Cycle Analyst V3 preview and first beta release

[oatnet]

I reset it to 0, and when i leave set up I still get faulty watt reading again, go back into set up and somehow the shunt number has been reset to 20.0. Reset it to 0 or 1.0 and error repeats.

I dont know what to do now..should i reflash the cycle analyst?

Will I lose all my settings for the throttle in throttle out?

Sounds similar to the problem I had, but I had the problem upon installation. In my case, I had an extra digit I could not see in b20, as a result from moving between high-mode and low-mode in b19. I had to reflash to b19 to see/delete the extra leading digit (I forget if it was hi-mode or low-mode where I saw it) and then flash back to b20 to have presets again. I lost all of my settings in the process and had to reprogram them, so write yours down first.

-JD

 

[amberwolf]

Has anyone else managed to actually blow theirs up yet? I think I shorted something in the Thun sensor cable to ground (via the bike frame) that smoked something in the CA. All I know at the moment is that the main power transistor in the CA has zero output, but not why, yet.
http://www.endless-sphere.com/forums/viewtopic.php?f=2&t=19540&p=676208#p676208
Am going to go poke at it with a meter and see if I can find anything else out.

 

[amberwolf]

I think I found the cause of the failure; might nto be a short after all but rather my failure to remember this:

The 10V bus can power the THUN sensor directly provided you are using a pack that is 48V or less. For higher voltages the CA's regulator would get hot and you'd want to feed the THUN 12V power from a DC-DC.

My pack is 66.2V hot off charger, and was at the time still almost fully charged, so that's a lot more than 48V. I shoudl've remembered that, before hooking it up, but I didn't, so that is probalby what killed the FET, and maybe whatever drives it.

http://www.endless-sphere.com/forums/viewtopic.php?p=676523#p676523

 

[hjns]

Hi AW,

Yes, I did the same a couple of months ago. Connected the Thun and the CAv3 B(early) to my 20S 84V pack. The CAv3 started smoking and stopped working.

Justin told me it is repairable. In the meantime, I had ordered another CAv3, which is currently working flawlessly after flashing to B21.

I have not had time to send the broken CAv3 back to Justin for repairs. I have an RMA number though.

 

[amberwolf]

Yes, I did the same a couple of months ago. Connected the Thun and the CAv3 B(early) to my 20S 84V pack. The CAv3 started smoking and stopped working.

I guess that makes yours more spectacular than mine. :lol: I missed your post about it somehow, though, both when following the thread and perusing it again since blowing mine up to see if I could find better info than I remembered.

 

[lollandster]

Hi, I have a feature request.
I have the CAv3 with the thun sensor and I like the idea of using presets. The problem I have is that the PAS level is not a part of the preset values and I would like to disable the PAS when in offroad mode, if PAS level was part of the presets I could set it to 0 to disable it. I would also like to disable the throttle when in assist mode to allow for legal PAS only mode. And lastly I would like to be able to use a LMH switch to select preset as mentioned in an earlier post.

I'm guessing PAS level doesn't change with presets because it can be changed with the AUX channel, but if I don't use the AUX or if in the future the AUX can be used for presets it would be very handy to use presets to change the PAS value.

There also seams to be a typeo on SLim->PSGain where it asks for V/uph which I guess is suppose to be V/kph

Thank you Justin for introducing high quality torque sensors to custom builds.

 

[izeman]

...And lastly I would like to be able to use a LMH switch to select preset as mentioned in an earlier post.

+1

 

[NeilP]

Well I was not going to put in a feature request .as I reckon it is a bit too late for this version.(unless it is a small firmware mod) but since it has just been mentioned....

I did have an idea while picking myself up off the floor after my rear wheel span out from under me on a corner. My thought was traction control/ wheel spin prevention.


So no good for this version or thread but I'll chuck the idea up any way.

I am thinking of wheel spin/slip control. So if the wheel is turning faster than ground speed, power is reduced.
On a single wheel drive bike, I guess a speed sensor on the front wheel comparing speed of the rear wheel,
Modern tractors use doppler radar speed sensors, but that is going to be a bit OTT for a bike!

I suppose that if this was implemented at some point in the future, with driven and non driven wheel sensors and a suitable controller with controllable (or modulated max) regen was used this could give some sort of regen braking ABS system. I know that my regen braking often causes the back wheel to lock on slippery surfaces.

edit...maybe it would be possible within this version? using the aux/PAS input as one of the speed sensor inputs?

We have wheel slip control modern farm tractors. You can dial in the percentage of wheelslip/spin allowed, to gain maximum pulling traction. On a tractor on soil, 8-15 % wheelslip gains maximum traction, On a tractor the forward speed is often measured via a downward/forward pointing doppler radar unit, and compared with axle speed and wheel size. This then adjusts implement depth to obtain maximum traction

 

[izeman]

how should we do our wheelies then? no rear wheel riding anymore?

 

[NeilP]

how should we do our wheelies then? no rear wheel riding anymore?

Switchable like modern ABS/ launch control

 

[izeman]

now that we started dreaming: what about a gyroscope installed inside the ca so we can make bullet proof wheelies?
i think it was audi who installed something like that in their first ebike.

 

[teklektik]

Yes, I did the same a couple of months ago. Connected the Thun and the CAv3 B(early) to my 20S 84V pack. The CAv3 started smoking and stopped working.

... I missed your post about it somehow, though, both when following the thread and perusing it again since blowing mine up to see if I could find better info than I remembered.

A special note has been added to the Setup Notes calling out this voltage limitation specifically. Hopefully this will give it slightly greater visibilty. Thanks guys for publishing the details of the difficulty.

 

[hjns]

Yes, I did the same a couple of months ago. Connected the Thun and the CAv3 B(early) to my 20S 84V pack. The CAv3 started smoking and stopped working.

I guess that makes yours more spectacular than mine. :lol: I missed your post about it somehow, though, both when following the thread and perusing it again since blowing mine up to see if I could find better info than I remembered.

I am not sure if I posted about it. If I did, I forgot where I put it.

 

[teklektik]

...I am not sure if I posted about it. If I did, I forgot where I put it.

This is a long thread and in the end it was your response to AW's post that flagged this as other than an isolated occurrence and as an installation consideration perhaps worthy of special mention.

 

[hillzofvalp]

now that we started dreaming: what about a gyroscope installed inside the ca so we can make bullet proof wheelies?
i think it was audi who installed something like that in their first ebike.

this is something I've been meaning to work on

 

[MattyCiii]

In the past the high range mode wouldn't let you input a shunt greater than 0.99mOhm, but in this firmware I let it so that you could still go up to 9.99 mOhm in high range, but if you then switch to low range it will treat that as 99.9 mOhm and you won't see the first digit. I'll put some proper clamps on the input shunt ranges again so that you can't do this. Main thing to keep in mind is that the Range selection effects much more than just the display (W or kW, and .01A or 0.1A), it determines the entire internal scaling factor used for all the power related variables (ie signed 16 bits = +-32,000, could be seen as +- 320 amps or +- 3200 amps).

-Justin

This functionality is still a bit... off.
Using two CAv3's, B20 on one, then B21 on both. Start with calibration set to Low --> change that to High. Now I want to change the shunt to 0001mOhm. It simply won't take. The first time I made the change and locked it in, the value (on both devices) changed to 18XX mOhm (eighteen hundred something something). I changed it again right away to 0001, and when I locked it in, it changed to 0008. Then tried 0001, 0007, 0009, 000x, and each time once locked in it changes to 0008.
With the value at 0008 mOhm in the rshut setup screen, I backed out to the Calibration setup screen and see the value listed there is 2 mOhm.

Unable to do any road tests tonight, I just simply put things back into Lo mode, and changed the shunt back to 1.000 mOhm. Which, by the way, it now won't do. I'm stuck for now with rshunt set to 1.008 mOhm.

Hope this helps doe B22

 

[hjns]

Just wanted to support MattyCii here.

I also was not able to use the High range mode well. I tried to enter my shunt value of 1.375 mohm, but failed in high range mode. I then moved to low range mode, but here I am limited to a max current of 99.9A. Also, despite my "high resistance" shunt value, I still would prefer to use all High range CA settings, because I strongly believe that anything over 3kW on an ebike can be considered High power.

Therefore, my feature request would be that the CA does not limit one to set the rdshunt at whatever value. Even in High range mode, one should be able to enter rdshunt values of more than .9999 ohm. Is that possible?

I am fully aligned with the concept that moving from low to high mode or the other way around scales everything a factor 10.

Alternatively, if in Low range mode, I set my rdshunt at 13.750 mohm, and I then switch to High range mode, will my rdshunt be converted to the correct 01.375 mohm?

 

[teklektik]

In the past the high range mode wouldn't let you input a shunt greater than 0.99mOhm, but in this firmware I let it so that you could still go up to 9.99 mOhm in high range, but if you then switch to low range it will treat that as 99.9 mOhm and you won't see the first digit. I'll put some proper clamps on the input shunt ranges again so that you can't do this. Main thing to keep in mind is that the Range selection effects much more than just the display (W or kW, and .01A or 0.1A), it determines the entire internal scaling factor used for all the power related variables (ie signed 16 bits = +-32,000, could be seen as +- 320 amps or +- 3200 amps).

This functionality is still a bit... off.

Yep. I finally had an opportunity to flash v3B21, having been running with v3B19 because of the jerky starts unique to v3B20. I tried a few experiments setting RShunt in High Range mode and it does look a bit broken. I may be wrong, but it seems tied to the display and data entry functionality - the actual run time calculations seem to work okay if you can get the proper RShunt value set.

Some issues:

1. In High Range mode there is a discrepancy between the displayed RShunt parameter value and the displayed value on the Calibration section summary screen.
2. As mentioned in an earlier post, the RShunt data entry format in High Range (xxxx) seems a step back from the v3B19 format (x.xxxx). The exact position of the decimal point is unclear. By experiment, it turns out that it's to the left of the entry field (.xxxx), that is, 0.9999 = 9999.
3. The High Range data entry field (.xxxx) cannot accept input of RShunt values 1.000 or greater since there is no available units digit.
4. The last digit of the High Range entry field does not (in general) allow entry of an arbitrary digit [0-9] but instead jumps to some new digit as soon as the field is closed and stored (e.g. entering 4030 saves and displays as 4033).

There are some other observed quirks, but in the present context of attempting to enter RShunt = 1.xxxx in High Range, (3) above presents a bit of a showstopper...

As Justin cautions, this Low/High Range setting is really designed to be a one-time affair, and as MCiii reported, some odd things appear to permanently corrupt the (displayed) RShunt value when swapping from High Range back to Low Range. Since this is an official no-no, it's not really a shortfall, but...

As a work-around, I flashed v3B19 and configured the Range to High and entered my desired RShunt value of 1.403 mOhm. I then flashed v3B21 and immediately adjusted it for High Range so the preserved RShunt value would be properly interpreted. (The Range setting is not a Calibration parameter and is reset to Low when flashed). The RShunt value was displayed correctly on the Calibration section screen (1.40) although the RShunt Setup screen showed 4033 instead (missing the leading "1." and with the puzzling trailing "3"). I was careful not to try to alter this value with v3B21. Going back to the start of Setup, I configured Setup normally, noting in particular that the default Max Amps and Max Power parameters had scaled up by x10 .

It was dark and rainy outside so no test run, but a run on the stand showed the main screen Watts display in kW with what appeared to be proper wattage readings. The set-in-B19-run-in-B21 gimmick seems to work and might be worth a try if you need this sort of RShunt setting plus presets - otherwise a downdate to v3B19 should do the job if the presets aren't really needed yet...

 

[teklektik]

Alternatively, if in Low range mode, I set my rdshunt at 13.750 mohm, and I then switch to High range mode, will my rdshunt be converted to the correct 01.375 mohm?

Well, it's not actually converted - the same stored fixed point value is just interpreted differently because the implicit decimal point is shifted according to Range. But - 'yes' - that would work exactly as you describe except that the Low Range entry field does not allow you to enter a value of 10 or greater... (good idea, though )

 

[MattyCiii]

*snip!*
The set-in-B19-run-in-B21 gimmick seems to work and might be worth a try ...

Thanks so much for the detailed analysis & write up Teklektik, and for all the amazing information you post. I'll be doing this trick!

 

[Philistine]

Sorry if this has already been asked on this thread but I searched and couldn't find it.

Can anyone tell me what the throttle input and output values should be in setup for a basic hall thumb throttle?

I have a Lyen 12 Fet, a basic thumb throttle from Lyen, on a HS3545, running on 12S lipo, @ 50amps. I have set the poles to "23", and the Rshunt value to 2.6. I have set it to speed based throttle. Limits are standard (ie 99, and 999amps etc.).

I have set everything up, and the bike goes but, it just stops giving throttle at about 100W (with full throttle applied). I am guessing it is the "voltage in and out" setting, but what should these be for a basic thumb throttle? Or is it something else? As I said, it runs smoothly, it is just as if I have the controller programmed at about 5 amps or something (it is programmed at 50amps).

Cheers, Phil.

EDIT: Just to be clear, at this stage I am just wanting to use my CAV3 as a basic CA (ie I don't yet have time to setup the current based throttle etc.., at this stage I just want to use it like a CA2 with speed based throttle). I presume it is the voltage input output settings for the throttle, but I don't understand why it is acting like I have limiting functions on. Should I flash it to the latest firmware (it is on 19), if so, do you flash it just like a programmable controller with the same cord (sensored or sensorless?)?

 

[amberwolf]

Can anyone tell me what the throttle input and output values should be in setup for a basic hall thumb throttle?

AFAICT from the few minutes I had with it before I blew it up, the "input" values are simply a readout of the actual throttle position, and the output values are what your controller responds to. On mine, I simply set it to the same as what the throttle actually puts out.

If you want a different throttle control range, like if your controller responds to a different range. So for instance, if a controller responds to 1.4V as zero throttle, and 3.8V as full throttle, but your actual hall throttle outputs 0.8V to 4.2V, you have a bunch of dead rotation at the top and bottom of throttle movement. You can eliminate that by changing the CA's output to match the controller's input spec, if you know what it is.

If it's not labelled, youc an determine it experimentally, by watching the input voltage and noting when the controller responds, and watching the speed (with a separate speedo?) and noting when the wheel doesn't get any faster but the throttle still moves.

 

[Philistine]

Thanks AW, I tried mucking around with the voltage output settings but it doesn't seem to help. I can't workout what it is, because everything works perfectly, it is just behaving like I have my controller programmed to about 5amps battery, but it is programmed to 50amps battery. When the CA is removed everything works perfect, so it is something in the CA that is limiting the output/performance to about 100-200W, but the amp limit and the speed limit are at max.

As I said, I basically just want to use this CA V3 like a V2 (ie I am not intending to use the functionality for a while due to time restrictions, and probably not on this bike, I only got this version so I could do both. I had hoped I could just set it to speed based throttle, and as long as I set the shunt values and poles right it would work out of the box. Is there an obvious limit setting I am missing? The Shunt value is right (I know from it being right on other CAs for this controller, and the watt readings on this CA seem right (ie, the 100W feels like 100W).

I guess I need to bite the bullet and read this whole thread....

 

[teklektik]

As I said, I basically just want to use this CA V3 like a V2 (ie I am not intending to use the functionality for a while due to time restrictions, and probably not on this bike, I only got this version so I could do both. I had hoped I could just set it to speed based throttle, and as long as I set the shunt values and poles right it would work out of the box. Is there an obvious limit setting I am missing?

... I have set it to speed based throttle. ...

Phil-
You have asked many good questions here, but the statement above is a little confusing and perhaps the source of the difficulty. I believe that you want simple monitoring without CA control or limiting. However, you have configured for speed throttle. I think this may be because of the common misnomer of basic controller operation as providing speed throttle (it's really PWM throttle) - the CA Speed Throttle is completely different - a true closed-loop throttle mode where the CA measures the speed and controls the throttle accordingly (not what you want just now).

I recommend that you follow the steps in this post carefully without modification. To get the old style CA V2 operation, configure for legacy mode, leaving the throttle connected to the controller. A critical step will be 11.2.1, setting ThrI->CntrlMode=Disabled. This will disable the CA control functionality and get you into legacy mode.

 

[teklektik]

... I'll be doing this trick!

Have at it - the rain let up today so I took it for a spin and it worked fine except that I had to scale up the associated gain setting by x10 as Justin called out on the Grin Tech V3 web page:

As well, all of the current and power feedback calculations are also affected by the chosen range mode, so an AGain of 50 in low range mode would be equivalent to 500 in the high range

The next post is the summary of this little experiment . Also, the setup notes have been updated to cover High Range RShunt configuration and linked to the next post as a temporary workaround.