Cycle Analyst V3 preview and first beta release

[Kepler]

Looks like the error was:

PC02117 U100000
T_Err U200000

I had the same error the other day. Occurred after an extended ride using the thun. Even after resetting power it continued to occur. I then disabled the torque sensor and the error went away. Switched Back to the torque sensor the next day and have used it even since to see if the fault would re occur. So far no repeat.

 

[teklektik]

With all the additional processing and functions going on, the B23/CA3Prelim code could with just the right circumstances take a bit too long to execute the main code routine (theoretically every 18.1818mS)...

I had the same error the other day. Occurred after an extended ride using the thun. Even after resetting power it continued to occur. I then disabled the torque sensor and the error went away...

As Justin pointed out this is due to a bit too much calculation in the main processing loop and as Kepler discovered, it can be just on the edge depending on the suite of features enabled and the phase of the moon. Justin reported that one of the contributing calculations was due to running configured for miles instead of km (distance/speed calcs appear to be natively based on the tire circumference in mm).

So - if you experience the problem, it continues to occur, and you live in the US (or another of the one or two metrification-resistant countries), try switching to km instead of forfeiting a more critical or desirable feature such as PAS. I ran some cursory tests when this was initially discovered and the mi/km trick proved effective. Since this is merely a contributor and not a cause, the problem can arise even if you are using km, so no guarantees...

 

[Sebbo]

Ich muss mich jetzt mal in Deutsch hier erkundigen. Leider kann ich es nicht in Englisch formulieren.


Ich hab den V3 mit der aktuellen Firmware. Dabei versuche ich eine Regelung über Drehmoment und KM/H zu verwirklichen.

Dazu habe ich 2 Settings. Eines mit dem ich die obige Steuerung machen möchte und das andere mit offenen Einstellungen. Also Gasgriff ohne Pedalzwang. 2 Motoren mit jeweils 2,5kw an 20s.

Ich hab aber das Problem das sich die Einstellungen von SLim, IntSGain, PSGain und DSGain auf alle Profile übertragen. Das heisst ich habe generell schwierigkeiten bei Vollgas anfahrten wenn das Vorderrad abhebt und er denkt ich komme gleich ans Speedlimit. Das in der offenen Einstellung auf 499KmH steht.

Wenn ich die Werte alle runter stelle: SLim->MaxSpeed to the maximum value, IntSGain = 1, PSGain = 0, DSGain = 0

Dann kann ich auch die Einstellung für 25KMh begrenzung vergessen... Und wenn ich die Einstellungen auf 25KmH stelle dann schwingt das ganze immer total über. Trotz einer Leistungsreduzierung auf 700w für beide Motoren zusammen...


Ich lese jetzt schon seit Wochen wie ein blöder und bekomme das nicht sauber gelöst.


HELP!


Grüße Sebastian

 

[teklektik]

Google Translate of Sebbo post:

I must now ask times in German here. Unfortunately I can not express in English.

I have the V3 with the latest firmware. Here I try a system of torque and KM / H realize.

In addition I have two settings. One with whom I would like to make the above control and the other with open settings. So throttle pedal without coercion. 2 motors with 2.5 kw at 20s.

But I have the problem that the settings of SLim, IntSGain, PSGain DSGain and applied to all profiles. That means I'm generally journeys difficulties at full throttle when the front wheel lifts off and he thinks I'm coming to the speed limit. It's in the open setting on 499KmH.

If I put down all the values​​: SLim-> MaxSpeed ​​to the maximum value, IntSGain = 1, PSGain = 0, = 0 DSGain

Then I can forget about the setting for 25kmh limit ... And when I put the settings on 25kmh then the whole vibrates always totally over. Despite a reduction of power at 700w for both engines together ...

I am reading it for weeks as a stupid and do not get the clean solved

Just trying to ensure the post is understood properly as a first step:

• 
  If I read this correctly you are using km/kph and have a 2.5kW two motor bike set for two presets:
  1. using a torque sensor and
  2. straight throttle control with no PAS limiting.
  The problem is that when you wheelie the bike (in mode 2), the front wheel spins up and seems to be introducing some limiting. This appears to be the DSGain issue of rapid wheel acceleration without actually hitting the speed limit which you have set to the max of 499kph. To counter this, you have set SLim-> MaxSpeed ​​to the maximum value, IntSGain = 1, PSGain = 0, = 0 DSGain. This seems to work for mode 2.
  
  However, when you set SLim->MaxSpeed to 25kph (to properly configure mode 1), the bike vibrates and even reducing PLim->MaxPower to 700W does not stop the vibration.
  
  The problem that you see is that IntSGain, PSGain, and DSGain are global parameters and apply to all settings - so you cannot successfully configure both the torque sensor and throttle-only modes to work properly.

Does this seem a correct statement of the problem (assuming here that you are pretty much okay reading English - otherwise we will need one our German V3 users to help out )

 

[Sebbo]

Yes!

If I want to reduce it to Germany to 500watt and 25 KMH with torque. no other mode is possible. Because the settings are not configured separately for each mode.

 

[Merlin]

hey sebbo, so that means that PRESET 1-3 have only one setting for Torque Mode? Like the Max Speed Setting what is for all the Modes the Same?

 

[Sebbo]

only one setting for the speed settings. IntSGain, PSGain, and DSGain

 

[teklektik]

Does this seem a correct statement of the problem?

Yes!
If I want to reduce it to Germany to 500watt and 25 KMH with torque.

My first suggestion would be to avoid the wheelie effect by picking up your speed from the rear wheel, either using the pickup or by tapping the SP signal on the rear controller CA-DP connector. This is really a workaround instead of a solution, but might as well throw it out there...

So - to actually try to solve the configuration problem, here's a few questions to better understand you PAS configuration and how it is failing:
For the PAS mode that is vibrating:

1. What are these parameter values?
   • PLim->MaxAmps
   • ThrI->CntrlMode
   • ThrO->UpRamp
   • PAS->PASMode
   • PAS->MxThrotSpd
2. What MaxSpeed, IntSGain, PSGain, and DSGain settings work acceptably in PAS Mode?
3. What limit flags are asserted (upper case) on the Diagnostic Screen when the vibrating is in progress?
4. What initiates the vibration? (start to pedal, apply throttle, etc)

Apologies for lots of questions - there's just lots of parameters in play.....

 

[Sebbo]

I have now first set without PAS. Only speed limit.

PLIM-> 35A 150gain , 700Watt 5Gain
THRL-> Power (was the best)
Thro UP 0.50 V / sec
FastRamp 2.25 V / sec (which means it? Doubt)

and I had tried to
PAS> TRQ | THR
PAS> MxThrotSpd 6kmh


MaxSpeed ​​- 25
IntSGain PSGain DSGain and I just tested without PAS and found no proper setting. Therefore I have set only pedaling.

What limit flags are asserted (uppercase) on the Diagnostic Screen When The vibrating is in progress?

Actually W and S.


What Initiates the vibration? (start to pedal, apply throttle, etc)

It drives everything is beautifully achieved up to 25 kmh. Then it turns off spontaneously. Until it is restored to about 20, and then it gets going hard. back to about 25kmh. and again



sometimes in CA shows top speed over 400!

 

[teklektik]

I have now first set without PAS. Only speed limit.

PLIM-> 35A 150gain , 700Watt 5Gain
THRL-> Power (was the best)
Thro UP 0.50 V / sec
FastRamp 2.25 V / sec (which means it? Doubt)

and I had tried to
PAS> TRQ | THR
PAS> MxThrotSpd 6kmh
MaxSpeed ​​- 25
IntSGain PSGain DSGain and I just tested without PAS and found no proper setting.

First - the units for ramping are Sec/Volt - this is the minimum number of seconds required for Throttle OUT to change one volt. If your Throttle OUT range from ZERO to WOT is 1.0v to 4.0v, then slamming the throttle WOT on getaway will cause a 3 volt change. Let's make this easy and always set FastRamp the same as UpRamp so if they are set to 2.0sec/V then it will take 3v x 2.0sec/v = 6sec to reach WOT. Start by setting SLim->MaxSpeed=499kph and adjust UpRamp and FastRamp the way you want them (make them equal to start - you can reduce FastRamp later after you get your PAS working so the power comes on more quickly when you resume pedaling and are already underway - read this post)

Next - Setting just the speed without PAS was a good strategy. Unfortunately, working out 'good' settings for speed limiting can be tricky and as you probably noticed, that section of Appendix B on page 34 of the User Guide is blanked out since the procedures are not finalized. Here is a part of that procedure that may work for you - it's designed for more powerful bikes that can overshoot the target speed - a different strategy is needed for lower powered bikes. Try this with PAS disabled - just focus on getting the speed limiting adjusted to work smoothly...

Unofficial User Guide Speed Tuning Procedure - Incomplete Draft Version

As mentioned in an earlier section, ramping logic can have a destabilizing effect on the control logic, particularly if it is suddenly engaged and introduces unexpected behavior into the feedback loop. Whenever possible, it is best to satisfy slower UpRamp requirements with more conservative (PSGain) adjustments so the controller has a nominal responsiveness in line with overall needs. This will give better stability than imposing throttle ramping to restrain controller tuning that has an unnecessarily aggressive response. As a bonus, it is somewhat easier to tune for a more damped response. Slower response can be achieved in the following tuning procedure in step (4) by further reducing PSGain beyond the recommended factor of 4 and/or by reducing IntSGain in step (5).

Note: Please be certain to have proper ramping values in place before proceeding. This is of particular importance for gear motors, mid-drives, and powerful DD motors since full throttle may be applied off the line resulting in driveline stress and potentially dangerous riding situations.

Begin by setting SLim->MaxSpeed to a conservative test speed e.g. (25kph). Set ThrI->CntrlMode=Speed. In the following tests, run the same stretch of level or very slightly inclined test road and apply WOT to achieve consistent test conditions.

1. Set IntSGain to 1 and DSGain to zero.
2. Calculate initial PSGain = 6 x (ThrO->MaxOutput - ThrO->MinOutput) / ( Slim->MaxSpeed ).
3. Increase PSGain to the point where a single modest overshoot occurs before the speed settles. If overshoot cannot be made to occur, this procedure is not appropriate - instead use the procedure for lower powered bikes (Low power procedure does not presently exist!)
4. Reduce PSGain to 1/4 of its value.
5. Increase IntSGain as large as possible without significant additional oscillations occurring after the initial overshoot.
6. Increase DSGain until settled operation is just jittery (power feels rough or ratty), then reduce 30%.

Restore the desired SLim->MaxSpeed and ThrI->CntrlMode.

No guarantees this procedure will work properly for you, but give it a try. If you can get this adjusted so there is no surging, then these same gain settings will work for both your PAS and unlimited throttle-only modes.

All builds will have different configuration, but to give you a notion of V3 settings for another 2WD bike,
My bike: 66v, 2 BMC v2S gear motors, 2 Lyen 12FETs configured for 36A rated/74A phase, block time=0.1
IntSGain = 15
PSGain = 0.16V/mph
DSGain = 30
upramp = 2.5
downramp=0.05
fastramp=2.0

What limit flags are asserted (uppercase) on the Diagnostic Screen When The vibrating is in progress?
Actually W and S.

Good - the power throttle accounts for the W and speed limiting accounts for the S - no other unexpected limiting.

What Initiates the vibration? (start to pedal, apply throttle, etc)

It drives everything is beautifully achieved up to 25 kmh. Then it turns off spontaneously. Until it is restored to about 20, and then it gets going hard. back to about 25kmh. and again

Okay - this is the speed oscillation or surging discussed in the User Guide. The Speed Tuning procedure above should tune the parameters so the bike overshoots 25kph somewhat, drops below it, then settles in to a constant 25kph. Overshoot and performance can be altered with different parameter settings, but let's just see if the surging can be eliminated to start - you can change things later...

sometimes in CA shows top speed over 400!

This is Not Good. The CA used to do this but the problem was addressed with a firmware upgrade quite a while ago. I have nothing to offer about this - this is a matter for Justin. I suspect that unless it is a real problem (motor cuts out for many seconds because of speed limiting), that it will be best to re-evaluate this issue when the next firmware comes out (it has re-organized processing to remedy timing loop violations mentioned in the posts above).

 

[Sebbo]

Okay, thank you first.

Today is Rainy. But after the work i will test something of the tips!



I write as soon as I could test it.




Grüße Sebastian

 

[Sebbo]

another question. I have now connected a KBS72101.

UpRamp to 0.05 sec / volt. Still going until long time after switching the VoltOut sync with the throttle.
Only after the first turn.

Do I need to change anything on the controller?

 

[teklektik]

another question. I have now connected a KBS72101.

UpRamp to 0.05 sec / volt. Still going until long time after switching the VoltOut sync with the throttle.
Only after the first turn.

Do I need to change anything on the controller?

A quick Google search shows that section 4.2 page 14 of the Kelly controller manual describes a throttle up/down ramp rate that can be adjusted.

In order to minimize interactions that may be troublesome to diagnose, I would suggest starting with the controller rates set as fast as possible and rely on the CA to provide the ramping. After things are basically working, you might return to these settings to see if there are advantages to allowing the controller to perform this task.

View attachment KellyKBSUserManual.pdf

 

[Sebbo]

Is already set to fast. In CA V3, it also shows the slow increase in the voltage. But if the upper voltage has been reached, it goes very fast.

Do I need a resistor or a diode in CA V3?

 

[teklektik]

Is already set to fast. In CA V3, it also shows the slow increase in the voltage. But if the upper voltage has been reached, it goes very fast.

Do I need a resistor or a diode in CA V3?

If you are seeing a slowly changing Throttle OUT signal on the V3 Diagnostic Screen then it has nothing to do with the controller connection.

There is probably a bad gain setting (IntSGain?) that is fouling things up - so many configuration and hardware changes all at once.

I would recommend restoring the default CA settings and re-evaluating this ramping issue in Pass-Thru mode (when you re-adjust the throttle) - basically start again instead of inheriting settings from your original controller. The defaults are on this form. Hopefully, as you add configuration changes the problem will not re-occur or it will be easy to identify the offending setting. Apologies - not a very directed solution, but...

 

[bspalteh]

Just wanted to add one clarification to the translation of Sebba's post. He was referring to an overshoot and large swings in the speed, not a vibration per se. I think that was already fairly clear a few posts later but just in case....

 

[teklektik]

Just wanted to add one clarification to the translation of Sebba's post. He was referring to an overshoot and large swings in the speed, not a vibration per se. I think that was already fairly clear a few posts later but just in case....

Thanks for the assist - that translation error was confusing, but as you point out, it got cleared up later and things made a lot more sense. I was thinking for a while that there was a new behavior mode

 

[justin_le]

It drives everything is beautifully achieved up to 25 kmh. Then it turns off spontaneously. Until it is restored to about 20, and then it gets going hard. back to about 25kmh. and again

sometimes in CA shows top speed over 400!

These things could be very much related. Right now is your speed signal coming from a speedometer pickup and spoke magnet, or is it from the controller hall signals? If the speed is from a sensor and magnet, then this means that you are either getting some switch bounce when the magnet passes the sensor, or you might have reed switch that is particularly touchy and affected by vibrations and bumps in the road. The first situation is easily resolved by adjusting the magnet/sensor spacing and placement.

Next time you ride, can you look at the speed readings and tell us if you see very high speed numbers just before every cutout?

-Justin

 

[Kepler]

In relation to the throttle out ramp settings being global, I have run into similar problems as Sebbo after setting up the CA3 on my 2.5 kW Fighter. I wanted a street legal 250W profile and an off road profile. I can tune each one perfectly on there own but need to compromise with the ramp settings to run both.

I am sure there are very good reasons for making these settings global however it sure would be nice if they could be set separately for each profile.

 

[Kepler]

For those running a torque sensor, wondering if you have noticed the following idiosyncrasies.

1: On first ramp up cycle, the power will drop off to zero then recover. After the first drop off it then in general holds throttle as it should.

2: pedaling with a constant crank load for a reasonable period of time, the power suddenly falls away to zero for no reason then pick back up to the correct throttle setting. This continues to happen intermittently through the ride.

I have tried many setting to try and improve this without success. Worth noting that if I set the CA to RPM PAS, I have no dropouts.

Trying to establish if this problem is common or something just to do with my setup.

 

[pendragon8000]

2: pedaling with a constant crank load for a reasonable period of time, the power suddenly falls away to zero for no reason then pick back up to the correct throttle setting. This continues to happen intermittently through the ride.

I to have experianced the same thing. but i am running throttle only at this stage, no aux or temp.
beta 21 version.
might be a coincidence but throttle has droped out for a tenth of a second or so and comes back for a few seconds or a minuet and drops again. only happened once that i am aware of(over a few mins during one comute) . honestly could have been a loose throttle wire but my electronics skill is prety sharp these days. and seems to inconsistent to be a physical wire issue and ive done the same trip 4 times and only happened on one of the trips, hadn't changed any wireing, might have been 'cos I lowered the power limit from1800 to 1650watt , dunno if that could make it hit a limit in the algorithm more agresively, watt ramp is 40 and throttle mode is power. throttle is set up quite well in regard to limits and ramp etc.

 

[teklektik]

In relation to the throttle out ramp settings being global, I have run into similar problems as Sebbo after setting up the CA3 on my 2.5 kW Fighter. I wanted a street legal 250W profile and an off road profile. I can tune each one perfectly on there own but need to compromise with the ramp settings to run both.

I am sure there are very good reasons for making these settings global however it sure would be nice if they could be set separately for each profile.

Limited parameter memory space and preset-specific parameters require three times the space.

UpRamp is preset-specific, but FastRamp is a new feature and is global.

Sebbo's difficulty has to do with using the trick to deactivate speed limiting - but he really needs it to work (in another preset) and so must tune it. Once the gain settings are properly set, they should be usable across all presets. Unfortunately, he hadn't gotten it working yet in the other preset and so turning it off in the first preset seemed like the only plan.

However, in a related vein, the User Guide calls out reduced gain as a means to alter behavior that might otherwise be addressed with ramping, but does not specifically mention the relationship to presets (my bad - I use a 3-position switch instead of presets). That section of the Guide seems to need revision.

 

[Kepler]

Yes I figured it would be a memory issue holding back on making FastRamp preset-specific. I actually find FastRamp a more important setting to be preset-specific then UpRamp because it is constantly in play when riding while on and off the throttle. UpRamp is important also but comes into play less because it only comes into play from a dead stop.

Other settings that would benefit from being preset-specific are PAS->MaxTrotSpd and Trq->AsstLevel. But of course if memory is already an issue I understand these are unlikely to change.

UpRamp I think is the parameter that needs to be considered most for possible inclusion as being preset-specific if memory allows.

The CA3 is so feature rich and I use just about every feature available on my setups. However, if memory does need to be free'd up to improve some of the functionality, personally I could live without the multiple battery profiles or perhaps just 2 battery preset profiles.

 

[teklektik]

UpRamp I think is the parameter that needs to be considered most for possible inclusion as being preset-specific if memory allows.

I'm thinking you meant FastRamp here....

Reducing the battery presets to two is creative thinking - if it gets down to this, it seems like a better than even trade for the majority of users. Of, course, with only one wired-in battery, I'm probably a little biased

 

[Kepler]

Yes I did mean UpRamp.

On another note, tried a heap more setting combinations for the Super Commuter with Torque PAS and managed to dial in the best setup to date. Still have couple of dropout on first start but much more stable once going with that annoying dropout all but gone. I will post up my settings later but for me it seems the key was to set Trq -> Asst Offset to a negative figure rather then positive. I have set this at -5 and in combination with my other settings, the PAS is far more responsive and stable.