Cycle Analyst V3 preview and first beta release

Get real world experience and user feedback on the electric bicycle products.
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 » Jan 07 2014 8:56pm

Another thought:
  • What kind of throttle do you have? (hall? Magura?)
  • What are your ThrI->MaxInput and ThrI->FaultVolt settings?
Maybe a little noise is pushing you into an input fault cutout....
These are the only remaining settings that I can think of that are uniquely in play when you are running flat out....
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

User avatar
Merlin   100 kW

100 kW
Posts: 1483
Joined: Jan 29 2013 10:00am
Location: europe

Re: Cycle Analyst V3 preview and first beta release

Post by Merlin » Jan 08 2014 1:58pm

on normal Cruising is not easy to say.
on daily to work i drive only with 2,5kw (throttle sprints) other time just 750w with pedaling....on out outside temps bout 5°c there comes not anymore then 40°c ...
driving with more fun i get staying 100°c....
pushing the limits i can squeeze out 10kw on accelerating ...running 8kw till 80-90kph..... and needing 5-6kw for tucked cruising on flat 100kph
With Outside temps 3-5°c i have after 5-6km reached 150°

(my limits are 120/150) because i have the tempprobe direct on one phase to windigs. in my opinion it is the hottest point.
after 150°c reading in my ca you have after 5 min only 90° still standing bike....
after reopened there was no black or brown melting....even the winding "sails" was ok (donno the right word :P)

vmin(77v) cant kick in because i have it 10v under my real LVC. (87v)
...because i want "full power" till LVC....i know to watch carefull about the voltage. but i look all the time to check voltage. no Prob for me.


I have a Hall Thumb throttle (but i like to have a magura but in germany i cant find/get one) since yesterday.
now i have a hall full throttle to test. (i get problems with my wrist cruising an riding all the time standing on my bike and using reversed thumbthrottle) now i try the cheap china lyen throttle....(and found another PRoblem ...i think i need a throttle mapper or something, cos it is hard to cruise on half thottle....MINIMAL movings on wrist the bike shoots forward (or get off throttle) that sucks a lot.

My throttle in and out and fail are: 0,83/4,24 /4,49

0,81 would be better for the full-throttle grip...but my pas dont start at 0,81...on 0,82 the PAS has 10 cutouts on 20meter
0,83 works.

maybe set the thottle fault on 4,79?


User avatar
Merlin   100 kW

100 kW
Posts: 1483
Joined: Jan 29 2013 10:00am
Location: europe

Re: Cycle Analyst V3 preview and first beta release

Post by Merlin » Jan 08 2014 4:53pm

teklektik wrote:
hjns wrote:
  • ramp values of V/sec takes some adjusting, but is actually more intuitive. I changed everything to 99.99V/sec for immediate effect. Will report my experiences.
You can transfer your old ramp values by taking the inverse to convert sec/v to v/sec. (Press the "1/x" key on your calculator). For example:

If you previously had a P5 ramp value of 1.25sec/v,
the revised P6 value should be 1/(1.25sec/v) = 0.80 v/sec.

...wished to find that before try and error these settings.....3200km and i never give a 100v/100a full Throttle from dead stop....
that was my first ebike backflip ever ...one mirror, a click-fix adapter, one fingernail, knee and elbow hurts.....

my full speed throttle test i will try another day *ouch :cry: :lol:

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 » Jan 09 2014 2:25pm

Merlin wrote:(my limits are 120/150) because i have the tempprobe direct on one phase to windigs. in my opinion it is the hottest point.
after 150°c reading in my ca you have after 5 min only 90° still standing bike.... after reopened there was no black or brown melting....
...
vmin(77v) cant kick in because i have it 10v under my real LVC. (87v)
...
My throttle in and out and fail are: 0,83/4,24 /4,49
Ok - well - it doesn't sound like temp, LVC, or throttle faults are causing the cutouts... seemed unlikely, but had to ask.
  • I might note though that silicon devices (halls) don't like 150degC. Justin's simulator considers 150degC to be 'overheated' although there are no doubt anecdotal forum posts about running hotter. Not saying one way or the other, just making mention of some stuff...
I'm flat out of ideas on this. I see that you have PAS - the only thing I can think of at this point is to disable PAS just to rule it out completely. Short of getting your whole configuration to see if there is anything odd, I think you are going to have to write to ebikes.ca and pursue the matter with Justin and the Support guys. This may be something simple, but I'm not having the big 'Ah-ha!'...
Merlin wrote: I have a Hall Thumb throttle (but i like to have a magura but in germany i cant find/get one) since yesterday.
now i have a hall full throttle to test. now i try the cheap china lyen throttle....(and found another PRoblem ...i think i need a throttle mapper or something, cos it is hard to cruise on half thottle....MINIMAL movings on wrist the bike shoots forward (or get off throttle) that sucks a lot.
My knee-jerk reaction is to have you switch to Current Throttle which will directly tie the controller current to throttle voltage - typically nice and controllable. On the other hand, if you have a junk throttle with really dreadful linearity, the improvement may not fulfill you expectations. For me, Current Throttle alone is enough reason to get a V3 - there's that much improvement. :)
Merlin wrote:maybe set the thottle fault on 4,79?
Ya - you could set it higher - maybe 4.9v. This might improve things if the max throttle voltage is jittery for some unknown reason, but what you have should work as it is.
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 » Jan 09 2014 2:43pm

Merlin wrote:whats about justin? is he still alive? working on CA v4 huh? ;P
Justin had scheduled a busy year in 2013 and the delays in getting PAS working properly for the projected 3.0 production release date were a real scheduling setback (we got all the way to P6 after the planned release of P1). He had to move on to address the other business matters that had slid because of the extended V3 effort. The V3 P6 release seemed stable so new V3 firmware work was of reduced priority.

In any case, if you've been following ebikes.ca developments, you know there have been many new product developments and releases - new programmable charger, a couple of new motors, StokeMonkey Kit, and some V3 plug-and-play accessories so the advanced features are not so DIY. Recently he has been working hard on the V3 'boot issue'. I know he planned that v3.1 would focus on display improvements - new functionality and improved configurability, but I'm not sure about a new V3 firmware schedule. I know getting back to this has been on his mind and perhaps with other new products now released he will have an opportunity to do so.
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

User avatar
izeman   10 GW

10 GW
Posts: 4749
Joined: Jun 21 2011 8:25am
Location: vienna, austria
Contact:

Re: Cycle Analyst V3 preview and first beta release

Post by izeman » Jan 10 2014 4:23pm

does the CA support ebrake switches that are NC as well? i'd like to use a NO reed switch which is closed when the lever is NOT pulled. so if it's pulled the switch opens instead of closes (what CA expects afaik). any chance to use them as well?

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 » Jan 10 2014 8:48pm

izeman wrote:does the CA support ebrake switches that are NC as well? i'd like to use a NO reed switch which is closed when the lever is NOT pulled. so if it's pulled the switch opens instead of closes (what CA expects afaik). any chance to use them as well?
Well, the V3 does not normally work with that type switch but you could make a small wiring change to make it possible. We just need to use a pulldown resistor that will drop the uproc port sufficiently low considering the worst case internal port 15K pullup resistor and the extra 1K resistor in series with the uproc input.

The switch normally is hooked across the PCB Ebk and G pads. Instead, hook it across Ebk and +5v and attach a 3.3K pulldown resistor on the PCB across Ebk and G.
  • With the switch NC, the CA will see +5v = no-ebrake. The 3.3v pulldown resistor will drain an extra 1.5ma from the CA +5v supply.
  • When you open the switch, the pulldown will send the uproc input port (worst case) - no higher than 5v * (3.3K+1K)/(15K+1K+3.3K) = 1.11v which the CA will consider 'asserted' (input < 1.5v).
I haven't tried it, but it should work without difficulty.
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

User avatar
izeman   10 GW

10 GW
Posts: 4749
Joined: Jun 21 2011 8:25am
Location: vienna, austria
Contact:

Re: Cycle Analyst V3 preview and first beta release

Post by izeman » Jan 11 2014 4:30am

thanks teklektik i will try that. but it sounds reasonable so it should work :)
will need to order another v3 for my next bike.

sacko   10 kW

10 kW
Posts: 816
Joined: Jul 04 2012 2:15am
Location: Kent, United Kingdom

Re: Cycle Analyst V3 preview and first beta release

Post by sacko » Jan 13 2014 1:59pm

Just ordered this temp sensor;

http://www.ebay.co.uk/itm/261240040694

Just wanted to check it was the right one before I install it.

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 » Jan 13 2014 2:47pm

With this one you wont need to mod your CA: http://www.ebay.co.uk/itm/100-Pcs-10K-O ... 43bbc358b8
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

sacko   10 kW

10 kW
Posts: 816
Joined: Jul 04 2012 2:15am
Location: Kent, United Kingdom

Re: Cycle Analyst V3 preview and first beta release

Post by sacko » Jan 13 2014 2:50pm

Brilliant thanks :)

This one ok?

http://www.ebay.co.uk/itm/390614553726? ... 1423.l2649

...quicker delivery from the UK.

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 » Jan 13 2014 5:06pm

Yep - good choice.
+1 for hjns's NTC recommendation - much simpler. :D
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

sacko   10 kW

10 kW
Posts: 816
Joined: Jul 04 2012 2:15am
Location: Kent, United Kingdom

Re: Cycle Analyst V3 preview and first beta release

Post by sacko » Jan 13 2014 5:14pm

Excellent, thanks for the clarification :)

icecube57   1 GW

1 GW
Posts: 3072
Joined: Apr 25 2008 8:42am
Location: Austell GA
Contact:

Re: Cycle Analyst V3 preview and first beta release

Post by icecube57 » Jan 16 2014 7:44am

ok Im making this post since there is absolutely no documentation for the tdcm torque sensing bottom bracket. I have mines installed. Zeroed the off set. I do register pedal pressure and the voltage changes. Its set to 100nm/v. Poles set to 12 and averaging is 24. In Pas mode it is detecting pedal strokes and offering assist based on rpm which is nice and all. But i really wanted to use torque pas because im getting a 60T chain ring and want it to assist at higher speed and have super human pedaling strength while still providing human assist. I switched it Torque Pas and I get this horrible stuttering. It seems to smooth out a little bit but it seems kinda odd that when i back off the pedals its actually smoother like its acting more like Pas vs Torque pas. It was cold last night on my initial tests so I didnt look into it deeper. Ive been reading the manual trying to figure out which setting will smooth things out. I believe my throttle is mapped out correctly no motor creep or dead zones in the throttle. I do have a 3 speed switch installed. 33 66 and 99 are programmed into it.

Looking to get this solved very soon and get more documentation for this BB.

icecube57   1 GW

1 GW
Posts: 3072
Joined: Apr 25 2008 8:42am
Location: Austell GA
Contact:

Re: Cycle Analyst V3 preview and first beta release

Post by icecube57 » Jan 16 2014 12:57pm

Ok so i got the torque sensing TDCM BB and i think i might have been better off with a PAS sensor. My problem is with a small chain ring there isn't enough force on the pedal due to get smooth engagement of the assist due to over spin of the pedals thats caused by agressive settings of the pas. A constant force isn't on each stroke. With a larger chain ring there will be a constant strain on the pedal that will smooth out the power hopefully. As of right now the Auto PAS on the CA with 500-750w limit is working pretty well. I still always have use of the throttle. The engagement of pas can be startling and dangerous at low speed. It seems a bit of a buck no matter how low you set the start watts to..

The Torque Pas also senses human force and provides a fixed amount of assist depending on the assist factor. You assist factor by default is 2x I preferred mine around 5x. At the higher numbers if you cant keep the pressure on the pedals when it kicks in you get a surging effect which is amplified by your reaction of being hesitant to pedal while this is going on.

When I change out my chain ring out to a 60T I'm probably going to find that AutoPas might not be adequate due to my cadence being alot lower and I will probably have drop outs and my ultimate goal was to get a substantial assist at higher speeds without having to be on the throttle. So im thinking i have to modify my riding style. Using the throttle to start and overcome the retarded pedal force to get the bike going with the large chain ring and maybe let pas take over at the higher rpms...

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 » Jan 16 2014 4:47pm

I don't have specific experience with this unit, but here's some thoughts:
  • The Torque functionality uses the Power PID controller as you can see by the 'W" in the Limit Flags display. You can soften the engagement and overall response of PAS in torque mode by backing down the PLim->WGain setting. You might go down to 10 for a start and work up from there. This should also allow you to address your surging issue.
  • Torque mode initially delivers force based on 55rpm and the actual pedal torque. The assist increases with both pedal torque and rpm (once rpm exceeds 55). You can scale the torque application by using Trq->SensrType=Custom which will enable configuration of Trq->TrqScale which is normally -200.0 Nm/V for a Thun. You can adjust this value to flatten/steepen the slope of the torque-to-power relationship. If set too flat, there will be little increase in assist with increased pedal force, etc.
  • Trq->AsstStart does not really adjust the PAS force at all - it sets the amount of power that is added or subtracted from the computed rider power. This can be used to delay the onset of PAS assist by forcing some threshold effort. It does not directly soften or alter the engagement or application of assist.
    • Gee - description of this variable is missing in the Guide - I will add it. Thanks!
    • There is a slightly more complete discussion of the torque computation in the AtuoTorqPAS document available here. This will be included in the next Guide release.
  • Although I believe you can get this squared away (you are breaking new ground here on this thread), you can also download the AutoTorqPAS document from here that will give you rpm-dependent assist with a simple PAS wheel instead of just fixed assist. This lets you get increased assist automagically at higher rpms by using the V3 torque mode instead of the simpler AutoPAS mode. I use this mode myself with a cheapie PAS wheel and it works very nicely for me -- I have a 52x11 sprocket setup and I can get get assist up into the 30's if I wish.

    In your case - since you actually have a torque input - you will need to make an adapter JST to disable the torque input. This is a PITA and I'm not really recommending this use-a-torque-sensor-as-a-PAS-wheel strategy except as a last resort.
In any case, please post back with your experiences - there is clearly a gap in the Guide for this new device and additional input could help remedy that.
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

User avatar
NeilP   10 GW

10 GW
Posts: 4962
Joined: Nov 27 2010 3:18pm
Location: 49.17303, -2.05354 Jersey, Europe
Contact:

Re: Cycle Analyst V3 preview and first beta release

Post by NeilP » Jan 16 2014 5:24pm

not trid my V3 yet..and months before i do..bck in plaster cast again..

but anyway.

No I have my shiney new CA datalogger and have run it with an old small screen V2 CA, i see it logs: Ah, volts, current, speed , distance

does the Data logger when coupled to the V3 log more data? such as human watts ? cadence ? or other data?
:) Only Irish coffee provides in a glass all four essential food groups: alcohol, caffeine, sugar, and fat :)

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 » Jan 16 2014 5:37pm

Ya -a few more goodies - here's a table from the 'Serial Port' section of the Guide.
CaV3_DataRecord.png
CaV3_DataRecord.png (18.53 KiB) Viewed 1805 times
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

User avatar
NeilP   10 GW

10 GW
Posts: 4962
Joined: Nov 27 2010 3:18pm
Location: 49.17303, -2.05354 Jersey, Europe
Contact:

Re: Cycle Analyst V3 preview and first beta release

Post by NeilP » Jan 16 2014 5:47pm

sorry, that is me being lazy..looked through the data logger guide..did not check the V3 guide..I have even printed it out..for night time reading..
I don't worry bout waking in th e middle of the night now and reading for a few hours....
http://endless-sphere.com/forums/viewto ... 26&t=56587
:) Only Irish coffee provides in a glass all four essential food groups: alcohol, caffeine, sugar, and fat :)

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 » Jan 16 2014 6:51pm

icecube57 wrote:ok Im making this post since there is absolutely no documentation for the tdcm torque sensing bottom bracket.
...
Looking to get this solved very soon and get more documentation for this BB.
Okay - I pinged Justin and here's his advice on setup:
Justin_le wrote:The #PAS poles need to be set at 12, the sensor type needs to be set to 'custom', and the value for the gain actually depends on the chainring diameter. So for this what we do is apply the back brakes and then apply a known mass (like 50lb) on the forwards pedal to produce a known torque, while watching the change in voltage scale on the setup menu screen. So if it shows 2.45V with no pedal weight, and then say 3.1V with a 50lb mass on a 175mm crank, then the scaling is:

50 * 4.44 N/lb * 0.175m / (3.1V - 2.45V) = 59 N-m / V.

Since it measures the backwards thrust on the crank then it is chainring size dependant since a larger chainring produces less force for the same pedal torque. Most setups have wound up between 30-80 Nm/V. If you just select 50 as an initial guess it will work fine, just that the human watts might be off by a bit.
He goes on to describe that with some experimentation he will be able to provide a table to assist in setting up Trq->TrqScale based on the chainring tooth count. I will get that into the Guide as soon as it is available.

Good questions - thanks for the posts! :D
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

User avatar
Scott   100 W

100 W
Posts: 269
Joined: Mar 10 2012 8:56pm
Location: WA

Re: Cycle Analyst V3 preview and first beta release

Post by Scott » Jan 16 2014 6:52pm

justin_le wrote:
Fishmasterdan wrote:Wow that is very nice..
Would it be possible to make the aux into a second thermo?? or possibly series connecting 2 probes and get a read out of the highest one?
In principle Aux could be made as a 2nd thermometer though I haven't coded for that. But your other suggestion actually works OK with NTC thermistors. If you get two 20K thermistors and wire them up in parallel, then it behaves like a single 10K thermistor that (due to the high nonlinearity) is mostly weighted to show the higher of the two temps.
-Justin
Does anyone know if this code ever got written? For now I am not worrying too much about my controller temp because of the cold weather and because of the new EB324 cooling design (FETs are heat sinked to the sides of the case instead of only to the top). I already have a 10K thermistor installed on a FET and would really like to see what kind of temps are going on inside of the controller and motor at the same time. If the motor could use normal temp input and the controller could use the AUX input that would be awesome. Then when I get my cycle analogger it would be cool to see controller temp vs. motor temp.
Trek Liquid 10 9c 48v 10ah, top speed 28.5 mph. 55.2 mph down a steep hill. - sold
Giant DH Team build http://endless-sphere.com/forums/viewto ... =6&t=51106 top speed 70 mph

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 » Jan 16 2014 8:15pm

Scott wrote:Does anyone know if this code ever got written?
No, it was not.
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

User avatar
NeilP   10 GW

10 GW
Posts: 4962
Joined: Nov 27 2010 3:18pm
Location: 49.17303, -2.05354 Jersey, Europe
Contact:

Re: Cycle Analyst V3 preview and first beta release

Post by NeilP » Jan 16 2014 11:21pm

The analogger has two extra inputs of its own, that can both measure temp.
Not enabled by default, so after first run you need to edit the settings.txt file that is created on the analogger.

You must enable the aux inputs
Set them to temp and deg F or C
Set temp coefficient of your thermistors
Solder leads to the analogger PCB aux pads to connect to your 2 new temp sensors
:) Only Irish coffee provides in a glass all four essential food groups: alcohol, caffeine, sugar, and fat :)

icecube57   1 GW

1 GW
Posts: 3072
Joined: Apr 25 2008 8:42am
Location: Austell GA
Contact:

Re: Cycle Analyst V3 preview and first beta release

Post by icecube57 » Jan 17 2014 1:10pm

With the hills in Georgia. Im often well below 55 RPM. I was playing around with the system and as long as there was sane values for the assist it was fine in Torque Pas and Auto Pas. As soon as you start trying to be Lance Armstrong the surging and the cut outs come into play. I haven't played with the gains yet. This will be Interesting to see how well i scales up. Its like on level ground pas works fine. On hills it works fine when I set it to use all the available power of the bike but I simply don't need that at speeds below 20. So i think this is where the limit switches come into play. My power band is so wide. This would be quite easy with a 1000w bike but trying to embrace the spirit of PAS on a 4500-5000w rig is a bit of a challenge. There also seems to be a jolt when it disengages and reengages at low speed which is quite startling. Its almost a run away bike scenario. I assume i need to change the Pas start and stop threshold. Does the limit flag for Start Speed also supersede PAS. Another person suggest I change my battery to phase amp ration to give it a softer more natural start. I still want to use my throttle and I dont need it to feel spongy off the line.

Post Reply