have to ride a lot more to fine tune, so that's a good thing :lol: However, I like spot on in 1 try, so I'll do it the corrective formula way, using the advice about DOD you provided. Thanks :D

Teklektik,

Thanks for explaining the hook up further to use the CA. Being one to take the easy way first I'll try the AH approach since I've got some shrink tubing around the phase wires and I'd rather leave them alone.

To both of you I much appreciate the advice, and, Teklektik, how to go from 3.01 to latest 3.1b12 (when available), using two installations of the CA Setup Utility for Windows, helps a lot.

Jim
 
Hello, I would like to download the v3.1 Beta 11 firmware. In the post by teklektik i get a message that the extension ZIP has been deactivated (see quote below). I tried to see if this is a local problem but get the same message on different devices and set-ups so I guess it must be a problem on the website. Is it possible to fix this?

teklektik said:
v3.1 Beta 11 Released (Repaired b10)

Thanks to Marc's bug report.....

.....Friday goodie...

Download!

(Previously implemented 3.1b10 beta features (part of this build) described here and in the Release Notes.)
 
v3.1 Beta 12 Released

Hey folks-
3.1b12 is ready to go with some interesting upgrades. Unfortunately, new features introduced changes to the EEPROM map, so a full 'from-scratch' config is required...

This release has bunch of small fixes but mostly focuses on features that simplify configuration and make issues easier to address without resorting to ES or Tech Support. Skipping some of the lesser changes, the biggies here are:
  • DigiAux support is migrating to use Schottky diodes
  • A new feature to restore (and save!) 'Factory Defaults'
  • Change to set the 'S' Limit flag on excessive DSGain
  • Redesign of the AuxA and AuxD menus
  • Provide a configurable minimum time for eBrake power removal

Here's a complete list of changes:
Code:
Release: 3.1b12                                                      2017-02-04

Note: V3.1b12 Setup files are **incompatible** with firmware or Setup files
      of **ALL** prior releases including any earlier 3.1 version.

 (1)  2430 - Temp Stats Screen 'Time Temp-Limiting' field displays even when 0
 (2)  2700 - Bikes w/o Speed monitoring can't activate features dependent on 'Moving'
 (3)  2566 - PAS TorqMode should not be selectable for non-torque PAS devices
 (4)  2834 - 'Low Volts' LVC cutoff message does not display properly
*(5)  1874 - (new) Provide means to restore all settings to (factory) defaults
 (6)  2020 - Y/N popup screens do not handle simultaneous button pushes
 (7)  2020 - Y/N popup screens do not allow [Y] for CAs with a single button
*(8)  2014 - Speed Limiting flag is not displayed for DSGain-induced limiting
 (9)  2086 - ThrIn 'Min/Max Input' parms suggest actual Throttle min/max values
*(10) 2020 - Convert battery Chemistry option 'LiMn' to 'Li-ion'
 (11) 2020 - Preset displays do not indicate the total number of presets
 (12) 2020 - AuxA 'Limit Change' Pop-up shows blank field for Speed=0
*(13) 2918 - (new) Change DigiAux circuit to use Schottky Diodes
*(14) 2918 - (new) Reconfigure AuxA/AuxD menus
 (15) 3396 - Precision of Per-cell LVC entry field in Setup Menu is too high
 (16) 2020 - ThrI Setup Preview Screen is incorrect for mode=Disabled(ZERO)
*(17) 3518 - (new) Add "watt-hours" as a custom display field option
 (18) 3502 - (new) Change Setup PAS Config Preview to show HW/W for Trq mode
 (19) 3594 - PAS bar graph is inoperative in 'PAS noTRQ' mode
 (20) 3554 - In AutoPAS mode the Aux Change PAS Assist Pop-up should show Watts
 (21) 3636 - Change PAS Sensor chooser option from 'PAS noTRQ' to 'Basic PAS'
 (22) 2566 - PAS output is generated in AutoPAS mode even if PAS is Disabled
 (23) 2020 - Difficult to differentiate initial change to temp alert icon
*(24) 1732 - (new) Provide a configurable minimum time for eBrake power removal

Here's a rundown of some of the more interesting items:

  • DigiAux support is migrating to use Schottky diodes

    Justin discussed this in an earlier post (DK-11 Digital AUX buttons see the light of day). There was a lot of hand-wringing about this change because of the loyal beta-testers who already built custom DigiAux switches, but in the end this change was determined to be in the best interests of the product. The new Grin switches based on the Wuxing DX-11 are pretty cool and also pretty inexpensive as these things go. There is a tiny pcb inside with SMD parts on it. See the post mentioned above - if you wired up a custom switch, you may just choose to hook up one of the little Wuxing PCBs in-line and slip a piece of shrink over it as a means of updating your rig. Otherwise, direct replacement with any low-power Schottky with a Vfwd of about 0.4V will do to replace your existing conventional diodes (e.g. 1N5711).

    To make this transition a bit easier, 3.1b12 is released in two versions: one for the older conventional diodes and one for the Schottky diodes. If you don't have or use DigiAux hardware, either version will work fine. Starting with the next beta release, the firmware will be released as a single version that expects Schottky diodes - conventional diodes will no longer be supported.

    An updated DigiAux document will be posted soon, but meanwhile, here's a summary of the new voltage thresholds using Schottys:

    SchottkyThresholds.png
  • Restoring all Settings to Factory Defaults

    Although the CA has always provided configuration of many operating parameters, there has never been a means to undo user configuration and return to the 'as delivered' configuration. This feature is now available and also allows presetting the 'Defaults' to the present user configuration. This second part of the feature is aimed at OEMs and kit retailer that sell the CA with a preconfigured setup for the kit or bike. Or course any user can now use this to save the complete configuration of all presets as a means to undo new configuration experiments or on bikes that are set up for non-techie friends and relatives...

    A new 'Cal->Defaults' item has been added to the Calibration category with a chooser menu to either:
    • Restore EEPROM settings to non-EEPROM defaults.
      All EEPROM settings including Calibration but excluding Lifetime Stats are overwritten and replaced with 'default' settings.

      View attachment 9
    • Save the present EEPROM settings to redefine a new set of 'defaults'.
      All settings including Calibration but excluding Lifetime Stats are saved. Previous 'Default' settings are overwritten and lost.

      CA3-1b12_Cal-SaveDefaults4.png

    Default settings reside in a reserved region of firmware memory that is initially clear when new firmware is flashed. EEPROM settings are automatically copied to this reserved region to define new default settings on the first boot after a firmware flash. This strategy sets the new 'defaults' differently according to the type of hex file used to flash firmware because of differing EEPROM contents after the firmware flash:
    • In the case of flashing a CA3-xxxx_firmware.hex file, EEPROM is freshly initialized with 'factory defaults' from the hex file so the 'defaults' are set to factory defaults.

    • In the case of flashing a CA3-xxxx_firmware_NoEeprom.hex file, EEPROM contains the existing user configuration so the 'defaults' are set to the existing user configuration *NOT* factory defaults. If you wish to preserve the existing defaults, first export the EEPROM settings to a file, then RESTORE the defaults prior to flashing the firmware. On the subsequent reboot, EEPROM with the old defaults will be saved as the new defaults. Finally, download your old EEPROM settings from file.
    To ensure data integrity, the CA will only restore default settings which were saved in their entirety without interruption. If an automatic or user- initiated save operation is interrupted before completion (e.g. power loss), a new automatic 'defaults' save will occur on the next program boot.

    A 'restore defaults' operation can be initiated either of two ways:
    • from the 'Cal->Defaults' Setup Menu, or
    • with a very long (6 sec) RIGHT button press from the main screen.
    The OEM flag option can be used to disable user access to the Setup menu Save/Restore parameter so the user only has easy access to the 'safe' option (b) above.

    View attachment 2

  • Speed Limiting 'S' Flag is Now Set for DSGain-Induced Limiting

    Although the CA has a handy display of 'Limit Flags' to assist in diagnosing problems, there has always been in gap in support in that speed limiting due to hard acceleration with an excessively high DSGain setting would not set the 'S' speed limiting indicator. This release extends detection to this type of limiting so that all cases of speed limiting now trigger the 'S' flag.

    CA3-1b12_SFlagForDSGain.png

  • New AuxA/AuxD Menus

    The Aux menus have been re-worked to eliminate some of the 'techiness' and secret incantations. As the images below show, the earlier catch-all voltage based configuration has been replaced with menu options for specific AuxA devices like pots and switches. These devices correspond to the on-the-shelf Grin add-ons and are also described in the [strike]Un[/strike]official Guide for DIYers. The pot configuration can be used to provide a configuration essentially identical to the old scheme if custom switches, etc are present that cannot be addressed with the new fixed options.

    View attachment 7
    As with the recent PAS menu changes, the CA menu operation hides unnecessary parameters, etc whereas the Setup Utility does not. Bogus or unnecessary values entered in the Setup Utility are validated and 'fixed' by the CA when downloaded so there are no ill effects.
    1. AuxA Menu

      To simplify configuration, setup parameters change according to device type as shown here:

      CA3-1b12_newParms.png
      • Limit percentages for 2 or 3 position switches are now set directly instead of the earlier lookup table or trial and error approach.

      • Pot configuration now uses 'deadband' settings instead of min/max voltages. The deadbands can be tweaked as needed for the upper and lower pot excursions, but standard Grin pots (where the pot voltage swings 0-5V) generally require no adjustment at all - just plug in a pot and go.
        • Additional techie details:
          • The top deadband is the difference between the highest pot voltage and 5V.
          • The lower deadband varies by configuration:
            It's the difference between the lowest pot voltage and one of:
            • 0.00V if there is no AuxD device
            • 0.86V if there is an AuxD device with Schottky diodes
            • 1.26V if there is an AuxD device with conventional diodes
          For a pot with output that swings rail to rail like the standard Grin pot, the deadbands can theoretically be 0V for all cases. However, there can be a little variation in diode forward voltage according to temperature or manufacturing, so the default settings are slightly higher than 0V and should work in pretty much all cases a, b, and c. Pots that don't swing rail to rail (across the whole voltage range) need a little more tweaking.

          A 'no meter' technique for adjusting these bands is to look at the Setup AuxA section preview screen with AuxA->Function set the desired limiting type (Amps, Spd, Pwr, Spd, or PAS) and adjust the pot to the extremes. The displayed limit percentage should swing from 0% to 99% and start and stop changing very close to the pot extremes. Adjust the respective deadband settings until this is the case.

          DeadbandAdjustment.png
    2. AuxD Menu

      The DigiAux menu is not changed too much except to separate the control type from the limit function. This is part of an effort to eliminate configuration interaction between AuxA and AuxD controls. The problem derives from the fact that AuxA voltages differ if the DigiAux circuit is present. In the prior implementation, adding AuxD buttons required reconfiguring an existing AuxA device. In 3.1b12, if DigiAux buttons are installed, the AuxD control must be set to '1-BTN' or '2-BTN' which allows the CA to adjust AuxA calculations without AuxA reconfiguration. If you want to disable the AuxD device but leave it installed, change the limit function to 'None'.

      The short form is that you can now plug in a new DigiAux button set, set the button type, and any existing standard Grin AuxA control will work as before without adjustment. DYI Aux controls may require a bit of AuxA tinkering...

  • Support for Middrive Shifting with eBrake

    With the proliferation of middrives, the need to momentarily relieve power for shifting has has become a recurring issue. This has generally been addressed with controllers or CAs by either tapping the ebrakes or adding some type of gear shift sensor. Particularly in the manual 'brake-tap' strategy, the timing of the power suppression is of primary importance. This firmware release adds a parameter to allow configuration of a 'minimum ebrake time' from 0.0 to 2.0 seconds in 0.1 second increments that is triggered by ebrake application. This allows setup of a specific no-power period that will always be present regardless of how quickly the brakes are tapped.

    CA3-1b12_EBk-MinTime.png
    The minimum delay operates in parallel with normal ebrake operation so that ebrake application in excess of the configured time results in normal operation as the timer just expires quietly with no visible effect. In most cases this eliminates brake release lag for normal braking while still providing the shift timer feature.

    minEbrakeTime.png

  • Misc Goodies...

    There are a few other small additions that are called out on the Release Notes. For instance, the old 'LiMn' chemistry has been renamed to a more generic 'Li-ion' battery type, some Setup parameter names have been revised to clarify meaning, and there's a minor tweak to the temp icon to make initial rising temperature easier to see (I think this was a mrbill request from long ago ;)).

More details excerpted from the packaged Release Notes:

Code:
Release: 3.1b12                                                      2017-02-04


Note: V3.1b12 Setup files are **incompatible** with firmware or Setup files
      of **ALL** prior releases including any earlier 3.1 version.

      ------------------------------------------------------------------------
      IMPORTANT:                               __________________
      Please install this package according to README_install.txt.
      Failure to do so may result in improper operation of the Setup Utility.
      ------------------------------------------------------------------------
      
      ------------------------------------------------------------------------
      IMPORTANT:
      After installing this zip package on a PC, the Setup utility will no
      longer be able to properly utilize 3.1 Setup files prepared for prior
      3.1 versions.  Prior 3.1 settings files will not display or edit properly
      and will download to the CA incorrectly resulting in improper CA
      operation. Similarly, setup files prepared with this release cannot be
      used with earlier CA 3.1 firmware versions. 3.0 Setup files are unaffected.
      ------------------------------------------------------------------------

      >> Any incompatible existing CA setup must be recorded and manually
         reconfigured over a fresh 'CA3-1b12_firmware.hex' installation.

      >> All existing CA Setups must either be:
	    1. exported manually from the CA console (at least the statistics)
	    2. OR read by the Setup Utility and preserved in screen snapshots
	       PRIOR to installing this package. Saving to a 3.1 Setup file and
	       reading the file after installing the zip package will not work.
	    3. OR read and displayed by a Setup Utility installation that has
	       not been updated with this release so that values can be manually
	       copied into a second Setup utility that has been updated.

      ------------------------------------------------------------------------
      IMPORTANT:
      Because a CA3-1b12_firmware.hex flash will load default settings, throttle
      voltage configuration may be incorrect. Treat this as an initial CA
      installation and place the bike on a stand or otherwise elevate the drive
      wheel to prevent a runaway when the flash completes and before a proper
      configuration can be restored.
      ------------------------------------------------------------------------


(1) 2430 - Temp Stats Screen 'Time Temp-Limiting' field displays even when 0

    Formerly, the rightmost field of the Temp Stats screen always alternated
    between 'Tavg' and 'Mins'.  For most users and riding situations, the motor
    is never temperature limited making the 'Mins' display unnecessary.

    The display has been revised to only alternate and display the 'Mins' field
    (total time temperature limiting) when the time is non-zero.
    
    Configuration data layout is unchanged; no existing CA configurations or
    settings files are affected.


(2) 2700 - Bikes w/o Speed monitoring can't activate features dependent on 'Moving'

    Display of 'Moving' vs 'Still' screens, accumulation of 'trip time', and
    accumulation of temperature averages are conditioned by detection of vehicle
    motion according to the Spd signal. Vehicles without this signal are never
    considered 'moving' which effectively deactivates these features. 

    'Movement' detection has been revised to be triggered by either activity of
    the Spd signal OR current greater than 1A (10A for HiRange shunt). The new
    current-based detection mode is disabled once a valid speed signal has been
    detected so that vehicles with working speed sensors preferentially use only
    sensor-based motion detection.

    NOTE: This change can result in slightly inaccurate average speed and trip
	  time if a bike without speed sensors is left standing with accessories
	  ON and drawing more than 1A since this will be considered 'moving'.
    
    Configuration data layout is unchanged; no existing CA configurations or
    settings files are affected.


(3) 2566 - PAS TorqMode should not be selectable for non-torque PAS devices

    It is possible to erroneously configure PAS for torque mode operation even
    though there is no installed torque device.

    PAS Mode selection has been revised to exclude the TorqMode option for
    non-torque devices. Additionally, changing to a non-torque device forces a
    previously configured TorqMode to PASOFF when either changing the device,
    changing to a preset with a conflict, or downloading a conflicting
    configuration from the Setup Utility.
    
    However, PAS Mode selection is not restricted nor automatically altered if
    there is no PAS device defined since there can be no actual operational
    conflict. This allows PAS to be deactivated by selecting NO DEVICE without
    adverse effect on the existing PAS Mode.

    Configuration data layout is unchanged; no existing CA configurations or
    settings files are affected.


(4) 2834 - 'Low Volts' LVC cutoff message does not display properly

    a) The 'Low Volts' screen message disappears after only brief display and
       the screen backlight instead pulses bright/dim. 

    b) When in 'Low Volts' mode there is no 'Entering Setup' message to prompt
       the user when the button can be released on the required long press.

    c) The 'Low Volts' message is not present when leaving SETUP if the
       condition persists.

    Display and button code has been revised to make the 'Low Volts' message
    persist, eliminate the backlight pulsing, force re-evaluation/re-display of
    the 'Low Volts' condition on exit of Setup mode, and add the missing button
    prompt.

    Configuration data layout is unchanged; no existing CA configurations or
    settings files are affected.


(5) 1874 - (new) Provide means to restore all settings to (factory) defaults

    Presently there is no means to clear all user configuration and restore
    EEPROM settings to (factory) defaults.

    A new 'Cal->Defaults' item has been added to the Calibration category with
    a chooser menu to either:

	1. Restore EEPROM settings to non-EEPROM defaults. All EEPROM settings
	   including Calibration but excluding Lifetime Stats are overwritten
	   and replaced with 'default' settings.

	2. Save the present EEPROM settings to redefine a new set of 'defaults'.
	   All settings including Calibration but excluding Lifetime Stats are
	   saved. Previous 'Default' settings are overwritten and lost.

    Default settings reside in a reserved region of firmware memory that is
    initially clear when new firmware is flashed. EEPROM settings are
    automatically copied to this reserved region to define new default settings
    on the first boot after a firmware flash.  This strategy sets the new
    'defaults' differently according to the type of hex file used to flash
    firmware because of the contents of EEPROM after the flash operation:

	a. In the case of flashing a CA3-xxxx_firmware.hex file, EEPROM is
	   freshly initialized with 'factory defaults' from the hex file so the
	   'defaults' are set to factory defaults.

	b. In the case of flashing a CA3-xxxx_firmware_NoEeprom.hex file, EEPROM
	   contains the existing user configuration so the 'defaults' are set to
	   the existing user configuration *NOT* factory defaults. It is not
	   possible to preserve factory defaults when flashing a 'NoEeprom' file.
	   
    To ensure data integrity, the CA will only restore default settings which
    were saved in their entirety without interruption.  If an automatic or user-
    initiated save operation is interrupted before completion (e.g. power loss),
    a new automatic 'defaults' save will occur on the next program boot.

    A 'restore defaults' operation can be initiated either of two ways:
    
	a. from the 'Cal->Defaults' Setup Menu, or
	
	b. with a very long (6 sec) RIGHT button press from the main screen.

    Configuration data layout is unchanged; no existing CA configurations or
    settings files are affected.


(6) 2020 - Y/N popup screens do not handle simultaneous button pushes

    The Y/N popups used for Reset and Restore operations select the operation on
    the first button push and ignore the second push. Also, subsequent release
    of two simultaneous button pushes causes the last released button to be
    handled by the normal status screen after the popup is dismissed.

    Behavior is revised so that simultaneous Y/N button pushes are interpreted
    as a 'No' response and popups are not dismissed until both buttons are
    released. For example, selecting a 'reset' operation can be canceled by
    simultaneously pressing 'No' before releasing the 'Yes' button.

    Configuration data layout is unchanged; no existing CA configurations or
    settings files are affected.


(7) 2020 - Y/N popup screens do not allow [Y] for CAs with a single button

    Historically some CAs have been configured to have a single button (the
    Right button).  Features of the CA are designed to allow operation with only
    the single Right button (e.g. entering Setup on boot via the Right instead
    of Left button). The recent y/n pop-ups for 'Reset' and 'Restore' allow
    single button CA's to only invoke the right [n] option.
    
    All Y/N popups have been reversed to place the [y] response on the right.
    
    Configuration data layout is unchanged; no existing CA configurations or
    settings files are affected.


(8) 2014 - Speed Limiting flag is not displayed for DSGain-induced limiting

    Some vehicles may experience cutouts during hard acceleration due to an
    excessively high DSGain setting. This particular speed limiting is not
    reflected in the Limit Flags display making the issue difficult to diagnose.

    The speed flag mechanism has been revised to include this case so that any
    speed-related limiting is reflected as an 'S' Limit Flag.
    
    Configuration data layout is unchanged; no existing CA configurations or
    settings files are affected.


(9) 2086 - ThrIn 'Min/Max Input' parms suggest actual Throttle min/max values

    The present 'Min/Max Input' nomenclature suggests that the actual throttle
    range should be specified instead of the required padded voltages that
    include high and low dead bands. This leads to a variety of Support issues.

    The nomenclature has been revised to 'Zero/Full/Fault Throttle Threshold'
    to suggest that there is some voltage discrimination intended for the
    settings rather than actual throttle min/max voltages.

    Configuration data layout is unchanged; no existing CA configurations or
    settings files are affected.


(10) 2020 - Convert battery Chemistry option 'LiMn' to 'Li-ion'

    Several common battery chemistries are not represented in the Battery
    Chemistry Setup options, however, they all share voltage and SOC
    characteristics with LiMn and all belong to the larger classification of
    Li-ion. Configuration of these other types presently requires selection of
    LiMn which is not obvious to many users.

    The battery chemistry option 'LiMn' has been renamed to 'Li-ion'.
    
    This is a text change only with no operational consequences.  Configuration
    data layout is unchanged; no existing CA configurations or settings files
    are affected.


(11) 2020 - Preset displays do not indicate the total number of presets

    The Preset popup, Setup PRESET preview screen, and Current Preset config
    screen display as '#n <name>' which does not reveal the max value for '#n'.

    Preset displays have been revised to the form '#n/m <name>' where m is the
    max configured preset (e.g. '#1/3 Medium').

    This is a display change with no configuration consequences.  Configuration
    data layout is unchanged; no existing CA configurations or settings files
    are affected.


(12) 2020 - AuxA 'Limit Change' Pop-up shows blank field for Speed=0

    The displayed speed field disappears in the AuxA 'Limit Change' pop-up when
    configured for speed limiting and speed = 0.

    Code has been revised to properly display '0'.

    This is a display change with no configuration consequences.  Configuration
    data layout is unchanged; no existing CA configurations or settings files
    are affected.


(13) 2918 - (new) Change DigiAux circuit to use Schottky Diodes

    The beta design for the DigiAux Circuit uses two conventional silicon diodes
    to extablish fixed thresholds for the digital button inputs.  Product
    considerations have led to a revision of these parts to Schottky diodes
    instead which lowers the button threshold voltages and expands the available
    analog range.

    Circuits constructed by early beta adopted must be upgraded to use schottky
    parts for compatibility.

    This is a hardware change with no configuration consequences.  Configuration
    data layout is unchanged; no existing CA configurations or settings files
    are affected.


(14) 2918 - (new) Reconfigure AuxA/AuxD menus

    (1) The classic 3-position switch configuration required guesswork or tables
    to configure desired limit percentages for Low and Med settings.

    (2) The Aux menu system presents configuration options that require
    reconfiguration of the analog device when the digital option is added
    because of a shift in the baseline analog voltage. This complicates
    configuration both for initial installation and DigiAux upgrades. In
    particular the 3-position switch configuration is non-obvious to configure
    for desired percentage levels and requires different tables than when
    configured without the DigiAux button assembly installed.
    
    (a) The menu structure has been revised to separate configuration of the
	particular installed controls (pot, 3-pos switch, etc) and the function
	the devices perform (Amps Limit, Presets, etc).
    
    (b) 'Standard' controls available off-shelf form Grin are called out as
        specific control options. Custom analog controls are supported by
	selecting one of the standard types.
    
    (c) Configuration of an installed pot has now presents parameters defining
	the high and low dead band voltages instead of the high and low absolute
	voltages so that adding/removing a DigiAux button assembly will not
	require reconfiguration of the pot.

    (d) Configuration of an installed pot is now in terms of the percentage
	limiting for the Low and Med settings with the High setting always
	assumed to be 100%. Configuration of the simple 2-pos switch
	installation is similarly supported.

    This revision affects EEPROM configuration.  Existing CA configurations,
    setup files, and prior firmware are incompatible.

    >> A fresh configuration must be instated after flashing the new firmware.


(15) 3396 - Precision of Per-cell LVC entry field in Setup Menu is too high

    The Per-Cell entry entry field allows entry to 1mv which suggests a
    measurement precision that the CA cannot actually achieve.
    
    The per-cell voltage LVC field width has been reduced by one digit reducing
    display/entry precision to 10mv.

    This is a user interface change with no configuration consequences.
    Configuration data layout is unchanged; no existing CA configurations or
    settings files are affected.


(16) 2020 - ThrI Setup Preview Screen is incorrect for mode=Disabled(ZERO)

    The ThrottleIn Setup Preview screen incorrectly displays the throttle
    percentage when mode = 'Disabled (ZERO)'. This percentage should be hidden
    as in the case of mode = 'Disabled (WOT)'.

    Preview screen display corrected to suppress the '%' field so both disabled
    modes have similar displays.

    This is a user interface change with no configuration consequences.
    Configuration data layout is unchanged; no existing CA configurations or
    settings files are affected.


(17) 3518 - (new) Add "watt-hours" as a custom display field option

    A 'Wh' custom display field is not available for users wishing to see an
    energy-based display (e.g. W, Wh, wh/mi, Wh/mi!).

    A new 'Wh' option has been added to the Pref->CustViews parameter screen.
    This value is the same value shown on the 'Energy Mileage' screen and shows
    only the Whrs the battery supplied without compensation for any regen
    contribution. The field shifts to 'kWh' over 9999Wh.

    This revision affects EEPROM configuration.  Existing CA configurations,
    setup files, and prior firmware are incompatible.

    >> A fresh configuration must be instated after flashing the new firmware.


(18) 3502 - (new) Change Setup PAS Config Preview to show HW/W for Trq mode

    The PAS Config preview screen shows the assist start watts which is not as
    meaningful as the W/HW assist gain.

    The PAS Config Preview screen has been revised to show PAS->AsstFctr in
    units of W/HW.

    This is a user interface change with no configuration consequences.
    Configuration data layout is unchanged; no existing CA configurations or
    settings files are affected.


(19) 3594 - PAS bar graph is inoperative in 'PAS noTRQ' mode

    When PAS->SensrType=PASnoTRQ the main screen Human bar graph does not show
    the scaled 0-120rpm rider cadence when PAS->PASMode is set to a non-torque
    setting.

    This issue arose when the PAS menus were revised in the recent beta release.
    The bar graph display has been corrected to properly display the rpm-scaled
    graph in PASnoTRQ mode and to suppress the bar graph animation/display if
    there is no PAS device installed.

    This is a user interface change with no configuration consequences.
    Configuration data layout is unchanged; no existing CA configurations or
    settings files are affected.


(20) 3554 - In AutoPAS mode the Aux Change PAS Assist Pop-up should show Watts

    The recently revised Aux Change pop-up shows the scaled PAS Assist as the
    Aux scaling percentage. This is not as useful as display of the resulting
    Aux-scaled Start Watts.

    The display has been restored to the earlier form showing scaled Start
    Watts.

    This is a user interface change with no configuration consequences.
    Configuration data layout is unchanged; no existing CA configurations or
    settings files are affected.


(21) 3636 - Change PAS Sensor chooser option from 'PAS noTRQ' to 'Basic PAS'

    The PAS noTRQ description may not be clear to non-technical users and
    does not reflect nomenclature in use by Grin Support personnel.

    The menu item has been renamed to "Basic PAS".

    This is a user interface change with no configuration consequences.
    Configuration data layout is unchanged; no existing CA configurations or
    settings files are affected.


(22) 2566 - PAS output is generated in AutoPAS mode even if PAS is Disabled

    Some default behavior predating the PAS menu organization allows PAS power
    to be developed in AutoPAS mode even though the PAS Device is set to
    'Disabled' (uninstalled),  An 'installed' condition should be required for
    any PAS operation.

    All logic including PAS power generation is now conditioned on configuration
    of an installed PAS device.

    This is an internal processing change with no configuration consequences.
    Configuration data layout is unchanged; no existing CA configurations or
    settings files are affected.


(23) 2020 - Difficult to differentiate initial change to temp alert icon

    Exceeding the temp threshold slowly fills the 'thermometer' icon pixel by
    pixel as temperature increases.  The first increase (2nd icon level) shows a
    single blinking pixel in the bulb which can be difficult to discern.
    
    The second icon has been revised to show two blinking pixels.

    This is a display change with no configuration consequences.  Configuration
    data layout is unchanged; no existing CA configurations or settings files
    are affected.


(24) 1732 - (new) Provide a configurable minimum time for eBrake power removal

    Middrives typically require that power is briefly removed while shifting.
    As with normal controller operation, the CA removes power when ebrakes are
    applied making the ebrake mechanism a convenient means to briefly remove
    power.
    
    CA ebrake operation has been enhanced to remove power for a preset minimum
    period even if ebrakes are released sooner. This minimum 'no power' period
    allows shifting by tapping the brakes or using a gear shift sensor.  Power
    is restored immediately if brakes are released after they have been applied
    longer than the preset period. The period begins with ebrake application and
    so will run from the last tap of a sequence of short taps.
    
    The ebrake period is configurable 0.0 to 2.0 seconds to allow deactivating
    the feature (0.0) or tuning operation for a particular drive installation.
    
    This revision affects EEPROM configuration.  Existing CA configurations,
    setup files, and prior firmware are incompatible.

    >> A fresh configuration must be instated after flashing the new firmware.

So, pick you poison below for either conventional or Schottky DigiAux diodes (either is okay if you have no DigiAux buttons) and give it a whirl.
As always, post up any thoughts or issues! :D

Download!
< CA3-1b12_Conventional.zip - withdrawn - see 3.1b13 >
< CA3-1b12_Schottky.zip - withdrawn - see 3.1b13 >

(Previously implemented 3.1b11 beta features (part of this build) described here and in the Release Notes.)
(Newer 3.1b13 beta release available here!)
 
v3.1 Beta 13 Released (Repaired b12)

Rats.
Fixing that last bug that postponed the b12 release also broke the Save/Restore feature, so we're having another 'Take 2' episode...

This b13 release is just a 'fixed b12' to address this single issue. There are no other changes.
Apologies to the early beta b12 downloaders --- but on the upside, you can just load the b13 NoEeprom version over b12...

So, a little Super Bowl pre-game excercise:

Download!
View attachment 1View attachment CA3-1b13_Schottky.zip
(Previously implemented 3.1b12 beta features (part of this build) described here and in the Release Notes.)
(Newer 3.1b14 beta release available here!)
 
Cool! Thanks!

My I get a wish?
Is it possible to enable changing the brightness of the CA screen backlight in setup?

Granted, I ride an unusual bike (a long wheelbase recumbent), but the CA sits just in front of my face and the CA backlight is a bit too bright for comfort when commuting in the dark.
 
Thanks for the release.

I got b12 loaded last night and rode extensively with it today. Tomorrow I will update to b13. I did not test the new default settings function, so I assume that my observations below apply to b13 as well.

teklektik said:
Pot configuration now uses 'deadband' settings instead of min/max voltages. The deadbands can be tweaked as needed for the upper and lower pot excursions, but generally require no adjustment at all - just plug in a pot and go.

It took me some trial and error to figure out how this works: the "low threshold "dead band" is the voltage difference between 0.0 and the minimum voltage of the connected pot. The "high threshold dead band" is the voltage difference between 5.0 and the maximum voltage of the connected pot.

My potentiometer gives a range of 1.15 to 3.95 volts. Using the default values with my potentiometer results in a usable range between 20 and 80 percent. To enjoy a potentiometer range of 0 to 100 percent I had to update the thresholds to 1.15 and 1.05 for the low and high thresholds respectively.

There are a few other small additions that are called out on the Release Notes. For instance, there is now a more generic 'Li-ion' battery type

My batteries are the Samsung INR18650-29E with cathode chemistry, LiNiCoMnO2.

I used the "new" Li-ion chemistry setting, but I find that it gives a consistently pessimistic capacity assessment, showing a half-charge when I know the battery has been charged to about 75%, and shows only 1 dot (6-7%) when the battery has at least 20% charge remaining. Does the battery capacity gauge require a "learning" period, like some smartphone battery indicators?

[...] and a minor tweak to the temp icon to make initial rising temperature easier to see (I think this was a mrbill request from long ago ;)).

(23) 2020 - Difficult to differentiate initial change to temp alert icon

Exceeding the temp threshold slowly fills the 'thermometer' icon pixel by
pixel as temperature increases. The first increase (2nd icon level) shows a
single blinking pixel in the bulb which can be difficult to discern.

The second icon has been revised to show two blinking pixels.

This is a display change with no configuration consequences. Configuration
data layout is unchanged; no existing CA configurations or settings files
are affected.

Ambient temperature today was 15C or lower, and conditions were at times wet. My highest temperature reading on the motor was 87.5C. I won't be able to test this until the weather warms. This year, unlike the last two years' "summer in winter", we're having a Pacific Northwest winter in Northern California. But, thanks for addressing my complaint from last summer.
 
Marc S. said:
Is it possible to enable changing the brightness of the CA screen backlight in setup?
AFAIK , it's not software controlled at all (not connected to the MCU); you'd have to open it up and change the resistance to the LED itself, to change the current and thus it's brightness. You could add a potentiometer if you want it adjustable on the fly, or you could put a switch that shorts out one of two series resistors to get two brightness levels
 
mrbill said:
teklektik said:
Pot configuration now uses 'deadband' settings instead of min/max voltages.
It took me some trial and error to figure out how this works: the "low threshold "dead band" is the voltage difference between 0.0 and the minimum voltage of the connected pot. The "high threshold dead band" is the voltage difference between 5.0 and the maximum voltage of the connected pot.

My potentiometer gives a range of 1.15 to 3.95 volts. Using the default values with my potentiometer results in a usable range between 20 and 80 percent. To enjoy a potentiometer range of 0 to 100 percent I had to update the thresholds to 1.15 and 1.05 for the low and high thresholds respectively.
Yep you got me.
This is a case of 'exhaustive documentation' - I documented until I got tired... :D

You are exactly right about the 'deadband' terminology.
Here's some additional details:

  • The top deadband is the difference between the highest pot voltage and 5V.
  • The lower deadband varies by configuration:
    It's the difference between the lowest pot voltage and one of:
    • 0.00V if there is no AuxD device
    • 0.86V if there is a Shottky AuxD device with Schottky diodes
    • 1.26V if there is an AuxD device with conventional diodes
For a pot with output that swings rail to rail like the standard Grin pot, the deadbands can typically be 0V for all cases. There can be a little variation according to temperature, so they default to something slightly higher than 0V that should work in pretty much all cases a, b, and c. Your pot is a little unusual in that it doesn't swing rail to rail (across the whole voltage range) making it need a little more tweaking. With the various voltage requirements tied to the DixiAux feature, this is a slippery user interface to make easy and obvious, so frankly we went with something that is essentially plug and play for customers with off-the-shelf controls, but that can still work for DIYers who have less standard custom setups and are more techie by their very nature.

A 'no measurement' technique for adjusting these bands is to look at the Setup AuxA section preview screen with AuxA->Function set the desired limiting type (Amps, Spd, Pwr, Spd, or PAS) and adjust the pot to the extremes. The displayed limit percentage should swing from 0% to 99% and start and stop changing very close to the pot extremes. Adjust the deadband settings until this is the case.

EDIT - This extra config info has been added to the release post above...

mrbill said:
There are a few other small additions that are called out on the Release Notes. For instance, there is now a more generic 'Li-ion' battery type
My batteries are the Samsung INR18650-29E with cathode chemistry, LiNiCoMnO2.

I used the "new" Li-ion chemistry setting, but I find that it gives a consistently pessimistic capacity assessment, showing a half-charge when I know the battery has been charged to about 75%, and shows only 1 dot (6-7%) when the battery has at least 20% charge remaining. Does the battery capacity gauge require a "learning" period, like some smartphone battery indicators?
I was a little imprecise above and should have explained this as do the actual Release Notes:
"The battery chemistry option 'LiMn' has been renamed to 'Li-ion'."

This is just the previous LiMn setting which should be appropriate for most Li-ion chemistries. We felt that the LiMn designation was too restrictive and confused folks who couldn't find their particular chemistry.

So - this should be working just as LiMn did before. All but the LiFePo4 work the same and use both Ah and open circuit voltage - no learning involved. So - what were you using before and did that have this same behavior? If it gave more satisfying results, I would just recommend swapping back to that setting.

mrbill said:
[...] and a minor tweak to the temp icon to make initial rising temperature easier to see...
Ambient temperature today was 15C or lower, ... My highest temperature reading on the motor was 87.5C. I won't be able to test this until the weather warms
Yep, I have the same problem and had to re-adjust my temp limit to force a reading on the bike... I confess that the bulk of the examination came on the bench where things were warm and cozy and I just had to twist the 'temperature' knob on the test rig... :)

As always - thanks for the feedback and thoughts - appreciated. :D
 
Marc S. said:
Is it possible to enable changing the brightness of the CA screen backlight in setup?
Nope, not adjustable by firmware/configuration.

The LCD backlight is actually part of the CA power regulator circuit - this is why it changes brightness a bit when you press buttons, etc. There is no external current limiting resistor but there is - as I recollect - a white/red jumper on the back of the LCD module which might resolve your issue (normally strapped to white). I haven't tried to adjust the LCD myself and although I can speculate on some likely strategies, there's really no point. We need to get Justin to weigh in for an authoritative answer and specific part values/advice. We'll get back straightaway...
 
teklektik said:
Stealth_Rider said:
I was thinking that based on Justin's post the 3.1 release was not a beta any more. Are we still working betas on 3.1?

I'm at 3.01 now so when I install 3.1b12 do I need to do anything special? I have written down my settings for gains, ramps, throttle in, throttle out, etc. so if I have to overwrite to go to 3.1b12 its no problem because I can put them back in by hand or using the Setup Utility.
Although we had hoped to have a formal 3.1 production release out by now, things don't always go as planned. At this point we expect that b13 will likely be the final beta before 3.1 is formalized. So - soon - just not yet.

There are a couple of README files in each package that discuss the install procedure, etc. There is no doubt that you will have to reconfigure the CA from scratch - the 3.01 firmware is miles away from the 3.1 beta versions.

The Setup Utility can differentiate firmware versions by the major '3.01' or '3.1' designation - it does not understand the various beta designations. So - although you can open a 3.01 and a second 3.1 at the same time (either in different tabs or in different executions of the same Setup installation), it is not possible to open a 31b10 and 3.1b12 at the same time - once the 3.1b12 package is installed, the utility considers all 3.1 configurations to have the b12 format unless you reinstall an earlier package (e.g. b10).

I need to rebuild configurations all the time and have found the easiest means is to make two installations of the Setup Utility in different locations. Install a package for your existing build on one installation and install a package for the new firmware on the other. Open Setup from both locations at the same time. This will allow you to pull your CA config into one SetupUtil and manually transfer settings to the other 'new' config - not very automatic, but easier than transcribing a bunch of settings when using only a single installation.

Teklektik,

I have upgraded my CAV3 on two bikes from 3.01 to 3.1b13(well, going from 3.1b12 and then to 3.1b13 :lol: ). Your suggestion of using two instances of the Setup Utility was very helpful in transferring settings from old to new. I placed the two Setup Utility screens side by side on the monitor and comparison/transfer was easy.

So far everything is working fine for both bikes and I have just completed my first Rshunt calibration using the Satiator and CAV3 Ah readings. Seemed to work very well and it's easy to track AH used and put back in after each ride so I've got an easy way to check the correction factor:

RShunt(new) = RShunt(old)*[CA_Ah]/[Satiator_Ah]

Now that I've got a good value for RShunt(new), and a convenient method to check it, I can tune other settings in a more reproducible manner knowing the CA display values are what is actually happening with the bike.

Thanks.

Jim
 
teklektik said:
Stealth_Rider said:
Are we close to seeing V3.1? I would like to use the 3 position switch configuration improvements. ETA?
3.1b12 should be out in the next few days - it's back in alpha test just now. We had an annoying bug show up a week ago that postponed things a bit.

Meanwhile, here's a teaser :D




????


SO do we have a second Aux input...?? is this a hardware change? or pads that are already there...or something I have forgotten about?
 
NeilP said:
SO do we have a second Aux input...?? is this a hardware change? or pads that are already there...or something I have forgotten about?
Hey Neil-
Good to hear from you! Well - what we have is a means to add some UP/DOWN buttons to the existing AuxPOT input so that there is a logical second Aux channel without actually changing the CA PCB. You can either build your own or buy a button set from Grin (really inexpensive). The buttons work either alone or with one of the existing analog devices (pot or switch) - so two Aux controls for different limiting/preset selection with just the one old Aux PCB input..

There's a doc (that will updated soon) explaining all this on the download page and a post back for 3.1b4 where the feature was introduced. This 3.1b13 release sort of tunes up the menus a bit more to make the whole Aux thing easier to use.

FWIW: Each of the beta release posts has forward/backward links. You might start with the latest b13 post and just link back to catch the highlights of what's been happening (viewing them in order first to last is the best plan). The download page always has a link to the latest beta release. This linked 'evolutionary release' tale is less overhead for us as the features go through development and everything can be transferred into the Guide at once when 3.1 goes to Production Release. All changes to date since 3.0 are in the Release Notes file of each zip package. The posts have the pretty much the same stuff but with pictures and more explanations....
 
Thanks, loads of new tabs open to start a bit of reading

first time i have looked back here in a long while. Ride the Harley now more than the e-bike...or take the car.
 
teklektik said:
Marc S. said:
Is it possible to enable changing the brightness of the CA screen backlight in setup?
...the CA backlight is a bit too bright for comfort when commuting in the dark.
Nope, not adjustable by firmware/configuration.
The LCD backlight is actually part of the CA power regulator circuit - this is why it changes brightness a bit when you press buttons, etc.
Here's some details:
There are two ways to go after your 'night running' request:

  1. There's a set of pads on the back of the LCD that are normally bridged to the 'W' or White pad. This illuminates the white LEDs. However, you can move the solder bridge to the 'R' or Red pad instead to illuminate the red LEDs. This will give you the familiar night-vision-preserving red illumination.

  2. As mentioned previously, the LCD module is part of the power regulator circuit (see below). As you can see, changing the current draw for the CA and accessories directly affects the current through the LCD and so its brightness. Justin points out that the brightness can be reduced by bypassing the LCD with a resistor. This will send some of the current around the LCD instead of through the backlight. The exact value will depend on your installation. There's about a 3.5V drop across the LCD module due to the backlight LEDs and the basic CA alone draws about 10ma - other accessories add to that current. A resistor of about 330ohm should completely extinguish the LEDs and about 2K or more should give near full brightness. You can trial and error a few resistors to get the desired effect or you might use a 2K pot or a 2K pot in series with a 330ohm resistor to make the brightness adjustable.

    CA_PowerRegAndLcdModule4.png

Hookup points are as shown (works for either CA2 or CA3 models):

CA_LcdColorBrightnessMods2.jpg

EDIT: See >> this post << for notes on brightening the backlight.
 
Whow!

The CA screen got red LEDs as well?
Adding wires and a remote switch to the pads, looks like a great way to change between a red 'night ride' and a green 'daylight' setting for the CA.

Adjustable brightness on top of that, would be the icing on the cake. :)

Thanks for this thourough explanation!

Cheers!
 
Marc S. said:
The CA screen got red LEDs as well?
Adding wires and a remote switch to the pads, looks like a great way to change between a red 'night ride' and a green 'daylight' setting for the CA.
Be aware that you are heading into uncharted waters switching the LEDs on the fly in lieu of using the permanent solder bridge technique.

The processor runs off the +5V supply and switching the colored LEDs will briefly interrupt the +10V power to the LCD and 5V regulator. There's a 150uf cap on the 5V regulator input and the switch transition should be brief, so there may be no effect, but there is a possibility that the processor will reboot when you flip the switch. The effect would be mitigated if the brightness pot was set 'darker' so the bypass current would tend to hold up the regulator cap voltage. Your strategy is not electrically 'Bad' in any way, but the exact behavior is a matter of specific situation and speculation - so just a heads up... :D

Hearing of your experiences would be interesting.
 
Or use a Make Before Break type of switch, so that it does not disconnect the first LED before connecting the other. Then there is no circuit interruption.
 
Ok, good to know. Thanks for the heads up!

Since I'm currently down to one running commuter bike, I'll postpone the 'CA open hard surgery' ;) when the others are up and running again anyway.
 
teklektik said:
Marc S. said:
The CA screen got red LEDs as well?
Adding wires and a remote switch to the pads, looks like a great way to change between a red 'night ride' and a green 'daylight' setting for the CA.
Be aware that you are heading into uncharted waters switching the LEDs on the fly in lieu of using the permanent solder bridge technique.
Nope, it'll be fine. What you actually want to do is leave the white LED jumper soldered together, then hook up two wires to a simple toggle switch or button which will short out the red LED pads. When the switch is closed, both the red and the white LED's are in parallel, but the red LED's have a lower forwards voltage drop so they'll take all the current while the white LED's won't have anything flowing through them. When you short those two wires with a switch, the backlight will go red, and when you open the switch it will revert back to white.

Red LED Backlight Wiring.jpg

That would actually be really cool to see a CA set up like this with a simple day/night selector switch so I encourage you to open it and perform the simple wiring mod. There is room inside the CA enclosure for the body of a small toggle switch so it would stay self contained.

Marc S. said:
The CA screen got red LEDs as well?

We've had this features for over 6 years now but forgot to document it elsewhere on our website other than the original announcement, so I'm glad to see it finally (re)discovered!
http://www.ebikes.ca/news/red-backlight/
View attachment 1
Pretty much everyone who has a large screen CA device, either V2 or V3, can do this mod for night riding.

Now what we are working on is trying to get a VA LCD module with a white on black pixel logic to still look good at the 10mA CA current, so the future screens might look something more like this:
VA LCD CA Example.jpg
 
Justin, thats a very nice feature build in the CAv3 from the get go!

I'll probably just change the 2 lead cable for the speedo pick-up of my CAv3 DPS to a 4 lead cable and hook-up a switch on it, to change screen brightness on the fly.
 
IIRC, there are six cable exit holes on the back of the CA at the mount, so if one is unused at present you could use just a new 2wire cable and not have to replace an existing one. ;)
 
I noticed weird issue after updating CA3-1b13_Conventional firmware. When I power on bike and twist throttle first time and let go, motor keeps running two seconds. Even if I try to slow motor with gently braking, it pulls about 50 watts power. This happen only once after power is turned on, after that all is normal... Untill I turn power off and wait a while, like 10 seconds, then power on and runaway appear again.

I made clean update with CA3-1b13_firmware.hex
 
Back
Top