Cycle Analyst V3 preview and first beta release

[Trailblazer]

Yep - you should have little difficulty getting that to work. It's a little harder to program an Infineon with a CA cable, but also doable...

I will PM you with the older downloader and you can give it a test drive...

EDIT - Oops - I see izeman PMed you already while I was making coffee. Good deal.

That'll teach you to take a coffee break.

Thanks so much teklektik and izeman for your help! You guys are the beez kneez! Justin is going to send me a cable. I shortened this frankinbeast I have and still no go on with the v1 or v1.1 of the flashing utility. Justin says it's very hit and miss with the Prolific chipset. Apparently I've got a miss on my hands.

 

[teklektik]

1. Is it the UpRAmp that will give me less of a lag time for throttle response to the wheel? I could only change the UpRamp setting it a tiny bit from .05 to .03 so I dont see much difference if any after changing the settings to this new number.

First - I'm assuming you are using P5 not P6. UpRamp became UpRate in P6. The manual needs an update to P6 - coming soon. As noted in recent posts, P6 uses the inverse value for configuration, which is actually nicer and incidentally gives better adjustment range due to implementation considerations.

The ramp units in P5 are sec/v, so for an throttle inout range of 3v it will take all of 3v x 0.05sec/v = .15 sec to ramp from zero to WOT - essentially instantaneous. Your change from 0.05 to 0.03 will not yield a ramp time change discernible by humans....

2.Is it correct that a lower number setting quickens the application of power?

It increases the allowable rate at which power can increase. The power is initially applied instantly without delay in all cases.

3. Maybe this ramping has more to do with the power curve once the power has reached the wheel and not so much with the lag time I am struggling with. For me its not so much that it takes and smoothes out the power at start ups but the speed at which it first reaches the wheel. I am fine to dish out the smoothness of the power on my own but I dont like this lag time before any power at all reach the wheel. Any ideas on this?

The ramp values set a limit on how fast Throttle OUT can change. As long as Throttle OUT changes more slowly than the configured value, the ramp setting has no effect. If the output tries to change faster than the ramp setting, the CA slows the change to comply. This does not actually introduce any delays in the Throttle OUT change although you may perceive a delay on startup if it takes a bit of time to ramp up to a throttle level that will actually make the bike begin to move.

The other big player is the relevant gain setting if you are using a closed loop throttle mode. If you are using Current Throttle and AGain is too low, then power may increase/decrease more slowly as the responsiveness of the whole control system becomes more damped. Since you have UpRamp set as low as possible, I'm thinking your gain may be set low (like 10 or so).

Since it seems your V3 was somewhat pre-configured for your MAC (mid-drive), the gain and ramping were probably set to relieve the strain on the gears and clutch on getaway. You can download the P5 Setup Parameters doc to check what the real defaults should be. However, keeping the gain low also prevents surging so you have a couple of parameters to jiggle to balance several considerations for responsiveness, surging, and drivetrain-friendliness. Everything is a trade-off and you can't have it all.....

4. Any ideas why the setting I entered like .02 or .01 would get reset to .03?

This is a representation artifact of the limited number of bits available to store the value and the minimum allowable fixed point binary value. '0.03' is the decimal representation of the smallest value allowed. You will find similar behavior for many of the Setup parameters.

Hmmm... I guess that behavior probably deserves mention in the Guide...

 

[John Bozi]

Could someone please tell me how to connect the 10k thermistor? I think this should be 5V and sense wire to the two pins of the thermistor.

Connect the thermistor across the two pins of the thermistor connector as shown on the Grin Tech V3 page. These are as noted - gnd (black) and sense (yellow). If you are doing custom wiring, use the two thermistor pads on the CA PCB (NTC and G).

hmmm - another answer while typing, but here you go anyway...

I bought a crystalyte 4065 with a white single wire coming out of the motor connected to thermistor. Being a newb I don't know how to connect this to the CA V3
I am worried the hardware of the V3 is not the same as the one I thought I was getting.

From what I read before buying: http://www.ebikes.ca/store/store_CAV3.php

Connector Pinout
The V3 CA device has a cable bundle bringing out all the signal wires into suitably terminated JST-SM plugs. The following shows the standard CA V3 wiring harness connector details for all cables coming out of the V3 CA:

http://www.ebikes.ca/drainbrain/CA3_Connectors.gif

I don't recall seeing these connectors anywhere in the box.
Forgive my newbieness, is this connector pinout a separate product?

I didn't buy from ebikes.ca.

 

[izeman]

... I don't recall seeing these connectors anywhere in the box.
Forgive my newbieness, is this connector pinout a separate product?

I didn't buy from ebikes.ca.

as you didn't buy from justin we can't tell you what the reseller did to your product. normally the CA comes with all these plugs installed.
maybe post some pictures of the CA board. 4 screws to open, and we can tell you if you've got a v2 or v3.

 

[izeman]

... Justin says it's very hit and miss with the Prolific chipset. Apparently I've got a miss on my hands.

yes. it's a NIGHTMARE. i tried 3 different chipsets, and none worked. even those that's should have the correct voltage levels. your best idea was to order it from justin directly.

 

[Trailblazer]

yes. it's a NIGHTMARE. i tried 3 different chipsets, and none worked. even those that's should have the correct voltage levels. your best idea was to order it from justin directly.

Exactly what I did. I'm used to routers and such for firmware upgrades. Give me a good ole' ethernet connection any day...it just works! Thanks again for your help, izeman.

 

[John Bozi]

... I don't recall seeing these connectors anywhere in the box.
Forgive my newbieness, is this connector pinout a separate product?

I didn't buy from ebikes.ca.

as you didn't buy from justin we can't tell you what the reseller did to your product. normally the CA comes with all these plugs installed.
maybe post some pictures of the CA board. 4 screws to open, and we can tell you if you've got a v2 or v3.

a v2 upgraded to v3 - I will get a refund and buy directly from ebikes.ca thanks.

 

[hjns]

... I don't recall seeing these connectors anywhere in the box.
Forgive my newbieness, is this connector pinout a separate product?

I didn't buy from ebikes.ca.

as you didn't buy from justin we can't tell you what the reseller did to your product. normally the CA comes with all these plugs installed.
maybe post some pictures of the CA board. 4 screws to open, and we can tell you if you've got a v2 or v3.

a v2 upgraded to v3 - I will get a refund and buy directly from ebikes.ca thanks.

Interesting. I thought that it was actually not possible to upgrade a V2 to V3 due to the differences in components on the PCB.

 

[izeman]

... a v2 upgraded to v3 - I will get a refund and buy directly from ebikes.ca thanks.

Interesting. I thought that it was actually not possible to upgrade a V2 to V3 due to the differences in components on the PCB.

afaik the board IS the difference between them. there is no sense in "upgrading" a v2, because you would have to change the brain

 

[izeman]

one more issue: i found it very hard to configure throttle dead zones and PAS at the same time.
i followed the manual, checking lowest voltage, highest voltage of the throttle output, enter values 0.1v above and below, checked voltages at which the motor starts to spin and stops accelerating, enter those values 0.1v above and below, and then i wondered why PAS was not working properly.
the dead zone was still a bit to large for me, so i played a bit with it. when it was perfect, PAS was not working anymore. problem seems to be, that the throttle overrides PAS, so if the CA thinks it sees throttle input, PAS is overwritten. but there is no throttle input, it's only a very small dead throttle zone.
can you understand what it try to describe? it's hard for me to understand, so even more to write it down
point is: to my understanding: no dead zones and PAS does not work together. or maybe i did not calibrate the throttle correctly.
i will investigate further ...

EDIT: ok. maybe i should have read all of the manual: "Note: The recommended 0.1v throttle adjustment offsets may be reduced to minimize throttle dead zones,
but motor creep, maximum power reduction, or auto-cruise/PAS problems may arise. Reducing the offsets is
not recommended, but if undertaken, should be accompanied with a policy of regular throttle setting
verification."

 

[teklektik]

i found it very hard to configure throttle dead zones and PAS at the same time.
...
point is: to my understanding: no dead zones and PAS does not work together.
or maybe i did not calibrate the throttle correctly.
i will investigate further ...

This phenomena has been reported in this thread before and appears as a note in section 3.7 of the Guide. Your initial calibration with the 0.1v guard zones was probably correct, but the subsequent fine tuning to eliminate the dead zones left the CA unable to reliably detect ZERO throttle - exactly as you describe.

Auto-cruise uses the same ZERO throttle detection logic and is similarly affected by trimming dead zones too closely.

Your understanding is correct.

EDIT - Okay - I see you located and added the relevant Guide reference to the post above while I was typing... spot on!

 

[waynebergman]

2.Is it correct that a lower number setting quickens the application of power?

It increases the allowable rate at which power can increase. The power is initially applied instantly without delay in all cases.

Yes I see that you are correct and thank you for this by teklektik .It's weird this is my first mid drive and I have noticed a few things out in the driveway this evening just watching all the FW's and interaction from pedal assist and the motor drive sprocket to the rear cassette in the wheel. What I feel is actually happening is not lag in throttle response at all but rather the pawls not instantly in gauging in the rear cassette when the motor sprocket tries to power up the wheel. I am not sure about this but I will post a question about this in the mid drive area.

 

[izeman]

i may have another request?

.) i use a precharge to avoid sparks when connecting the battery. every time i turn the switch off voltage drops slowly and at 11V the controller shuts down - if it happens to see this voltage. it takes around a second for the voltage to drop to zero. so if i do a break, turn off the controller, then switch it on again, and during the ride want to know what the minimum voltage was, it shows me something between 30-50v - the last voltage the CA saw before voltage was totally gone. so this value is useless, as the battery still is at 70v+ and never was below that.
maybe there could be some watchdog that recognizes super fast voltage drop and doesn't take those values into consideration for Vmin?

.) other request, now that i have my pas installed, and got it working: i use pas together with "torq pas" and set 25Nm/V regarding to 7.4 in the manual. but 700w sometimes is too much, and this makes it hard to pedal slower, maybe to make it possible for a non-ebiker to pedal along side. so if i set it to 15Nm/V or 400W, this is too little to go above 25km/h. even when pedalling very fast.
so i would ask for a steeper curve rpm->watt output. this way you can pedal very slowly (55rpm) and ride slowly, and when you need to climb the hill, or want to go faster you just do 90rpm.
i know another way could be to add a 3-way switch to change support level, but my handle bar is already quite crowded, and it would be nice to solve it by CA programming.

thanks. i hope i didn't ask too much

 

[teklektik]

• other request, now that i have my pas installed, and got it working: i use pas together with "torq pas" ...
  i would ask for a steeper curve rpm->watt output. this way you can pedal very slowly (55rpm) and ride slowly, and when you need to climb the hill, or want to go faster you just do 90rpm.

thanks. i hope i didn't ask too much

Not at all! I'm glad to hear there is some interest in this. I had the same thought a while ago and implemented exactly what you have requested. It works really well and is particularly nice when you downshift - the rpms jump and the Watts leap right up to assist.

Unfortunately, this one is a little less 'cookbook' to configure. I spent some time testing and trying to put together something simple, but got interrupted with Life. Anyhow, I will get back on this straightaway and will get something posted very shortly (US holiday stuff going on for the next days anyway...)

It's good to hear from another AutoTorqPAS user who likes rpm scaling with a simple PAS wheel - I'm hoping Justin can get it scheduled to appear someday as a standard feature.

 

[izeman]

cool. thanks!

 

[shorza]

I have a weird problem that i am sure is just a setting that i have changed.

Upon start-up, for the first 30 seconds or so, the bike is very slow. Even at full throttle, the bike ramps up very slowly. Batt voltage and throttle in looks normal on the CA, but throttle out is low and rises slowly.
After 30 seconds or so, the bike is back to normal.

Any idea which setting could cause this?

 

[izeman]

It's good to hear from another AutoTorqPAS user who likes rpm scaling with a simple PAS wheel - I'm hoping Justin can get it scheduled to appear someday as a standard feature.

maybe i could be solved this way (from the user side): set a RPMmin and RPMmax and give them Wmin and Wmax values.
eg. 30rpm=300w and 100rpm=1500w. and the CA linearly ramps up according to rpm.
btw: why is the lower limit 55rpm? i sometime like to pedal slower ...

 

[izeman]

by "throttle out" do you mean power output, or do you really mean output voltage (as observed in screen #11).
how are your throttle values? please post them here - remember those values are preset dependent. i happened more than once, that i change values for the wrong preset and wondered that it didn't change

 

[teklektik]

It's good to hear from another AutoTorqPAS user who likes rpm scaling with a simple PAS wheel - I'm hoping Justin can get it scheduled to appear someday as a standard feature.

maybe i could be solved this way (from the user side): set a RPMmin and RPMmax and give them Wmin and Wmax values.
eg. 30rpm=300w and 100rpm=1500w. and the CA linearly ramps up according to rpm.

Yep - that was my suggestion to Justin: replace PAS->PASWatts with PAS->Watts55 and PAS->Watts120. Maybe in a future version...

btw: why is the lower limit 55rpm? i sometime like to pedal slower

This was chosen to yield sufficient power to make PAS getaways effective - any lower and modest motors lug and are in a poor operating range. This way as soon as pedaling is detected, the bike can apply meaningful assist. As always, additional configuration parameters eat up valuable memory, so I think this was chosen as a pretty fair compromise - I believe that as a rule of thumb the typical range for recreational riders is generally considered to be 60-80rpm.

 

[izeman]

... I believe that as a rule of thumb the typical range for recreational riders is generally considered to be 60-80rpm.

i agree. i had a long ride (around 60km) today, and the power application was perfect. 600w pas torque, and the power went up to some 1200w. going 25mph/40kmh was easy and not too power consuming.
i wished for a lower rpm setting for going up steep hills. but i has absolutely no priority.
btw and off topic: today after the long ride (i never discharged the pack before, so i had no idea of the real total Ah) just when i road next to the office where my car was parked, the LVC happened and the bike turned off. 50m more and i had to pedal. now i know my battery can deliver 18.6Ah and my bms works great!

 

[teklektik]

Here's a revised AutoTorqPAS setup procedure that allows selection of the rate at which power increases with rpm. This is a rework of section 7.4 of the v3.0p5 UUG. Select the desired power levels at 55rpm and 90rpm in a table and enter the Setup Trq parameters from the selected table cell.

There are three settings required and they interact in sort of non-intuitive ways. If you need in-between values, a little interpolation and fudging from neighboring table cells should get you close without too much trouble.

If during testing you find the 55rpm wattage to be a bit off, diddling the AsstStart value is the easiest fix even though there will be some effect on the upper power level as well. Just take the error in watts, divide it by the AsstFactr and use the result to adjust AsstStart up or down as appropriate. Again, this is a cheat for small corrections, but easy to do...

For purposes of comparison, the technique presently published in the Guide (Trq->AsstStart=0) has a 55rpm-90rpm rate factor of about 1.8x (see top title row of the tables).

I gave this a few spot checks on the road and things seemed to work okay, but feedback is appreciated. Enjoy!

EDIT - file relocated to UUG download post.

 

[izeman]

thanks teklektik. i will definately try as soon as my motor is up and running again

 

[rolf_w]

I'd like to use a CAv3 to set up 3 modes on my newest e-bike built in accordance with the 2012 Swiss regulations:

e-Velo: 0.5kW max. motor power, up to 20km/h 'motor only mode' allowed, up to 25km/h only 'motor assist mode' allowed, above 25km/h pedal only (no motor allowed)
e-Bike: 1.0kW max. motor power, up to 30km/h 'motor only mode' allowed, up to 45km/h only 'motor assist mode' allowed, above 45km/h pedal only (no motor allowed)
e-Moto: open for off-road

I also would like to integrate a Foundingpower programmable BMS to limit / disable the controller throttle (either via the CA or directly) in case of a low voltage alarm of any cell or other faults in the battery pack.

How could a possible scenario look like? Mode selection on the CA is a must I guess; would it be possible to have the CA starting in the slow mode upon booting? thus being in the 'e-Velo' mode regardless of any auxiliary settings?
thx, r

 

[izeman]

I'd like to use a CAv3 to set up 3 modes on my newest e-bike built in accordance with the 2012 Swiss regulations:

e-Velo: 0.5kW max. motor power, up to 20km/h 'motor only mode' allowed, up to 25km/h only 'motor assist mode' allowed, above 25km/h pedal only (no motor allowed)
e-Bike: 1.0kW max. motor power, up to 30km/h 'motor only mode' allowed, up to 45km/h only 'motor assist mode' allowed, above 45km/h pedal only (no motor allowed)
e-Moto: open for off-road

I also would like to integrate a Foundingpower programmable BMS to limit / disable the controller throttle (either via the CA or directly) in case of a low voltage alarm of any cell or other faults in the battery pack.

How could a possible scenario look like? Mode selection on the CA is a must I guess; would it be possible to have the CA starting in the slow mode upon booting? thus being in the 'e-Velo' mode regardless of any auxiliary settings?
thx, r

+
yes. yes. yes.
and yes
everything possible. download ca-v3 manual - teklektik did a great job, and read the part about power preset 1.2.1
you will need a pas as well.
a bms will work totally independent from CA. this has nothing to do with it. when the bms shuts down the power, CA has no influence on that. it's not the throttle that gets disabled. the bms totally CUTS the power to the controller, shutting everything down.

 

[Architectonic]

I'd like to use a CAv3 to set up 3 modes on my newest e-bike built in accordance with the 2012 Swiss regulations

Slightly off topic, but do you have a link officially showing these regulations?