Page 42 of 182

Re: Cycle Analyst V3 preview and first beta release

Posted: Jan 30 2013 7:18pm
by MattyCiii
Here's a behavioral quirk - I don't think it's intentional:

This behavior is observed on an R/C setup.

I have the [Throttle Out --> Ramp Down] dialed way down. I have it set to 10 on a scale of 0 - 999. This achieved the following desired behavior - when I let go of the throttle, the motor speed drops off quite slowly. That's what I want, and that's what I get. I have my reasons why, if curious just ask.

Here's the strangeness. When I hit the e-brake cutoff, the motor speed drops off just as slowly as if I snap the throttle to zero. These should not necessarily be processed the same way. I would prefer that the motor drop off quickly - as in ""immediately". If I'm hitting the brake, I don't want the brakes to fight the motor.

Is this intentional? Can it be changed, such that e-brake cutoff either has its own ramp-down (I can't imagine why), or such that e-brake cutoff drops throttle output to zero?

Re: Cycle Analyst V3 preview and first beta release

Posted: Jan 31 2013 4:34am
by amberwolf
For those few running the CA off 12V and using a separate battery voltage input like I am (and I think Hjns is?), a note:

If the 12V isn't steady, like say a turn signal momentarily causes a voltage sag, the throttle output will also dip a teeny bit. Not much, but it's enough to notice "surging" (sagging, really) in sync with the turn signal. The more sag, the worse the problem is.

Thankfully it is a dip in trhottle voltage, not a rise, so it won't cause a power surge from the motor, but it is discomfiting. It's not a big deal on my CrazyBike2 setup with LED turn signals and a 12V 20Ah lighting pack as the sag is tiny, but on my DayGlo Avenger or Delta Tripper which use incandescent turn signals and a much smaller saggier 12V lighting pack (DT) or a DC-DC converter (DGA) that sags a LOT (almost blinks off the headlight during signalling!) I expect it would cause an extremely noticeable throttle surging, at best.


I suspect I can fix it by putting larger capacitors at the CA, for CrazyBIke2, but might require a diode on the input and capacitors after that for the other vehicles. (since I don't yet have the right stuff for really bright LED turn signals on both of them, or sufficiently powerful 12V sources to resist the sag caused by incandescent turn signals, at least in the space avaiable on them).

Re: Cycle Analyst V3 preview and first beta release

Posted: Jan 31 2013 5:19am
by hjns
amberwolf wrote:For those few running the CA off 12V and using a separate battery voltage input like I am (and I think Hjns is?),
Not anymore I am. Some time ago I ran 30S and inquired about the CA accepting that high voltage. Justin answered that the CA can accept that no problem. Therefore, my CAv3 is still hooked on the controller using the DP. Later I moved back to 20S, so no reason to change anything for me.

I did fry a CAv3 when connecting it to the Thun, and then connecting it to the 20S lipo pack. As mentioned several times now in this thread, the CAv3 can not convert that high voltage to power the Thun, therefore, the Thun needs it's own 12V when using more than 10S as battery. I recently sent the fried CAv3 back to Adam at Grin under RMA. Fortunately, my other CAv3 is working quite well. Keeps my cromotor from overheating when climbing Swiss hills.

Re: Cycle Analyst V3 preview and first beta release

Posted: Feb 02 2013 2:21am
by pendragon8000
Hi Justin and all,

Can I get a CA v3 and run 24s straight in?

Did a quick search but wasnt conclusive. I have a CA v3 on my first build and love it. Totally worth it, considering it can improve efficiency so easily, it pays for itself to some degree.

So 100 volt is OK?

Thanks

Re: Cycle Analyst V3 preview and first beta release

Posted: Feb 02 2013 2:26am
by amberwolf
http://www.ebikes.ca/store/store_CAV3.php
"Explanation Standard CA-DP Pads"

Re: Cycle Analyst V3 preview and first beta release

Posted: Feb 02 2013 2:56am
by pendragon8000
Thanks Amberwolf.
" Battery Power (V+): This is the V+ supply of the battery pack, used both to power the CA and also to sense the battery voltage. The maximum supply voltage with no accessories is 150V, but this must be derated if there are other devices (Torque Sensor, Input Throttle etc) also drawing power from the CA."

Re: Cycle Analyst V3 preview and first beta release

Posted: Feb 03 2013 10:10pm
by teklektik
The 'Advanced Features' section of the Setup Notes has been expanded and posted as a printable PDF file. Additional notes on operation with high voltage batteries has been added. Since this is largely a reference section, most hyperlinks have been replaced with material imported from other posts as well as the Grin Tech site to make a more complete printable document. The web version is not yet updated.

The existing Basic Installation and Console Operation section PDFs have been renamed but have no material change in content (the filename timestamps are unchanged from the previous posting).

Since this has pretty much gotten out of the 'notes' category, it has been renamed as an Unofficial Installation Guide.
Enjoy! :D

Re: Cycle Analyst V3 preview and first beta release

Posted: Feb 04 2013 8:22pm
by teklektik
teklektik wrote:The 'Advanced Features' section of the Setup Notes has been expanded and posted as a printable PDF file.
Oops - important omission regarding power sequencing in new section on High Voltage Vehicles. Reposted...

Re: Cycle Analyst V3 preview and first beta release

Posted: Feb 06 2013 4:43am
by hjns
I love my CAv3.

Yesterday at work, I left my bike outside in the cold. It started raining, then snowing. My lipos were charged inside, so no problems there. However, after I had mounted my lipo on the bike again, and wanted to drive home, a bad thing happend. I turned the ignition key, and immediately the bike jumped 50 cm forward with a 3kW push. In reflex, I gripped my ebrake, cutting the power, switched off the bike, and got off. You can bet I was a bit shaken.

I started doing some analytics. When I got to my bike, I had to swipe a lot of snow from the saddle, but also from the twist throttle. I immediately assumed that the throttle was soaking wet, leading to an increased conductivity.

I gripped my ebrake, turned on the bike, and went into the Throttle-in configuration screen. Look and behold, the throttle showed a constantly changing ThroIn between 0.90V and 1.80V. With the ThroIn at 0.87V mapped for ThroOut of 1.25V, and the controller reacting to 1.30V, it was no wonder that the bike jumped at no throttle. Because the No-throttle could actually translate into a 1.80V ThroIn, which would be translated into something like 2.5V ThroOut. And yes, that would make the bike produce something like 3kW at standstill.

The solution was clear. I mapped ThroIn to 1.90V, delivering ThroOut 1.25V. This obviously created a huge starting lag of my throttle, but at least I was able to get home safely.

Overnight, the bike was drying in my wintergarten. This morning, I checked, and everything was normal again with ThroIn resting voltage at 0.85V. I remapped to ThroIn 0.87V and all was well. I am SOOO happy with the CAv3 that allowed me to diagnose the problem within seconds, and create a working solution within minutes.

Any suggestions how to protect the throttle from the rain? Should I open up the throttle and fill it with grease or vaseline? What happens with vaseline at -10oC? Any other ideas?

Re: Cycle Analyst V3 preview and first beta release

Posted: Feb 06 2013 5:26am
by amberwolf
I suspect it would be a good solution, but I have no recent experience at those temperatures. :shock:

I also have praise for the throttle readout/range control: I have had a wierd intermittent problem with teh 18FET on CrazyBike2, which seems to only happen when the pack is closer to full charge--the lower the voltage gets, the less often it happens.

WOThrottle voltage can be up to 4.2something volts, out of my thumb throttle. Never a problem on the other controllers so far, but on this one, at full pack voltage, it causes complete cutout, just as if I engaged an ebrake (which I don't have wired up on this one). This would not usually be a big deal, but I naturally hit WOT trying to accelerate away from a stop sign or traffic light, or when certain traffic conditions suddenly happen that I have to acclerate thru or away from. Having power cut out right then is pretty unhelpful, and it's not instinctive to roll throttle back to get more power. :(


Anyway, I hadn't done anything about it because it was easy to deal with, but tonight one of those traffic situations came up with an idiot deciding to try to pull out of a side street jsut as I was about to pass in front of him, and my only option was to go faster really quick to get out of the wya---braking would've just caused me to skid and crash, at that speed with those conditons (sandy road edge), and traffic in the lane next to me so I couldn't just swerve left out of the way.


So after easily getting out of the way (but with an instant's panic when power cut out because of the problem), I continued to a parking lot and pulled into a space, and went to the first "left" screen on the CA, with the awvst limits indicator, and the throttle voltages. I tilted the bike on the right pod up so the wheel was off-ground, watching the Throttle in and out voltages, and goosed it WOT. As soon as it reached 4.10V out, the motor cut out. 4.09V it was fine.

I entered setup, throttle output section, and set max out to 4.09V. Retested, and no problems since. Way easy fix, and took almost as little time as just rolling back the throttle to recover! (which I won't have to do now. :))


So anyway, another "YAY!" from me. :)

Re: Cycle Analyst V3 preview and first beta release

Posted: Feb 06 2013 12:38pm
by teklektik
hjns wrote:Look and behold, the throttle showed a constantly changing ThroIn between 0.90V and 1.80V. With the ThroIn at 0.87V mapped for ThroOut of 1.25V, and the controller reacting to 1.30V, it was no wonder that the bike jumped at no throttle. Because the No-throttle could actually translate into a 1.80V ThroIn, which would be translated into something like 2.5V ThroOut.

Any suggestions how to protect the throttle from the rain? Should I open up the throttle and fill it with grease or vaseline? What happens with vaseline at -10oC? Any other ideas?
Judging by the Zero throttle voltage this looks like a hall throttle - I don't really understand the cause of the failure unless it's stray voltage across wet hall leads or throttle connector pins. I guess you could try drying it out with a heat gun or hair dryer. It that works, you might goop the exposed internal connections with Liquid Lectric Tape, Vasoline, or Permatex 22058 dielectric grease (available from auto stores and basically a kind of fancy silicone grease with no damaging petroleum products). The throttle connector is a good place for a dab of dielectric grease in any case.

A quick search of ES shows this has come up before, but no real follow-up so the reasons and eventual fix were not clear.

Re: Cycle Analyst V3 preview and first beta release

Posted: Feb 07 2013 3:06pm
by bowlofsalad
Hello,

I've tried searching, but I am unable to locate something that sort of spells out the change log in versions of cycleanalyst. Does anyone know of some kind of list?

Re: Cycle Analyst V3 preview and first beta release

Posted: Feb 07 2013 3:23pm
by teklektik
There is no change log. If you read the Release Notes you will get some idea of new features, but they do not address all changes or bug fixes. You can post a question about a specific concern.

Re: Cycle Analyst V3 preview and first beta release

Posted: Feb 08 2013 12:08am
by Green Machine
Hi guys,

It is my understanding that the v3 cycle analyst allows you to use a resistor type throttle even with a hall sensor hub motor. Is this correct to assume? Is there any draw backs to using a resistor throttle in this way?

Re: Cycle Analyst V3 preview and first beta release

Posted: Feb 08 2013 6:21am
by amberwolf
Well, it doesn't matter if the hub motor has hall sensors or not--nothing to do with throttle type.

However, if you have a controller (for any kind of motor, hub or not, sensored or not) that has an input voltage range typical of hall sensor throttles, that doesn't go from 0 to 5V (typical of potentiometer throttles), but more like 0.8V(ish)-4V(ish), you can still use the pot (resistive) throttle on there via the V3 CA. You'd set the CA's throttle input range to whatever comes out of the throttle, and the output range to whatever the controller requires. The CA will scale between them for you.

Re: Cycle Analyst V3 preview and first beta release

Posted: Feb 08 2013 7:42am
by teklektik
Green Machine wrote:It is my understanding that the v3 cycle analyst allows you to use a resistor type throttle even with a hall sensor hub motor. Is this correct to assume? Is there any draw backs to using a resistor throttle in this way?
The difficulty has not been with the motor but with the controller - most of which are built with a roughly 1-4v input range similar to the output of hall throttles. The 0-5v output of resistive throttles results in 'dead zones' at the ZERO and WOT ends of the range and has classically been addressed by adding some external resistors or trimpots. One of the very nice features of the V3 is that it is specifically designed to provide configuration options to match throttle output to controller input range. (See the colored graphs in the middle of this post.)

Resistive throttles (Magura) provide a linear output which has generally made them more desirable. Halls have typically been pretty non-linear with some annoying mid-range dead zones leading to twitchy throttle control. There do seem to be some newer hall throttles with improved magnet designs that improve the linearity.

In your case (if we're talking about your MAC mid-drive), running the V3 in Current or Power Throttle mode in addition to the resistive throttle will make the motor much more controllable and will eliminate the sudden midrange step in throttle response.

Re: Cycle Analyst V3 preview and first beta release

Posted: Feb 11 2013 1:35am
by Kepler
I am up to the wiring side of setting up a my CA3 on an RC friction drive with magnetic PAS. I am thinking of running the +5V and ground from the ESC and then wiring the PAS signal to the surplus SP wire within the 6 wire harness.

Can anyone see any issues with this wiring diagam?

Re: Cycle Analyst V3 preview and first beta release

Posted: Feb 11 2013 2:30am
by teklektik
Kepler wrote:I am thinking of running the +5V and ground from the ESC and then wiring the PAS signal to the surplus SP wire within the 6 wire harness.

Can anyone see any issues with this wiring diagram?
Re-purposing the yellow CA-DP(5) lead for the PAS connection is fine. I'm unfamiliar with the RC ESC signal nomenclature, but assuming that ThrO->PPM is correct, the throttle looks good, too.

However, as per section 3.3 of the CA V2.23 manual regarding external shunts, and the wiring of the CA SA shunt, you should have the CA-DP(2) Gnd on the battery side of the shunt with the blue (S-) shunt wire. This will raise the ESC Gnd slightly relative to the CA under high current (0.05v for 0.5mOhm shunt at 100A) but this will not adversely affect interpretation of the PAS ON/OFF state. Alternatively, connecting the PAS Gnd to the same point as the CA-DP(2) Gnd might be slightly tidier electrically.

Re: Cycle Analyst V3 preview and first beta release

Posted: Feb 11 2013 3:00am
by Kepler
Thanks for the feedback Teklektik.

In relation to what side to put the ground, I am a little confused. Manual seems to say the opposite to the CA RC thread http://endless-sphere.com/forums/viewto ... 6&start=30

Re: Cycle Analyst V3 preview and first beta release

Posted: Feb 11 2013 9:03am
by teklektik
Kepler wrote:In relation to what side to put the ground, I am a little confused. Manual seems to say the opposite to the CA RC thread http://endless-sphere.com/forums/viewto ... 6&start=30
The confusion seems to arise from the bulleted items on that page and the term 'Gnd' which might be either ESC Gnd or CA Gnd. It is a little unclear if the page is noting that this is an alternative (ESC Gnd) or is the same as Shunt- ('normal' CA Gnd). The latter would make it consistent with the CA v2.23 manual and construction of the CA-SA shunt which is also recommended for CA-RC use.

Although I could make some guesses, I'm unfamiliar with the exact way the shunt signals are handled on the CA PCB and would recommend going with the CA v2.23 interpretation (CA-Gnd). You might like to shoot an email to Grin Tech Support for an authoritative answer if it is more convenient in your wiring to tie the leads as you have have drawn them. If you do, please post back.

Good page, though -thanks for the link.

EDIT - Here's what the older V2 CA-SA shunt looked like as mentioned in the CA-RC thread. The four CA lines are identical in function and color code to the V2 or V3 CA-DP connector.
CA-SA Shunt Dongle2.gif
CA-SA Shunt Dongle2.gif (6.3 KiB) Viewed 3302 times

Re: Cycle Analyst V3 preview and first beta release

Posted: Feb 12 2013 1:23pm
by izeman
i have some problems connecting my lyen 312 sensored controller to the ca. i want to use the throttle control feature of the ca as well.
when i connect the throttle directly to the controller i have a butter smooth start, but topspeed is 16km/h only.
when i connect the throttle to the ca directly nothing happens. i disconnect the green wire from the diode pad on the board, and connect it directly to SL1 the motor spins at full speed, but stutters a bit when starting, and stalls sometimes when releasing the throttle and going wot again.
i must say that i didnot move the yellow cable to SA, and left it where it's connected. i don't want to desolder and resolder any cables if not necessary.

Image

it's also weird, that when repositioning the green cable, then the directly connected throttle doesn't work anymore.

any help? i'm a bit confused, and a afraid to experiment a lot as i don't want to short anything, and destroy the controller and/or CA. thanks!

Re: Cycle Analyst V3 preview and first beta release

Posted: Feb 12 2013 3:18pm
by teklektik
Sorry to hear you are having issues. A few questions:

What is the source of this wiring diagram? Are you attempting to add a CA connector to your controller which lacked it originally or to modify an existing V2 CA controller connector to be V3 compatible?

Apologies for some unfamiliarity with your controller connection labelling, but what is the relationship of SL1 to SP? Are these the same electrically? For a generic controller upgrade, the CA-DP(6) Throttle Out signal would go to SP (controller Throttle In). SL1 may be an alternate throttle input Lyen has provided, but its purpose needs a bit of clarification.

You did not mention your V3 configuration settings. Have you gone through the throttle set up procedures in the Unofficial Installation Guide? If not, then the throttle and controller may not be matched and you could be experiencing input faults from the controller or CA causing the 'stalls'.

I run a couple of Lyen EB312s with a V3 on my bike and I am certain you will get good results if this snag can be identified. :)

Re: Cycle Analyst V3 preview and first beta release

Posted: Feb 12 2013 3:40pm
by mrbill
Has a "clutch saver" or more accurately, "one-way bearing saver", feature been considered?

Some of us ride with drivetrains that include one-way bearings that are usually engaged after the drivetrain takes up what is usually but not always a small amount of backlash (e.g. a slightly slack chain). The moment the drivetrain becomes engaged is one of high instantaneous stress on any clutch or one-way bearing. Most bicycle one-way bearings (e.g. ratcheting freewheels) are usually pretty sturdy under this kind of load, but roller-bearing clutches have torque limits that on e-bike-sized parts are marginally adequate.

I have tried to lower the throttle gain to effect this, but I find this does not quite have the behavior I'm looking for. Reducing throttle gain merely caps the speed with which the throttle lever can be thrown but still allows the drivetrain to be "slammed" if enough backlash is in the system and the current limit is moderately high.

This feature might best be associated with the Throttle Out programming.

I was thinking that an easy way to enable a mode like this is to have a current gain ramp from zero. This would be different from the throttle pull-back gain that affects the rate of pull-back when current exceeds the maximum current. When the throttle is engaged the current limit is initially zero, then ramps to the set limit over a period of seconds or milliseconds to be set by the user.

A more complicated program would involve some feedback. The current limit would be fixed to a low number sufficient to spin an unloaded drivetrain until a significant load is sensed. Then after another short delay the current limit would ramp to the full programmed current limit.

The user should probably be able to turn on and off this feature as it probably is not useful to DD hub motor users. The "low" current limit, "short delay", and ramp could also be set by the user, since some drivetrains could conceivably involve multiple stages whose backlash would be taken up in stages rather than in one step (like a long freight train getting started).

Anyway, I thought I'd throw the idea out there in case it hadn't been considered.

Re: Cycle Analyst V3 preview and first beta release

Posted: Feb 12 2013 4:12pm
by teklektik
mrbill wrote:Has a "clutch saver" or more accurately, "one-way bearing saver", feature been considered?
Yes. Re-engagement at speed was first discussed back on page 3 of the thread in the context of gear motors.

Re: Cycle Analyst V3 preview and first beta release

Posted: Feb 12 2013 5:43pm
by izeman
teklektik wrote:What is the source of this wiring diagram? Are you attempting to add a CA connector to your controller which lacked it originally or to modify an existing V2 CA controller connector to be V3 compatible?
the source of the diagram? if i understand your question correctly: this is directly from justin's post.
lyen's controller has the CA connector installed. but i don't know if it's v2 or v3
Apologies for some unfamiliarity with your controller connection labelling, but what is the relationship of SL1 to SP? Are these the same electrically? For a generic controller upgrade, the CA-DP(6) Throttle Out signal would go to SP (controller Throttle In). SL1 may be an alternate throttle input Lyen has provided, but its purpose needs a bit of clarification.
SP is fine for me. on the controller board there CA-DP(6) is connected to a diode, which itself is connected to SP. i desoldered pin6 and soldered it to SP directly. that made the ca thru throttle work. otherwise the wheel wouldn't turn.
the question remains if CA-DP(5) has to go to SA, or if it may stay connected to where it is - which is the pad where it should work with a v2.
You did not mention your V3 configuration settings. Have you gone through the throttle set up procedures in the Unofficial Installation Guide? If not, then the throttle and controller may not be matched and you could be experiencing input faults from the controller or CA causing the 'stalls'.

I run a couple of Lyen EB312s with a V3 on my bike and I am certain you will get good results if this snag can be identified. :)
i read the whole setup guide and adjusted all min/max values, and it works really nice w/o any dead zones. to ask a precise question: what did you have to change to connect the v3 with connected throttle to a eb312?
thanks :)