Cycle Analyst V3 preview and first beta release

Hello.
I'n sorry if this has been answered already.
I have a very old CA3-DPS from 2015, with goldenmotor controler, with the integrated speed sensor.
It had an unrecognized version by the CAsStup utility.
It was something like r1b16 (I didn't put too much attention to it).
When I updated to the latest firwmare (3.14) I realised that the CA didn't report any speed anymore.

The wheel circumference is correctly setup.

I tried with older firmware like the CA3-1b13, and I could get sporadic speed report at 1/10th the real speed, with very laggy display (took a few seconds to show 4-5km/h).

Do I have to setup something special for the DPS? I couldnt find any settings relative to the speed senror in the utility.
I can hear the sensor "clicking" when the magnet pass in front of it, I'm pretty sure the sensor is still working.

I also confirm spontaneous power surges with ERider T9 and 3.15b3 firmware. When pedaling stops for half second, I get power surges when start pedaling again. Over one second pauses not cause power surges. When set PAS ramp up as low as 0.5V/s it will tame power surges, but also increase responce time. PAS ramp 1.0V/s or more is uncomfortable, 5V/s generate dangerous power surges.

I also ran into a problem with speedo, Cycle Analyst no longer shows speed. In settings I have 23 poles (DD motor) and when I spin wheel, indicator in CA settings move up and down. With older firmware 3.13 which I had speed show normally.

I made fresh Firmware update as set all settings to default to avoid glitches.

It would be nice to have a comment on these issues.

Thanks.

There is also a thread here tracking issues with the CA3 and the E-Rider T9: https://www.endless-sphere.com/forums/viewtopic.php?f=2&t=112368&p=1672030#p1672030

I have an issue with the display of the CA V3. It switches from one screen to the other without me touching the MFSwitch (latest CA switch).
I have a GMA 10T, baserunner L10, 48V9Ah battery (High Power VTC6 cells), e-rider T68
My kit started running in June 2020 and I didn't have the issue at the beginning. I cannot remember what components I added before the issue started (if it's related to that). It could be the e-rider, not sure. I must have tried to test a normal PAS a few months ago and disconnect the e-rider it but I don't remember it solving the issue. I can try that again.

I tried everything (I think) - It's not coming from a "bad" Switch or CA because I changed both and nothing changed.
My trike is putting a lot of vibration but it's not that either (I tried with the CA in my hand while riding).
I cross my finger that I am not alone and someone has a solution because it's a pain to ride with a display that is constantly changing.

zorbas said:

I have an issue with the display of the CA V3. It switches from one screen to the other without me touching the MFSwitch (latest CA switch).
I have a GMA 10T, baserunner L10, 48V9Ah battery (High Power VTC6 cells), e-rider T68
My kit started running in June 2020 and I didn't have the issue at the beginning. I cannot remember what components I added before the issue started (if it's related to that). It could be the e-rider, not sure. I must have tried to test a normal PAS a few months ago and disconnect the e-rider it but I don't remember it solving the issue. I can try that again.

I tried everything (I think) - It's not coming from a "bad" Switch or CA because I changed both and nothing changed.
My trike is putting a lot of vibration but it's not that either (I tried with the CA in my hand while riding).
I cross my finger that I am not alone and someone has a solution because it's a pain to ride with a display that is constantly changing.

Click to expand...

Hello, is it cycling regularly?
If I remember well there is a setting that will enable automatic cycling of the menus.
Maybe you activated it by mistake.
If not, you should check connection issue with the cycling button, maye there is a fault and it activate from time to time.

Question about the aux power output on the CA3:

The current is rated at 1 amp. How is this protected? Normal fuse, slow blow fuse or PTC? I am drawing .5 amps (@48V) for lighting loads and I want to make sure I'm not fatiguing a standard-blow fuse with whatever the turn on surge is.

Also is the current from the aux power output monitored by the shunt in the controller? I am using a Phaserunner and I wasn't sure if the current drawn by the lights will show up in the total power display on the CA.

xmtx said:

I tried with older firmware like the CA3-1b13, and I could get sporadic speed report at 1/10th the real speed, with very laggy display (took a few seconds to show 4-5km/h).

Do I have to setup something special for the DPS? I couldnt find any settings relative to the speed senror in the utility.
I can hear the sensor "clicking" when the magnet pass in front of it, I'm pretty sure the sensor is still working.

Click to expand...

For a spoke magnet system you need to make sure that the #poles setting is at 1, and not the default value of 23. Can you confirm if this was the case?

JackFlorey said:

The current is rated at 1 amp. How is this protected? Normal fuse, slow blow fuse or PTC?

Click to expand...

It is protected by a PTC, so feel free to trip it as much as you like! One small detail to note is that higher currents in the DC jack do cause a ground potential difference between the CA and the motor controller from voltage drop on the ground return wire, and that in turn causes an offset in the throttle output voltage of the CA and the throttle input voltage of the motor controller. For people who set their min throttle output of the CA just a hair above where the motor controller kicks in, then a 1A current on the DC jack could lift this enough that the motor starts to spin even without you touching the throttle.

xmtx said:

zorbas said:

..
I tried everything (I think) - It's not coming from a "bad" Switch or CA because I changed both and nothing changed.

Click to expand...

Hello, is it cycling regularly?
If I remember well there is a setting that will enable automatic cycling of the menus.
Maybe you activated it by mistake.
If not, you should check connection issue with the cycling button, maye there is a fault and it activate from time to time.

Click to expand...

Can you be more specific with which parameter is allowing automatic display “circling”? I checked again today and found nothing.
The only aumatic display modification is from the parameter “Autoreturn to mainscreen”. This is set to “no” currently.

The display button is not the issue because as I said, I bought a new one and the issue remains. I also bought a new CA (as suggested by Grintech) and the issue is still there.
I forgot to mention that this is only happening when I am moving.
Would be nice to have a solution from them..

I did some additional check today. It seems to be related the some kind of interference between the engine and the CA.
https://endless-sphere.com/forums/viewtopic.php?f=49&t=113161&p=1674943#p1674943

Does that make sense?

justin_le said:

It is protected by a PTC, so feel free to trip it as much as you like!

Click to expand...

OK thanks for the info.

Also discovered that the tap for the aux power is indeed after the current sense resistor, so the CA's reported power includes the power used on the aux port.

Hello again all. A few more questions:

I have a V3 with the new five-pin connector for the power/aux switch. Does this work with the solar build? Do the switches on the aux switch use the same aux line as the solar current analog in? (don't want to break anything)

Also does anyone have a pinout for that 5 pin connector? Seems like there's too many pins in that connector for what it does.

justin_le said:

xmtx said:

I tried with older firmware like the CA3-1b13, and I could get sporadic speed report at 1/10th the real speed, with very laggy display (took a few seconds to show 4-5km/h).

Do I have to setup something special for the DPS? I couldnt find any settings relative to the speed senror in the utility.
I can hear the sensor "clicking" when the magnet pass in front of it, I'm pretty sure the sensor is still working.

Click to expand...

For a spoke magnet system you need to make sure that the #poles setting is at 1, and not the default value of 23. Can you confirm if this was the case?

Click to expand...

Wow, this is so obvious that I feel stupid now.
It's working great with the latest firmware.

Thank you very much for the answer and all your work, I really love your products (got the thru axle motor, baserunner and CA of course).

zorbas said:

Can you be more specific with which parameter is allowing automatic display “circling”? I checked again today and found nothing.
The only aumatic display modification is from the parameter “Autoreturn to mainscreen”. This is set to “no” currently.

Click to expand...

Yep, it was this one. Sorry it didn't help.

justin_le said:

It is protected by a PTC, so feel free to trip it as much as you like! One small detail to note is that higher currents in the DC jack do cause a ground potential difference between the CA and the motor controller from voltage drop on the ground return wire, and that in turn causes an offset in the throttle output voltage of the CA and the throttle input voltage of the motor controller. For people who set their min throttle output of the CA just a hair above where the motor controller kicks in, then a 1A current on the DC jack could lift this enough that the motor starts to spin even without you touching the throttle.

Click to expand...

yeah I noticed that when I installed a 12-80v spotlight to the dc out
draws like 20w+ or 0.3-0.5A
reports on my CA as drawing 50-60w
and yeah the initial installations was annoying due to it triggering minor throttle until I increased minimal throttle on Phaserunner by 0.10v
also if I have the light switched onto high beam on boot it triggers/loops the CA's booting outside of it though its perfectly fine and functional

I can confirm the wacky speed displayed with 3.15b3 (and 3.15b1). The speed is nowhere near correct and seems to max out at about 10mph but is constantly jumping all around. Reverting back to 3.14 fixes it but doesn't work with the new ERider PAS.

It took me a while to figure out that I needed 3.15b to get my new ERider T9N to work. I would rather have a working PAS than a working speedometer so I'll stick with 3.15b3 for now. Hopefully the speed issue is fixed soon.

Another suggestion. It might be helpful to pop up an error message in the CA setup utility when a manually entered setting is over ridden by a default setting. Most of my struggle getting my ERider setup correctly was that I when typed in 36 pas poles in the utility, I didn't notice that it kept reverting to 32 before I wrote it to the CA. Of course now that I have 3.15b3 and can set it to 36 it works fine.

I love my Grin bikes. Thanks.

Does the CA V3 remaining battery capacity graphic represent empty as the cell being 100% discharged or does it represent the cell being 80% discharged or?

100% discharged of usable capacity. Normally figured at 3V per cell. But as all things are not perfect. You can tweak your CA by altering you battery's rated Ah in the CA. If it shows to little rise Ah if to much lower it.

Depends on how you use your battery. Pulling low amps your battery will seem to have more amps. If used at higher discharge rates then it will seem to have less amps available.
Not sure if it uses the cut off voltage in its calculations.

Here is a quote from the CA pdf web version.

Set Your Capacity
Here you input the approximate amp-hours of your
battery pack in order to help the battery SOC icon
remain accurate at tracking changes during high
discharge currents. The value does not need to be exact as the battery icon will
always gradually readjust itself based on the voltage reading.

Click to expand...

by BVH » Sep 19 2021 6:19pm

Does the CA V3 remaining battery capacity graphic represent empty as the cell being 100% discharged or does it represent the cell being 80% discharged or?

Click to expand...

ZeroEm said:

Pulling low amps your battery will seem to have more amps. If used at higher discharge rates then it will seem to have less amps available.

Click to expand...

Nit pick, change those capacity units from amps to Ah.

Also remove the "seem".

Batteries actually **do** deliver more energy (Ah) at low C-rates and less energy at high C-rates.

:lol:

ZeroEm said:

100% discharged of usable capacity. Normally figured at 3V per cell. But as all things are not perfect. You can tweak your CA by altering you battery's rated Ah in the CA. If it shows to little rise Ah if to much lower it.

Depends on how you use your battery. Pulling low amps your battery will seem to have more amps. If used at higher discharge rates then it will seem to have less amps available.
Not sure if it uses the cut off voltage in its calculations.

Here is a quote from the CA pdf web version.

Set Your Capacity
Here you input the approximate amp-hours of your
battery pack in order to help the battery SOC icon
remain accurate at tracking changes during high
discharge currents. The value does not need to be exact as the battery icon will
always gradually readjust itself based on the voltage reading.

Click to expand...

by BVH » Sep 19 2021 6:19pm

Does the CA V3 remaining battery capacity graphic represent empty as the cell being 100% discharged or does it represent the cell being 80% discharged or?

Click to expand...

Click to expand...

yeah would be nice if we could configure the graphic scaling with the batteries cutoff instead.
my cells have a 2.5V cutoff that i have set to cutoff around 2.8V thanks to the CA's LVC scaling safely for emergencies in a large travel.
telling me batteries dead at 3/3.2V doesnt help lol

In the battery setup of the CA3. (below is out of the manual)

6.3.4 Low Voltage Cutoff
This optional feature allows the Cycle Analyst to automatically scale back power
as the battery approaches a low voltage cutoff point. While lithium batteries have
a Battery Management circuit (BMS) that will shut the battery off to prevent
overdischarge, the experience for the rider is an abrupt loss of power with no
warning. If the Cycle Analyst low voltage cutoff is set 1-2 volts higher than the
BMS shutoff voltage, you'll have a gradual reduction in power instead, and will
usually get more amp-hours and range from the pack.
When the CA is at the low voltage rollback mode, you will see the 'V' character
flashing on the main display, and a capital 'V' on the diagnostics screen.

Click to expand...

Dartman said:

I also confirm spontaneous power surges with ERider T9 and 3.15b3 firmware. When pedaling stops for half second, I get power surges when start pedaling again. Over one second pauses not cause power surges. When set PAS ramp up as low as 0.5V/s it will tame power surges, but also increase responce time. PAS ramp 1.0V/s or more is uncomfortable, 5V/s generate dangerous power surges.
…
It would be nice to have a comment on these issues.

Click to expand...

Bump! Why is none of the developers responding to the power surge issue, which is clearly a big and common issue?

The solution is also very clear: please implement the possibility for a torque sensor to directly control the throttle out voltage, without any feedback loop. Just like a normal throttle, which a torque sensor essentially is (pedal operated throttle).

The power surge issue has nothing to do with the type of torque sensor, only with the control logic that is trying to perform some closed loop correction to achieve a certain power level when you briefly stop pedaling. An open loop direct implementation immediately solves the problem, so please give us the option to let the torque sensor control a throttle out voltage without feedback loop.

I have tested this solution/theory and it works fine: when i set the torque sensor to perform like a basic PAS, and let the PAS control the throttle out voltage (fixed level), then the result is very well behaved: the bike pulls up at a certain (constant) throttle level without any power surges, no matter how often you stop pedaling. The only problem is that the throttle is now at a fixed level, so the torque part of the sensor has no effect.

So devs, please do not ignore this issue and address this dangerous behavior of closed loop power control. Give us the option of open loop throttle out voltage control by the torque sensor signal, and all problems and dangers are gone. So we can finally enjoy the ca as the torque control enabler that it currently is pretending to be (but clearly is not safely doing).

Please take this seriously, respond to us, and solve it once and for ever. Thank you.

Which devs do you think you are addressing with your suggestions?

I suspect there are more effective channels, they are likely not reading these threads.

I want to know who US is?

Iambuilderman said:

Dartman said:

I also confirm spontaneous power surges with ERider T9 and 3.15b3 firmware. When pedaling stops for half second, I get power surges when start pedaling again. Over one second pauses not cause power surges. When set PAS ramp up as low as 0.5V/s it will tame power surges, but also increase responce time. PAS ramp 1.0V/s or more is uncomfortable, 5V/s generate dangerous power surges.
…
It would be nice to have a comment on these issues.

Click to expand...

The power surge issue has nothing to do with the type of torque sensor, only with the control logic that is trying to perform some closed loop correction to achieve a certain power level when you briefly stop pedaling. An open loop direct implementation immediately solves the problem, so please give us the option to let the torque sensor control a throttle out voltage without feedback loop.

Click to expand...

You might have some luck getting the desired behavior by increasing the Amps and Watts Gain in the CA. See the Unofficial CA Manual for an explanation of the feedback loop.