Cycle Analyst V3 preview and first beta release

Get real world experience and user feedback on the electric bicycle products.
User avatar
justin_le   10 MW

10 MW
Posts: 2239
Joined: May 28 2007 3:27am
Location: Vancouver
Contact:

Re: Cycle Analyst V3 preview and first beta release

Post by justin_le » May 03 2012 1:14am

hjns wrote: Also, even with the temp log off on the Analogger, hooked onto the CA v3 beta, the Analogger shows me 3 to 5 additional collumns that I still need to identify. Having the CA send out a header when starting up, or sending out the summary stats with a header when shutting down would be a great help.
Right, the serial output stream spec hasn't been finalized, but at the moment it sends the following:

Ah, Volts, Amps, Speed, Distance, Temperature, RPM, HumanWatts, Torque(Nm)

There should be a header on the start of the datastream that goes:

Code: Select all

Ah     V      A      S      D      T      R      W      N
I know having a clock in the CA would be a bit too much, but if it could retrieve the date-time from the Analogger, and send it along in the CA dataset as timestamp, it would be great.
If you have the GPS analogger, then it should automatically time stamp all the current files in their filename, as explained in AppendixA here http://www.ebikes.ca/drainbrain/Analogg ... al_Web.pdf
Previously competed in the Suntrip race on a back to back tandem solar powered row/cycle trike. 550 watt solar roof, dual Grin All Axle hub motors, dual Phaserunner controllers, 12 LiGo batteries, and a whole wack of gear.

Now back in Vancouver learning to be a dad with my Big Dummy Frame (yes This One, thanks ES!) with GMAC 10T rear hub motor, Phaserunner controller, and 52V 19Ah EM3EV pack
My website: http://www.ebikes.ca
Please contact via email, info@ebikes.ca, rather than PMs, which are disabled

User avatar
hjns   100 kW

100 kW
Posts: 1273
Joined: Aug 20 2011 11:05pm
Location: Basel, Switzerland

Re: Cycle Analyst V3 preview and first beta release

Post by hjns » May 03 2012 6:35am

Thanks for providing the new firmware with 3-digit LVC. I fully support the request for speed limiting higher than 99.9km/h as well, if only in high current settings. Not that I am running really high currents, though, as with 126V 4A can get my HT to fluctuate between 99 and 103 km/h.
justin_le wrote:Right, the serial output stream spec hasn't been finalized, but at the moment it sends the following:
Ah, Volts, Amps, Speed, Distance, Temperature, RPM, HumanWatts, Torque(Nm)
There should be a header on the start of the datastream that goes:

Code: Select all

Ah     V      A      S      D      T      R      W      N
Thanks for the prompt response. Well, it should provide a header, but it doesn't. These are the first few lines that are being produced.
-0.0000 125.64 0.00 0.00 0.0591 -273.0 0.0 0 -10793 2999
-0.0000 125.64 0.00 0.00 0.0591 -273.0 0.0 0 -10793 3000
-0.0000 125.63 0.01 0.00 0.0591 -273.0 0.0 0 -10793 3000
-0.0000 125.64 0.01 0.00 0.0591 -273.0 etc
I will try again tonight. It does correspond nicely with Ah, V (30S), A, S, D, T (Celsius), RPM, humanW, and N...., and also shows I have no torque sensor nor thermistor attached as yet, so thanks for your answer. It saves me a lot of time. Maybe there is a setting in the CA that tells whether or not to send the header?
justin_le wrote:If you have the GPS analogger, then it should automatically time stamp all the current files in their filename, as explained in AppendixA here http://www.ebikes.ca/drainbrain/Analogg ... al_Web.pdf
Yes, I have, and that it does. However, that does not allow for linking individual rows of data to specific time points (or specific parts of the road), only the time of generation of the file is being recorded.

My idea was to record the CA data on a specific hilly road at a predefined speed driving for about 30 minutes, recording the changes in altitude using the GPS, so that I can correlate the increases in current and/or temp with the steepest incline that I encounter. Now, in order to do that, at this moment I need to assume that the CA log starts at the same timepoint as the first GPS date-time stamp, then calculate the difference in seconds between the time of start of recording until the time that I start with the hill, next multiply the number of seconds with the frequency of datatransmission, in order to calculate where the CA data is co-inciding with the hill. It works, but it is a bit fault prone due to the multiple derivations.

Anyway, the temp threshold - current limiting features should result in recordable differences in speed and/or motor/controller temp. Only the CA v3 beta can do that! So hurray!! This thing has got a very nice future!
Henk


All men dream: but not equally. Those who dream by night in the dusty recesses of their minds wake in the day to find that it was vanity: but the dreamers of the day are dangerous men, for they may act their dreams with open eyes, to make it possible. T.E. Lawrence

User avatar
teklektik   10 GW

10 GW
Posts: 4020
Joined: Mar 26 2011 1:15pm
Location: CT, USA

Re: Cycle Analyst V3 preview and first beta release

Post by teklektik » May 03 2012 9:50am

justin_le wrote:Also, the B13 code I just uploaded here is a bit different than the one I sent you by email yesterday which hadn't yet been finished, so when you get a chance update with this copy.
Thanks for the heads-up. Sort of stuck in a spate of rainy days anyhow... :(
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

User avatar
Doctorbass   100 GW

100 GW
Posts: 7434
Joined: Apr 08 2007 2:24am
Location: Quebec, Canada East
Contact:

Re: Cycle Analyst V3 preview and first beta release

Post by Doctorbass » May 03 2012 1:59pm

justin_le wrote:
Doctorbass wrote: Justin, Is it possible to get 3 unit digit for the max speed limit? ( >99km/h...) ex: instead of having 99.9km/h, we could choose 999km/h :wink:
Yes, it would be annoying for most people to have an extra digit to go through when setting max speed, so what I'll probably do is have it so that in high range mode you can set max speed to 3 digits with no decimal, or 0-999, while in low range mode the max speed is 3 digits with one decimal digit, or 0.0-99.9

Pretty much guaranteed that anyone going over 100 kph is running in the high range with kW display and <1mOhm shunt, so this should accommodate all I would hope!

-Justin

Thanks Justin!
CURRENT PROJECT: 2WD duo MXUS/Max-E ebike 32kW
http://endless-sphere.com/forums/viewto ... =3&t=65764
-Fastest speed record 117 km/h on flat
-Fastest 1/4 mile@ 114km/h on flat and 16.316 sec
-Succeded to haul a 19200 pounds schoolbus!
113kmh Gianthttp://www.evalbum.com/3406
Mongoosehttp://www.evalbum.com/1947
YOUTUBE---https://m.youtube.com/user/Doctorbasss
http://twitter.com/DocbassMelancon
I speak FRENCH and english

lizardboy   10 mW

10 mW
Posts: 34
Joined: May 31 2011 6:19pm

Re: Cycle Analyst V3 preview and first beta release

Post by lizardboy » May 03 2012 4:13pm

About the voltmeter readings someone reported on...

Hitting reset brings up a different voltage for me not much.. but every 2nd time or so the motor blips a bit and the diplayed voltage changes a few tenths of a volt. Sometimes there is no motor blip but the voltage changes a couple of tenths from before the reset.

It's like sometimes it pays attention to the v/v calibration and sometimes it doesn't. I have changed my v/v value from the factory setting and it seemes to ignore my changes on power up but if I reset it a few times the voltage drops and matches the calibration.

I'm not sure if theres a difference between the version you emailed me and the posted version but suddenly my throttle is cutting out the voltage input stays the same but the output just drops to the minimum output setting and stays there reactivating the throttle doesn't help if you hold the throttle in place after about 5 seconds it seems to start the ramp up.

I have the protection set at 1.5 volts higher so it isn't this. It does it in current, speed, passthru and disabled modes. all the wiring seems good. I'm thinking a voltage spike is activating the throttle protection but the real time screen doesn't show any spikes. It does show that the output drops back but I can't see why. The throttle slider on the main screen moves up but the throttle output voltage on the real time screen stays at the minimum.

To early to call this software but it only did it after I reloaded the firmware again but it could be unrelated seems intermitant I have times it works fine. but it makes me nervous to get far from home. It would be great if there was an on screen indication that the protection has been tripped.

UPDATE: reflashed with the emailed version and all is fine could be the reflashing reset something and it has nothing to do with reloading the software with the version posted here on ES
Last edited by lizardboy on May 04 2012 11:44am, edited 3 times in total.

User avatar
Doctorbass   100 GW

100 GW
Posts: 7434
Joined: Apr 08 2007 2:24am
Location: Quebec, Canada East
Contact:

Re: Cycle Analyst V3 preview and first beta release

Post by Doctorbass » May 04 2012 11:10am

What is the input impedense of the current sensing input?

I plan on using my C-A v3 with an external current sensur that is not a shunt but it agalvanic insulated sensor with a 0-5V output.

Ijust wonder if i could use a voltage divider to drop the output range to the +/-200mV range of the C-A.. unless it have changed and have a wider voltage range input?

The current sensor is the one on the zero motorcycle system and it also have a 2.5V ext voltage ref.. i dont know if i could use it to offset the output?


This is the exact model i have: http://media.digikey.com/pdf/Data%20She ... -600-S.pdf

it's the 200 model.. ( ± 600A)

The output is proportional to :
VOUT Analog Output voltage @ IP VOE ± (0.625. IP/ IPN) V

any idea Justin?
CURRENT PROJECT: 2WD duo MXUS/Max-E ebike 32kW
http://endless-sphere.com/forums/viewto ... =3&t=65764
-Fastest speed record 117 km/h on flat
-Fastest 1/4 mile@ 114km/h on flat and 16.316 sec
-Succeded to haul a 19200 pounds schoolbus!
113kmh Gianthttp://www.evalbum.com/3406
Mongoosehttp://www.evalbum.com/1947
YOUTUBE---https://m.youtube.com/user/Doctorbasss
http://twitter.com/DocbassMelancon
I speak FRENCH and english

User avatar
justin_le   10 MW

10 MW
Posts: 2239
Joined: May 28 2007 3:27am
Location: Vancouver
Contact:

Re: Cycle Analyst V3 preview and first beta release

Post by justin_le » May 04 2012 12:18pm

Doctorbass wrote:What is the input impedense of the current sensing input?
It's pretty high impedance
I plan on using my C-A v3 with an external current sensur that is not a shunt but it agalvanic insulated sensor with a 0-5V output.
Ijust wonder if i could use a voltage divider to drop the output range to the +/-200mV range of the C-A.. unless it have changed and have a wider voltage range input?
You can totally do that, just make sure that the common mode voltage stays between about -0.4 to +2 volts relative to ground. The input range of the CA is still +-200mV. The main issue with galvanically isolated hall current sensors is that they have problems with offset hysteresis and have nowhere near the dynamic range as a shunt resistor for accurately measuring Ah. So they can work OK for looking at and accumulating amps when you are riding under normal load, but if for instance you want to measure the current draw of a small load like your headlight etc. they'll be fairly inaccurate.

If you have one, try measuring the output with no current, then briefly apply full scale current through the sensor, and then measure the output with no current again. Repeat this after putting full scale negative current though the sensor.

With a shunt resistor, no current will always equal 0mV. But the datasheet on the LEM transducer you are using says up to 0.4% offset error after exposure to the full current (Ipm) of 600A, which means the amperage shown with zero current can vary by +-2.4 amps or so.

-Justin
Previously competed in the Suntrip race on a back to back tandem solar powered row/cycle trike. 550 watt solar roof, dual Grin All Axle hub motors, dual Phaserunner controllers, 12 LiGo batteries, and a whole wack of gear.

Now back in Vancouver learning to be a dad with my Big Dummy Frame (yes This One, thanks ES!) with GMAC 10T rear hub motor, Phaserunner controller, and 52V 19Ah EM3EV pack
My website: http://www.ebikes.ca
Please contact via email, info@ebikes.ca, rather than PMs, which are disabled

User avatar
Doctorbass   100 GW

100 GW
Posts: 7434
Joined: Apr 08 2007 2:24am
Location: Quebec, Canada East
Contact:

Re: Cycle Analyst V3 preview and first beta release

Post by Doctorbass » May 04 2012 11:00pm

Thanks Justin i now 100% understand your point.

well adding a shunt seem now like the easyest option :wink:

Doc
CURRENT PROJECT: 2WD duo MXUS/Max-E ebike 32kW
http://endless-sphere.com/forums/viewto ... =3&t=65764
-Fastest speed record 117 km/h on flat
-Fastest 1/4 mile@ 114km/h on flat and 16.316 sec
-Succeded to haul a 19200 pounds schoolbus!
113kmh Gianthttp://www.evalbum.com/3406
Mongoosehttp://www.evalbum.com/1947
YOUTUBE---https://m.youtube.com/user/Doctorbasss
http://twitter.com/DocbassMelancon
I speak FRENCH and english

User avatar
justin_le   10 MW

10 MW
Posts: 2239
Joined: May 28 2007 3:27am
Location: Vancouver
Contact:

Re: Cycle Analyst V3 preview and first beta release

Post by justin_le » May 05 2012 1:03am

lizardboy wrote: I'm not sure if theres a difference between the version you emailed me and the posted version but suddenly my throttle is cutting out the voltage input stays the same but the output just drops to the minimum output setting and stays there reactivating the throttle doesn't help if you hold the throttle in place after about 5 seconds it seems to start the ramp up.
Hey Lizardboy, I think I know what might have been going on there. If you do reflash the posted code on ES try changing the Watts Gain from 200 (which would have been the default setting) down to about 50 or so and see if you still have that issue with the throttle ramp. -Justin
Previously competed in the Suntrip race on a back to back tandem solar powered row/cycle trike. 550 watt solar roof, dual Grin All Axle hub motors, dual Phaserunner controllers, 12 LiGo batteries, and a whole wack of gear.

Now back in Vancouver learning to be a dad with my Big Dummy Frame (yes This One, thanks ES!) with GMAC 10T rear hub motor, Phaserunner controller, and 52V 19Ah EM3EV pack
My website: http://www.ebikes.ca
Please contact via email, info@ebikes.ca, rather than PMs, which are disabled

lizardboy   10 mW

10 mW
Posts: 34
Joined: May 31 2011 6:19pm

Re: Cycle Analyst V3 preview and first beta release

Post by lizardboy » May 05 2012 1:14pm

okay I gave that a try I reinstalled the ES version and it's exactly the same I reduced the Wgain to 50 and to 10 but no difference at all. I reinstalled the version you emailed me with gain set to 500 as it defaulted and it works perfectly.

The throttle slider on the main display stays at full when the signal fails and goes up and down in response to the throttle. There is just the minimum output voltage when I look in the real time screen. My understanding was that the throttle slider reflected the throttle output voltage. What conditions could cause the output slider and the output voltage in the real time screen to be different?

User avatar
justin_le   10 MW

10 MW
Posts: 2239
Joined: May 28 2007 3:27am
Location: Vancouver
Contact:

Re: Cycle Analyst V3 preview and first beta release

Post by justin_le » May 05 2012 4:09pm

lizardboy wrote: The throttle slider on the main display stays at full when the signal fails and goes up and down in response to the throttle. There is just the minimum output voltage when I look in the real time screen. My understanding was that the throttle slider reflected the throttle output voltage. What conditions could cause the output slider and the output voltage in the real time screen to be different?
No, it's the input throttle voltage that is shown on the slider. For the output signal, go to the diagnostic screen that shows Vi and Vo directly in units of voltage. We've got the same firmware that I posted here running on 3 different ebikes in the shop and have tried to replicate the behaviour that you describe but without luck, they all work fine. Could you list exactly what settings you had changed from the default?

-Justin
Previously competed in the Suntrip race on a back to back tandem solar powered row/cycle trike. 550 watt solar roof, dual Grin All Axle hub motors, dual Phaserunner controllers, 12 LiGo batteries, and a whole wack of gear.

Now back in Vancouver learning to be a dad with my Big Dummy Frame (yes This One, thanks ES!) with GMAC 10T rear hub motor, Phaserunner controller, and 52V 19Ah EM3EV pack
My website: http://www.ebikes.ca
Please contact via email, info@ebikes.ca, rather than PMs, which are disabled

User avatar
teklektik   10 GW

10 GW
Posts: 4020
Joined: Mar 26 2011 1:15pm
Location: CT, USA

Re: Cycle Analyst V3 preview and first beta release

Post by teklektik » May 05 2012 4:42pm

justin_le wrote:Could you list exactly what settings you had changed from the default?
The v3B13 Setup Summary post has been updated to list the default values.
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

lizardboy   10 mW

10 mW
Posts: 34
Joined: May 31 2011 6:19pm

Re: Cycle Analyst V3 preview and first beta release

Post by lizardboy » May 05 2012 5:49pm

Well if thats how the slider works then I'm really confused.. When I look in the real time screen the input voltage is there and moves with the throttle just fine but the output does not move. Which makes sense with the dead throttle I'm experiencing. When I watch during a cut off the voltage goes down to the minimum right away as it happens. LIke you blew a fuse. When I ride with the email version there is a pronounced down ramp that I'm trying to make shorter

Okay heres my settings:

Setup Calibration

Cal -> Range
{ [Lo (W)] |
Cal -> RShunt
[5.122] mOhm
Cal -> V Scale
[31.09] V/V
Cal -> Zero Amps
(Press/Hold to normalize the currently detected Amp reading to 0.0)
-2.45 -2.48
Setup Speedometer

Spd -> Units
{ [km] }
Spd -> Wheel
[2019] mm
Spd -> #Poles
[1]
Spd ->TotDist
[02186] km
Setup Speed Lims

SLim -> Max Speed
[40.0] kph
SLim -> Start Speed
[00.0] kph
SLim -> IntSGain
[200] Gain
SLim -> PSGain
[0.59] V/kph
SLim -> DSGain
[002] Gain
Setup Power Lims

Plim -> Max Current
[45.0] Amps
Plim -> AGain
[150] Gain
Plim -> Max Power
[1500] Watts
Plim -> W Gain
[50] Gain
Setup Throt In

ThrI -> Cntrl Mode
| Current |
ThrI -> Min Input
[0.86] Volts
ThrI -> Max Input
[3.68] Volts
ThrI -> Fault Volt
[4.49] Volts
Setup Throt Out

ThrO -> Output Mode
{ [Voltage] }
(if ThrO->OutputMode = { Voltage }

ThrO -> Min Output
[1.20] Volts
ThrO -> Max Output
[3.95] Volts
)
ThrO -> Up Ramp
[500]
ThrO -> Down Ramp
[200]
ThrO -> KV Comp.
[0.95] V/kph
Setup Battery

Batt -> Chemistry
{ LiPo }
Batt -> String#
[10] Cells
Batt -> Capacity
[9.0] Ah
Batt -> RBatt
[238] mOhm
Batt -> Vlt Cutoff
[31.5] Volts
Batt -> V Gain
[0800] Gain
Batt -> TotCyc
[0132] Cyc
Batt -> TotAhrs
[00768] Ah
Setup Display

Disp -> Main Disp
{ [Watts] }
Disp -> Averaging
[5] Duration
Disp -> RS232
{ [1] | 5 } Hz
Disp -> Vshutdown
[10.0] Volts

So when this happens the throttle works fine holding the bike in the air or on the stand but as soon as i head out of the driveway and up hill it cuts off usually at around 300 watts but it could be more time related as keeping power below this level doesnt help either.(i'm using current throttle) Backing off or retwisting the throttle does nothing. If you retwist and hold it will come on in about five seconds but cuts out a second or two later cycling the power gets you back to where it runs a few seconds. It can't be a ramp issue as the throttle responds right away initially (if it works off the bat). And there is no down ramp when it happens it just cuts off right away. it seems exactly like the output voltage is too high and it's tripping the throttle short protection.
Initially when I was adjusting the throttle parameters (on the Email version) I had a similar problem the first time out. adjusting the max throttle voltage lower (it was already below the protection level) solved it. I tried this with the new version made no difference. The voltage in the realtime screen doesn't move from 1.20 volts so I don't think this is it.

All I can tell you is putting all the above settings into the CA for the Email version works fine beautiful even but try it on the ES version and the problem happens everytime with same settings in both . (I left the watt gain at 500 in the email version)

I'm currently working fine with the emailed version but I'd love to get this sorted. Hope the info helps.
Last edited by lizardboy on May 05 2012 9:50pm, edited 1 time in total.

User avatar
teklektik   10 GW

10 GW
Posts: 4020
Joined: Mar 26 2011 1:15pm
Location: CT, USA

Re: Cycle Analyst V3 preview and first beta release

Post by teklektik » May 05 2012 7:41pm

I don't want to interrupt or confuse the debug session here, but until Justin gets back on the thread with more informed advice: have you tried setting ThrO->{Up/Down}Ramp to 999?

You have a very small (020) down ramp setting so the delay effect will be almost as long as possible (001). I presently have these turned off (999) since they seemed to be causing surgey behavior. Setting these to 999 will get them out of the equation, and you can see the effects of Plim->AGain in a simpler environment.

You might also try running Plim->MaxCurrent and SLim->MaxSpeed up to very large values to guarantee that they are not participating. This should help isolate and identify the troublesome limiting parameters (probably Plim->MaxCurrent and Plim->AGain) and ensure that any effects you experience are due (more or less) exclusively to the associated limiting logic.
Last edited by teklektik on May 05 2012 9:53pm, edited 1 time in total.
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

ashwright   100 mW

100 mW
Posts: 42
Joined: Oct 01 2011 3:33am
Location: Brisbane, Australia

Re: Cycle Analyst V3 preview and first beta release

Post by ashwright » May 05 2012 8:49pm

Can the Aux input be used to control max throttle out?

I am trying to think of a way of using the new CA, to control the performance of a PAS, mid mount bike. So you can use the three way switch on the handle bars to say, Pedestrian mode (10%, 5kph, very slow accell), bike mode (70%, 30kph, medium accell) and road mode (100%, fast accell).

My current bike has a PAS controller, with a three way switch, which seems to have this behaviour. But most newer good controllers don't have this, and it looks like the new CA would be perfect, with a basic, cheap, high performance controller.


Would the be any other way of using the CA in this way?
I am not sure a current limit would get what I want. Usually around pedestrians, I am on flat ground, in a low gear, and so it does not take too much current to accelerate fast (which is bad, as pedestrians are very unpredictable).

User avatar
teklektik   10 GW

10 GW
Posts: 4020
Joined: Mar 26 2011 1:15pm
Location: CT, USA

Re: Cycle Analyst V3 preview and first beta release

Post by teklektik » May 05 2012 9:33pm

ashwright wrote:Can the Aux input be used to control max throttle out?
Yes. I am presently using the Vaux input to provide three power levels using current limiting (Ctrl->AuxFunct = AmpsLim). I have a little post with an Excel resistor calculator that will go up shortly to illustrate this using an off-the-shelf Crystalyte LMH switch.
ashwright wrote:I am trying to think of a way of using the new CA, to control the performance of a PAS, mid mount bike. So you can use the three way switch on the handle bars to say, Pedestrian mode (10%, 5kph, very slow accell), bike mode (70%, 30kph, medium accell) and road mode (100%, fast accell).

My current bike has a PAS controller, with a three way switch, which seems to have this behaviour. But most newer good controllers don't have this, and it looks like the new CA would be perfect, with a basic, cheap, high performance controller.

Would there be any other way of using the CA in this way?
Justin will need to verify, but it appears that there is mode that will do exactly what you want with Ctrl->AuxFunct = PasLevel. If you check the Setup summary, you can see two other associated parameters (Ctrl->PASMode and Ctrl->AssistLevel) that appear to determine operation at the 100% value for the POT pad (Vaux) on the CA PCB.

As with Ctrl->AuxFunct set to any of the other possible control modes, it remains only to use a resistor divider to obtain a couple of lower percentage values of the voltage range Ctrl->MinAuxIn to Ctrl->MaxAuxIn and to select one of these fixed voltages with a switch. One possible configuration might use a SPDT on-off-on switch arranged like this:
CA_v3_LMH_fixedR.gif
CA_v3_LMH_fixedR.gif (4.48 KiB) Viewed 5828 times
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

lizardboy   10 mW

10 mW
Posts: 34
Joined: May 31 2011 6:19pm

Re: Cycle Analyst V3 preview and first beta release

Post by lizardboy » May 06 2012 10:44am

Thanks Teklektik;

I'll give that a try today. It isn't like limiting at all it'a complete cutoff..

On another note why doesn't setting the throttle to pass-thru or disabled bypass this problem? It would be great if one of these modes was just a limp home diagnostic type bypass or pass through. So that what came in the throttle input went out the throttle output exactly with no transformation limiting or compensation.

I didn't find out about my current problem till I rode down a really big hill. Would have been great to have been able to set to pass thru and have it just ignore whatever setting it is that's wrong and just get me home. The drive by wire throttle needs a work around at least for me. If I smash the CA off on a tree or something It would be game over for propulsion. I think I'll add a micro size switch and do a hard bypass for the one signal line involved.

I'll report back on the changed settings. I'll try and turn every feature down or off one at a time and alter every parameter. The hard part is I have to go up the hill everytime to test it as it won't do it on the stand.

I'll let you know

User avatar
teklektik   10 GW

10 GW
Posts: 4020
Joined: Mar 26 2011 1:15pm
Location: CT, USA

Re: Cycle Analyst V3 preview and first beta release

Post by teklektik » May 06 2012 12:09pm

lizardboy wrote:I'll give that a try today. It isn't like limiting at all it's a complete cutoff...
The idea is not to go into this with a preconceived notion of how it will fail or behave regardless of the names of the parameters - just to gather some information for Justin that involves the smallest number of interacting control features as possible. If this first test works fine, you can add in ramping, etc one at a time until you have a clear "I did this and suddenly it misbehaved". This may help localize the issue whether it's code or our (mis)understanding of parameter interactions that he can clarify.
lizardboy wrote:On another note why doesn't setting the throttle to pass-thru or disabled bypass this problem?
ThrI->CntrlMode = Pass-thru bypasses the closed-loop throttle feedback logic utilized by ThrI->CntrlMode = {Current|Speed} modes so you are running in open-loop mode - pretty much what goes in, goes out without embellishment save for ramping (and throttle voltage scaling). I believe that with ramping disabled, there should be no effects on the throttle save for the max limiting parameters (amps, watts, speed) much like v2.23. When I make these settings, the v3 behaves identically to the old unit.
lizardboy wrote:It would be great if one of these modes was just a limp home diagnostic type bypass or pass through. ... The drive by wire throttle needs a work around at least for me. If I smash the CA off on a tree or something It would be game over for propulsion. I think I'll add a micro size switch and do a hard bypass for the one signal line involved.
I had the same concern about a single point of failure and so added a second JST3 connector under the dashboard that runs my controllers directly as originally intended. All three throttle connections (Gnd, +5, Sense) are sourced from the controllers - nothing from the CA. The normal CA-throttle JST3 gets its three connections from the associated pads on the CA PCB instead. Eliminating common connections avoids ground loops and allows the CA to be completely unplugged/removed without affecting the backup mode e.g. 'tree smash'.
Last edited by teklektik on May 08 2012 6:26pm, edited 1 time in total.
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

User avatar
teklektik   10 GW

10 GW
Posts: 4020
Joined: Mar 26 2011 1:15pm
Location: CT, USA

Re: Cycle Analyst V3 preview and first beta release

Post by teklektik » May 06 2012 1:23pm

lizardboy wrote:It isn't like limiting at all it's a complete cutoff...
Rain finally stopped and I only had limited time yesterday before dark to try v3b13 - worked okay. It's sunny today so I'm doing more tests.

But - I am experiencing the same failure mode with the throttle cutout using ThrI->CntrlMode = Current. The symptoms come and go. I can get into a state where it works perfectly, then get into a state where it cuts out abruptly.
...
More tests, etc, but I thought I'd get this posted so lizardboy would know he has company :D

EDIT-

Justin-

Tests were all done heading up a hill with two BMC V2S motors, 66v20Ah Headway, 2 12FET Xlyte analog controllers, and Ctrl->AuxFunct = AmpsLim with Vaux set to 3.0v (3.0v/5.0v = 60% of Plim->MaxCurrent = 50 A yielding an effective MaxCurrent = 30 A).

After a bit more testing, it appears that the surging/cutout failure lizardboy reported is random on power-up making this look like an uninitialized variable. If it does start in the failure mode where it is surging badly (cutting out, zipping fwd, ...), I seem to be able to get it out of that mode by running down a short incline giving it throttle so the surging stops and all runs okay in the downhill. Then I cut the throttle to zero, stop, and the symptoms stay gone until another power cycle occurs. There is no CA button-pushing whatsoever between working perfectly, power cycling the bike, and having the failure (possibly) re-occur. I didn't test other action sequences beyond the roll/throttle/stop sequence - that magic works consistently, so I left it at that... it seemed that once it was heading downhill and it was able to achieve control w/o overshoot, it got the ability to do so in any situation.

Once the CA is happy and running properly, my ramping parameters seem to work properly which leads me to believe that the surging I saw with them previously was related to this other issue (just a guess). I didn't do very much with the ramping beyond seeing that non-999 values did not precipitate surging.

Here's my setup:
CA_vB13a_ConfigSettings_teklektik_20120506_1510.zip
(1.22 KiB) Downloaded 164 times
Last edited by teklektik on May 09 2012 3:25pm, edited 1 time in total.
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

lizardboy   10 mW

10 mW
Posts: 34
Joined: May 31 2011 6:19pm

Re: Cycle Analyst V3 preview and first beta release

Post by lizardboy » May 06 2012 5:16pm

Well it is good to know I'm not alone;

My enviroment is all hills so I may have run into this a bit sooner than other people. Not trying to have preconcieved ideas here. But I'm an out of the box thinker in troubleshooting and I tend to follow a different pattern to other people. My understanding of how the CA works is evolving as we go. Seems exactly like a variable being initiated twice.

You'll notice that the voltmeter stuff I was talking about earlier only happens when the random variable problem comes up. reset like five times and the voltmeter drops the 0.2 volts(for me) that I expect for the calibration value and then the current throttle works no problem until you cycle the power.

Should be a simple fix I'm so glad you stayed on this... Takes a bit of hill to occur. Thanks for the tip about the ramp values I was decreasing the ramp value to shorten the ramp as this seemed logical but it's the other way around. Love that feature and the current throttle. My motor will last eons longer this way. I can ease on the power instead of every throtlle change being a full power speed change.

Now on to temp sensing, Did a few massive climbs of almost 2000 feet of vert. today Motor was fine but the controller was hot so I may start by sensing the controller. It's kind of the mine canary in my system The weather has been stellar today already came home and charged up again 4 times.

Because I wired my throttle using 2 wires out of the DP bundle. one to send the throttle signal to the CA and one to take it back I'm just going to mount the hard switch on the end cover of the controller. Easy bypass one hole required to mount the switch.

Tench   100 kW

100 kW
Posts: 1033
Joined: Jul 13 2010 12:48pm
Location: Derby UK

Re: Cycle Analyst V3 preview and first beta release

Post by Tench » May 06 2012 6:46pm

I have just been catching up on this thread and the cutting out problems others are experiencing, mine is currently running in current mode with the ramp up set to 100, everything is working well except i can turn it on and off 5 times and see several different voltages for the main pack on the main screen, as much as .4v difference, sometimes i get a blip from the motor on start up too and also when the reset is pressed. when a different voltage is displayed on the main screen the min in and min out preset voltages also change a little, it occured to me that this maybe related to the cutting out some are experiencing, when the throttle voltage settings change if they are set very close to the controllers starting voltage it could trigger the over ride to prevent an open throttle start? ie; if the throttle output voltage is above the controllers start point on start up the controller will not activate the motor, closing the throttle to reduce the voltage below the controllers min activation voltage will reset it, just a thought chaps, lower your min output voltage a little more to see if this is your problem.

Simon.
Project one http://endless-sphere.com/forums/viewto ... ilit=tench
Project Two http://endless-sphere.com/forums/viewto ... =6&t=37489
The uk's first Stealth Bomber

There is a box inside your head, inside amongst others are the words "Wont fit" stop looking for answers in the box, there are no answers only excuses.

User avatar
teklektik   10 GW

10 GW
Posts: 4020
Joined: Mar 26 2011 1:15pm
Location: CT, USA

Re: Cycle Analyst V3 preview and first beta release

Post by teklektik » May 06 2012 7:42pm

Tench wrote:...when the throttle voltage settings change if they are set very close to the controllers starting voltage it could trigger the over ride to prevent an open throttle start?
My controllers do not that functionality.
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

User avatar
teklektik   10 GW

10 GW
Posts: 4020
Joined: Mar 26 2011 1:15pm
Location: CT, USA

Re: Cycle Analyst V3 preview and first beta release

Post by teklektik » May 08 2012 12:34am

With the cutout/surge issue identified and a solution in hand to get out of that mode, I moved on to other experiments and testing today. The bike was configured for current throttle, power (Watt) limiting, and used both up and down ramping. The Vaux input was configured as a three level LMH power limiting (Watt) selector. Unlike v3B12 where I could not get Power Limiting to work even with WGain=1, v3B13 works pretty well although I cannot use values above WGain=15 without getting dramatic surging from a dead stop.

Settings:
CA_vB13a_ConfigSettings_teklektik_20120507_1720.pdf
(14.35 KiB) Downloaded 202 times
Tweaking complete, I went for a 25 mile ride with minimal pedaling. This was one of my usual trips with speeds varying from 12mph to 40mph - part rolling hilly roads, part paved bike paths.

Once I got it out of the initial power-up cutoff/surge mode, the bike ran very well for the remainder of the ride. Startups were smooth and the power limiting worked very slick delivering the prescribed power levels even with battery sag due to discharge or heavy two motor loads. Switching the Vaux levels on the fly worked smoothly with no appreciable delays as new higher/lower power levels took effect.

I had some very minimal 'hunting' on getaways but additional WGain adjustments should address that. Although I could not feel it, there were some rapid power oscillations visible on the CA Main Display in certain circumstances as the CA attempted to home in on a target power level while underway climbing hills - the average power came out okay, and the hunting was rapid enough that there was no rider sensation of the action. This may be related to running at Vaux power levels different than the one used to tweak the WGain setting. Again, this was an observation of CA displayed values, but the ride itself was fine. Some additional WGain adjustment may lessen this effect.

Someplace in the middle of the ride, I switched back to Pass-thru Throttle because the Current Throttle was coming on too strong and I could not feather it to re-engage the load while rolling - (although it worked well from a dead stop). This is related to the planned but as yet unimplemented logic allowing the CA to accomplish more gentle re-engagements at speed. Until that feature appears, manual control (for me) offers less stress on the gear motors.

I need to test Speed Limiting, but that will need to wait until after the next few days of rain.

Meanwhile, the unit is running well with the exception of the cutout/surge bug and perhaps the low WGain scaling (015). The rider experience on getaways is much improved over v2.23 and the Power Limiting in lieu of Current Limiting keeps the bike peppy and responsive right up until max DOD is reached. The new ThrI->FaultVolt feature to auto-shutdown on open (runaway) throttle works very nicely and is a welcome addition. The earlier issue with throttle burp on Reset has not re-occurred.
Last edited by teklektik on May 09 2012 3:23pm, edited 2 times in total.
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

User avatar
teklektik   10 GW

10 GW
Posts: 4020
Joined: Mar 26 2011 1:15pm
Location: CT, USA

Re: Cycle Analyst V3 preview and first beta release

Post by teklektik » May 08 2012 12:46am

Justin-

Although this would affect a noteworthy change in the operator interface over earlier versions, I would really recommend switching the top-level button functionality so that the left button invokes Reset and the right button invokes Setup. With only minor exception, the entirety of the Setup procedure utilizes the right button and it seems counter-intuitive to use the left button to initiate the Setup procedure then to rely on the other button to actually carry out all the configuration tasks.

...I have accidentally Reset my CA many times this last week while trying to reconfigure a parameter... :oops:
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

lizardboy   10 mW

10 mW
Posts: 34
Joined: May 31 2011 6:19pm

Re: Cycle Analyst V3 preview and first beta release

Post by lizardboy » May 08 2012 11:44am

For me it's that I can't see the ok when it comes up because my right hand is blocking just a bit of that edge of the screen but not a big deal just have to get used to a new hand position.

Just to confirm the battery gauge only works when the throttle is released and shows full all other times correct? It also seems to hardly move only losing a few pixels by halfway discharged then catches up rapidly on the bottom part of the discharge curve. So far it's way better to just do the math in my head and subtract total ah from consumed a./h could be because I'm still running the emailed version as the posted version is useless for me currently. But if there is still work to do there I understand and I'll be patient. But otherwise I'd just like it to be run by the A/H totals cause thats what I'm doing when I look down anyway and I'm not sure what the advantages are for going with some other system. ROde my pack dead on a big uphill last night and the gauge didn't move from 7/8's full till the last 200 yards when the battery died (then it showed empty).

I've also forgotten to reset (many times wish that was automatic somehow) and even after running to 150% of what the pack capacity should have been the battery gauge still told me I had 7/8's of a tank. So clearly it's not looking at ah anymore. Maybe internal resistance is set wrong. I used my icharger's built in special function to measure the pack resistance. Is this wrong?

And also maybe I'm missing it but the second screen used to have a spot that showed the amps , I set the main screen to watts and I'd like to check the amps without having to change any settings. Is one of the realtime values amps? or am I just missing the screen that shows amps? I'd really love to have the first screen the way it is although I'd like the throttle slider to be switchable to show me what the output of the throttle is. and I'd like the second screen to show me everything else I need while riding like.... the temperature readings and the amp draw and maybe an estimate of distance left at the current wh/km sort of two screens easy to flip through. And how about an adjustable time setting that would default back to the main screen after a set amount of time. It would save having to have your eyes off the road to return to the main screen.
Last edited by lizardboy on May 08 2012 7:09pm, edited 2 times in total.

Post Reply