Cycle Analyst V3 preview and first beta release

[irq]

Ca died again

Bike was standing still, just hooked the battery and did something else. After ~10min Isle i heard a rattling noise which came from the Motor which stuttered. The ca only showed squares.

Pulled battery plug, replugged, backlight alive but nothing else.

R9 was very hot, r8 dead (cut). Now even with 12v feeding into 12v + led it stays dead except abcklight. Just to get sure i tried a different controller without any dofference.

:[

 

[teklektik]

Really sorry to hear the saga has another chapter...
Sounds like a part or two may have been compromised during the earlier mishap and finally punched out.
Looks like it's time for an email to Reinhard to set up a repair/PCB exchange.

 

[irq]

I already replaced the known broken parts, last things i changed wäre:

Increased shutdown voltage since the ca started to late saving the current data, imho i went up to 15v. Could this cause the ca to no more booting? And latest b9 3.1 has Bern flashed 1 HR after release. I am not yet done

R8 is 4.7 k, right?

 

[teklektik]

Increased shutdown voltage since the ca started to late saving the current data, imho i went up to 15v. Could this cause the ca to no more booting?

No - don't see that as an issue. Some CAs had trouble booting but if yours never did, then it's not that strange problem (seemed to be a silicon problem from PIC).

R8 is 4.7 k, right?

I don't have a parts list, just the schematic from a year or so back. It looks like the value is 3.32K - it's a series resistor on the (S-) shunt pad, R5 is the same value on the matching (S+) pad. The schematic has some overprinting of parts designators and values, so it's a little hard to make out, but I'm 98% sure that's correct. If your R5 is okay and legible, that should be the value of R8 as well - just in case the schematic was revised in a subsequent edition (I need to get the latest...).

 

[irq]

OK, was just an idea. But i think i am at least onbthe right track, no sign oft 5v anywhere. No wonder that the mcu wont show any lifesign. Just need to figure out the 5v regulation part for the beginning

I cant measure the 12v anymore at the positive pad of c8 (the tantalum cap vor data Store)

 

[irq]

Its alive again, a bit at least. Feeding the 5v rail with an 7805 makes the mcu happy and the 7805 hot. Seems like u3 is toast. I can also measure 60ohm in the + & - negative pads of c8, feels not right?

 

[irq]

U3 L0UB, a 5 Pin 5v SMD Regulator is my primary guess. But since my Neighbars already started to raise attention (again) about that strange guy in his basement at 2 am in the morning (must be someone building a thermonuclear warhead on 2 wheels, including fluxcompensator and lots of very suspicius wires) i stopped for today.

is there any schematic available about that 5v regulating part?

 

[teklektik]

Frankly, there's not much going on with the 5V regulator and the part number is absent on the schematic for some reason.
C8 is 150uF from regulator Vin to Gnd and C6 is 4.7uF from Vout to Gnd. The regulator enable pin is strapped to Vin.

The FET regulator you repaired earlier produces 10V which is available to run the torque sensor and also powers the LED backlight. The 5V regulator Vin is powered from the LED return voltage so the two regulators operate in cascade fashion.

 

[irq]

its a LP2985, the L0.. was the Imprint on the SMD Chip. 5v PMIC 150ma Regulator.

I should probably do some math about the current which all components i hooked onto the CA are using, Analog Pot, Digiaux 2 Button Switch, 2 Dir PAS, OpenLog @ Serial (which is one thing which raised attention, even since this Device got its own Power Source the CA Display flickers slightly when the OpenLog writes on its SD Card which is indicated by an blue LED on the OpenLog. Led Activity triggers CA LCD Flickering), 2 Ebrakes

uhm, hope i didnt forget something

 

[lightrush]

LP2985's data sheet says it has overcurrent protection. Does that mean the 5V lines are in fact fused (protected) from shorts?

 

[irq]

jep should be, thats also what tekletik/justin mentioned: 5v is protected, 10v not

but i am sure that there can be situations where even a protected power source dies. abuse makes nearly everything possible

 

[lightrush]

Oh yeah, for sure. I'm rather asking because I also have some dead CAs on my hand and I think I killed one of them by shorting the 10V line with a bad Thun. So I'm thinking of ways to protect these lines. If the 5V has protection, then I only need to worry about the 10V line.

 

[irq]

i am the prove that the 5v can be killed, too

will keep updating how things go on, thanks to justin i have several ideas where the problem could be.

 

[Allex]

First is the annoying throttle delay. I have the default passthru setting on the CA. When throttle goes through CA nothing happens for about 0,5-1sec(feels like it anyway) when I twist it. When I connect throttle to phaserunner everything is fine I have a direkt response.
I read about that you can change:
ThrO upRamp
ThrO fastRamp
But I dont even find those settings in my CA

Have you set up the CA according to the [strike]Un[/strike]official Guide? This will get your throttle dead zones squared away and help you set the ramping. I'm unsure why you cannot find the ThrO parameters - they should be prominently displayed in the 'Throt OUT' setup category of both the CA console setup and the Setup Utility. The ebikes.ca CA V3 page explains the parameter locations, etc. That page should fill in all the basics and get you the firmware and programming tools, etc. The following image shows these parameters (from 3.1b8 release, but same place in 3.01). Post back if this does not help in locating the ramping parameters.


You can inspect any delays introduced by the CA by examining the IN and OUT voltages on the Diagnostics screen. With a high UpRate, OUT should track IN instantaneously in PassThru mode. Those particular display fields ignore the Display Averaging parameter and works off the immediate raw voltages so there is no need to jiggle Pref->Averaging to eliminate artifacts of CA display smoothing.

As you can see in the image, the default UpRate is 2V/sec so if you are using the defaults and your throttle has the usual 3V range, you can get a maximum of (3V) / (2V/sec) = 1.5sec throttle delay if you slam it WOT on getaway.

Second is when I activate PAS and torque sensor, the throttle ramp is changed when using throttle - accelerations is very sluggish compared to when I disable pas/torq sensor) Can I have a crisp acceleration with throttle while PAS/torq is enabled?

It's a little difficult diagnose what is going on since the basic PassThru throttle tuning (ramping at least) is not fully configured and is operating improperly. You really need to get the basics squared away before you move on to these more advanced features - bad foundation and all that...

In any case, this should not be happening - torque PAS processing has nothing to do with throttle processing and throttle application always suppresses PAS processing. The only way I can see this happening is if you are switching presets to activate PAS which would bring a different UpRate and power limiting settings into play (see image above) -- but you didn't indicate you were using presets and haven't changed the UpRates, so this behavior is confusing...

I recommend that you flash 3.1b8 and start again with the Guide install procedure to get your throttle set up, then move on again to the TDCM. The re-flash will get you a few useful things:

• default settings which will eliminate a lot of interaction quesitions when following the Guide
• rate flag indicators on the Diagnostic Screen so you can see when those settings are affecting your throttle
  View attachment 1
• new PASRate parameter to allow separation of throttle (UpRate) and PAS (PASRate) ramping so you can tune each operating mode separately
  (see this new parameter in the above image)

If you have no cable to re-flash or reload default parameters, then proceed with the install procedure, etc using the factory-installed 3.01 firmware (verify on splash screen).

If issues remain unresolved, post up your settings file and firmware version so we can look at the particulars of your configuration. Settings can interact, so it's not always easy to diagnose an issue with just a high level description of what has been set up...

Thank you! After update it does work much better. But now a few functions are gone, like holding left button to exit Setup, have to fiddle around that...

 

[izeman]

alexx: another hint: if you press LEFT button from startup screen once you will see the diagnostic screen teklektik mentioned, you will see a live throtte input/output screen. with a throttle rampage of 99V/s you will see that there is no noticeable delay between input and output.
i flashed b9 firmware this weekend, and all features are there. so long LEFT press exits setup immediately, and long RIGHT press shows a confirmation screen asking to confirm the saved data reset.

 

[teklektik]

Thank you! After update it does work much better. But now a few functions are gone, like holding left button to exit Setup, have to fiddle around that...

Glad to see things are improving.

It's good to get feedback on the beta releases, but we need to know the firmware release you are running and the circumstance of the failure - that is, what settings or actions replicate the problem. Usually we can assume it's the last beta version, but in your case we had talked about b8 but b9 was released in the interim, so the situation is a little unclear...

 

[teklektik]

i flashed b9 firmware this weekend, and all features are there. so long LEFT press exits setup immediately, and long RIGHT press shows a confirmation screen asking to confirm the saved data reset.

Thanks for the report - much appreciated!

 

[r3volved]

lol, Tek is an indispensable asset! And as a former tech writer I endorse that RTFM!!

 

[adriftatsea]

Problem:
My CA3 takes a long time to shut down.

I have it in a 72v (20s) vehicle with two Curtis 1238 controllers, who's caps I believe are feeding the CA3 when I shut the contactor off.

Is there a suggested way to address this? I've thought about inline resistors and optionally breaking the 72v wire to the CA3 with a 12v relay from my dc-dc converter.

 

[teklektik]

• Have you tried raising Pref->Vshutdown? Bring this up to something about 5V under Batt->VltCutoff.
  
  
• If you have an unisolated 12V DC/DC converter, you can rewire your CA according to the scheme laid out in the Guide Section 5.9 "Powering the CA / High Voltage Vehicle Support" and power the CA from your 12V ignition switch and use Vext to monitor the 72V battery. This will shut down the CA more or less immediately when you switch off the ignition. Just needs a couple of resistors.
  
  
• If your primary 12V DC/DC converter is isolated, you can add a small unisolated DC/DC converter exclusively for the CA and do as in (B). This is about a $10-$15 (or less) solution. See the Guide section 6.1 "DC/DC Converters".
  
  
• That said, there's nothing wrong with your relay strategy if that's a solution technology with which you are comfortable, although I'd go with a reed relay so that the thing was sealed and would last about forever. Here I'm assuming you're not using the CA accessory power plug and the CA is just instrumentation on the dash and so will only draw CA operating current to keep the tiny reed relay contacts within spec...

 

[izeman]

My CA3 takes a long time to shut down.

if you build/use an inrush limiter with some FETs, this will pull down the voltage seen by the controller in no time.

 

[lightrush]

Am I right in the understanding that the 10V pad is directly connected to end of the main regulator and to LED+?

 

[irq]

back in the game, with an ca replacement unit. so far so good, first ride was absolutly fine. arrived at work, placed it in a corner where it was waiting until 5pm. after ~2hr it suddenly started accelerating itself, noone was even near it

Pulled the Battery Plug, plugged in again -> everything back normal. what the heck... ?!

 

[izeman]

back in the game, with an ca replacement unit. so far so good, first ride was absolutly fine. arrived at work, placed it in a corner where it was waiting until 5pm. after ~2hr it suddenly started accelerating itself, noone was even near it

don't you turn off your bike while not riding? i would never leave the battery connected over a longer period.
maybe the throttle calibration was a way off, or there was some kriechstrom

 

[irq]

Theres no thumbthrottle connected, pas only.

Some leak voltage was also one oft the first ideas, but i think that i have Seen threads where users reported strange things caused by the wire Route? Magnetical interferences, but since it was idle...

Maybe some ferrite rings wont hurt?