Cycle Analyst V3 preview and first beta release

[teklektik]

It sounds like the test you performed successfully is exactly what you want... yes?

Sorry that was a typo. I meant, by hooking up an ebrake lever to the CA V3 it did not cut power to the motor. ... Going to look over all my connections and test it again.

No worries - puzzling that it didn't work, though. I can't get at my CA just now to test, but I might try just jumping the EBK to Gnd to be sure the ebrake isn't an issue.

A description of the EBK connection can be found in this post:

Ebrake Cutoff: Brake inhibit input, has onboard pullup as well so you can use either a mechanical brake switch or a hall effect device. This way you don't need brake signal wires going all the way back to the controller, they can go into the CA and then the CA cuts the throttle signal.

EDIT - Hmm - if this still doesn't work, you might try setting ThrI->MaxInput=0.2v instead of ThrI->MaxInput=0.02v just in case the ebrake functionality is handled in hardware directly on Ti instead of by software - it will give a little larger margin for EBK to pull down Throttle In. The revised setting of 0.2v is still well below the 0.37v default voltage of Ti so there should be no ill effects by making the Setup change.

 

[Degull]

It sounds like the test you performed successfully is exactly what you want... yes?

Sorry that was a typo. I meant, by hooking up an ebrake lever to the CA V3 it did not cut power to the motor. ... Going to look over all my connections and test it again.

No worries - puzzling that it didn't work, though. I can't get at my CA just now to test, but I might try just jumping the EBK to Gnd to be sure the ebrake isn't an issue.

A description of the EBK connection can be found in this post:

Ebrake Cutoff: Brake inhibit input, has onboard pullup as well so you can use either a mechanical brake switch or a hall effect device. This way you don't need brake signal wires going all the way back to the controller, they can go into the CA and then the CA cuts the throttle signal.

EDIT - Hmm - if this still doesn't work, you might try setting ThrI->MaxInput=0.2v instead of ThrI->MaxInput=0.02v just in case the ebrake functionality is handled in hardware directly on Ti instead of by software - it will give a little larger margin for EBK to pull down Throttle In. The revised setting of 0.2v is still well below the 0.37v default voltage of Ti so there should be no ill effects by making the Setup change.

I see what happened, it was the way I was testing it, but this brings another problem. The motor stops producing when using the ebrakes but only when the just throttle is used.
I have a cruise switch on my half twist throttle. I used the cruise switch feature to keep the motor under power while I shorted the two ebrake wires together. If I engage the cruise and then short the ebrake wires, nothing happens, the motor keep spinning under power.

 

[teklektik]

I have a cruise switch on my half twist throttle. ... If I engage the cruise and then short the ebrake wires, nothing happens, the motor keep spinning under power.

Yep - this is in the same vein as the earlier remark about regen:

As a possible downside, if your controller is programmed/jumpered for regen when ebraking, that functionality will be lost if the ebrake is fed to the CA. If this doesn't apply to your configuration, then it looks like you could save that extra ebrake wire running back to the controller .

My bad - I was too specific in calling out (only) regen as a consideration. More generally: if you are using any controller function that relies on an ebrake input then you must run the ebrake connection to the controller.

Too bad this didn't work out for you, but the discussion called out an interesting point. I've updated the unofficial setup instructions to reflect these considerations.

 

[Degull]

Not a big deal to run a couple of ebrake wires to the controller but I plan on rewiring the throttle into the CA V3. The plan is to streamline the amount of wires going back to the controller. Thanks for clearing that up for me.
The cruise feature is a great, is there anyway to wire a cruise switch to the CA V3?

 

[teklektik]

The cruise feature is a great, is there any way to wire a cruise switch to the CA V3?

Although it's possible to hook up a primitive cruise control feature with the CA, there's no means to use a conventional cruise control switch or anything that will release cruise control mode permanently when you hit throttle/ebrake. This release mechanism seems a valuable safety feature and although such functionality wouldn't be too hard to implement with a few parts, wiring to the controller is so much easier...

Hmmm - but you might be on to something here....

• Justin-
  The v3 torque sensor feature is very nifty but is arguably one of the least used features. However, it offers inputs that might do double duty to implement cruise control for those who don't have a torque sensor. And since torque sensing and cruise control reflect sort of competing philosophies, usage conflicts should not be too common.
  
  How about an alternate mode for the torque sensor Dir input that would allow it to be tied to a pushbutton to hold the current Ti (throttle input) value when the button is released after being held for ~5 seconds? Cruise would disengage immediately when the button is depressed or on ebrake. To adjust the cruise up/down just hold the button in, adjust speed with throttle and release. To disengage, tap the brakes or the button.
  
  Re-purposing one of those throttles with an integrated Regen/Ebrake button as a Set Cruise button would give a nifty cruise functionality with minimum clutter. (Or some other equivalent scheme... :wink: )

EDIT - proper answer by Justin here for planned CA cruise control functionality.

 

[lizardboy]

Well Thats it for me for a bit.

Every single software update brings a new deal breaking problem. No one else seems to have them

I haven't at this point had a correctly functioning unit for longer than two days. Thanks but I've just rewired and gone back to my old cycle analyst. Nice box doesn't try and do too much but works everytime. I'll miss current throtle but it's jerky and cuts out randomly anyway. It's been just brutal to try and keep this thing working. I'm happy to have a unit that works again. I just want to ride for a bit with no issues and just enjoy. Hopefully all the things that don't work will be fixed and we'll get a working software next revision but I for one have had enough of the moving target this thing is if I can't measure amperage it doesn't do much folks. The software should be getting fine tuned by now not starting a whole new set of trouble. I have two other bikes under construction for other people but I can't recomend the v3 box to them at this point and I can't sell them the old box due to it's age and expected life cycle.

 

[teklektik]

Well Thats it for me for a bit. ... I'll miss current throtle but it's jerky and cuts out randomly anyway.

Sorry to hear that your beta test experience has not gone well. In any case, you can get current throttle functionality from your v2.23 unit by following the procedures described in section 9.3 of the CA v2.23 Manual.

I have two other bikes under construction for other people but ... I can't sell them the old box due to it's age and expected life cycle.

My understanding is that both the old and new units will be sold but at different price points and to different user audiences. At such time as the v2.23 product is eventually sunset, its case, cables, and wheel pickup should still be readily available as crash replacements since they are identical to those of the v3.

 

[hjns]

Well Thats it for me for a bit.

Sorry to read this, and I think that your feedback was actually very useful. I reread the thread, and it looks like Justin was able to reproduce at least one bug that you identified, so your feedback actually made a difference!

I have two other bikes under construction for other people but I can't recomend the v3 box to them at this point and I can't sell them the old box due to it's age and expected life cycle.

I have to partially agree here. I also would not recommend a beta-version to other people. However, I feel that the main purpose of selling a beta-unit to the people is in getting all the bugs out while at the same time provide early adopters a chance to get their hands on something really new. That means to me that I don't expect any beta-unit to be completely functional, and I expect that I will encounter bugs or other faults. At the same time, I expect to find and use new features that are not in the latest "normal" version.

Both of my expectations have come true, and last week I saw smoke coming out of my first CAv3-beta, probably because I connected it in the wrong way. I still need to further diagnose things, but I had connected the PAS unit directly to the PCB of the CAv3-beta. As I was running 30S 126V, the internal DC converter may have gone a bit too warm, I saw the magic smoke and then the unit went black. I have another CAv3-beta in the mail, and it is stuck in Customs, but I am looking forwards to continue tinkering. To me, this is part of the game when tinkering with beta-units, whether it be software, or are combo of new hardware + new software.

However, for other people I would not hesitate to recommend the CAv2 to them. I think it is a great unit, and for me it works flawlessly with my Cromotor and a current throttle. I think it is more about managing expectations of other people. If you want to recommmend a Bionix-like unit, then obviously it is not going to work. Recommending the CAv2 as a unit that does the following means it will sell itself:

Once connected to the vehicle's power bus, the Cycle Analyst calculates and shows the following information:

• Volts, Watts, Amps: The instantaneous electric power being drawn from the battery.
  Amp-Hours, Watt Hours: The net energy that has been pulled from the pack since the meter was reset. The accumulated amp-hours let you know the remaining energy in the battery pack with far greater accuracy than a voltage or LED indicator.
• Speed, Distance, Time: All of the basic features of a bicycle computer, including the average and maximum trip speeds, plus a total distance odometer.
  Regen, Wh/km: There are several quantities this meter will calculate that you don't get with other instrumentation, such as the % extra range that was gained from regenerative braking, as well as the vehicle's average energy use in Watt-hours per kilometer or mile.
• Peak Currents and Voltage Sag: Records the maximum forwards and regenerative currents as well as the minimum voltage on the battery pack.
• Total Battery Cycles and Amp-Hours: Retains life cycle information on the vehicle's battery pack, such as how many charge and discharge cycles the pack has experienced and the total amp-hours that were delivered over its entire life.

Furthermore, the Cycle Analyst has the ability to over-ride the user's throttle and regulate the power delivered to the motor, turning an otherwise dumb ebike into an intelligent device with a user programmable speed limit, current limit, and low voltage cutout.

• Speed Limit: This has the utility of providing legal compliance to the speed cap imposed on ebikes in most jurisdictions. When used with a full throttle, it serves as a cruise control on the electric bicycle. It can also be beneficial to riders who want to increase their torque and power by using a higher voltage battery, without simultaneously increasing their top-end speed.
• Current Limit: An adjustable amps limit is useful to prevent damage to the batteries from excessive current draw, to increase the range that you'll get on a charge, and to protect the motor controller and motor in setups that draws too many amps.
• Voltage Cutout: A low voltage cutout is used to protect a battery pack from being discharged too deeply, which can cause cell reversals in NiMH/NiCad packs, permanent cell damage in Lithium packs, and sulfation in Lead Acid batteries. The programmable low voltage rollback allows you to set an appropriate low voltage point tailored to your pack.

 

[lizardboy]

The pinout i am using for the programmer cable is as follows:

with the notch at the top the furthest left pin on the lyen programing connector is the ring the next to the right is the tip and the next or center pin is the sleeve. i have the gnd supplied all the time to the sleeve. ( no switch)

 

[shorza]

*FIXED*

 

[MattyCiii]

Quick question for those who use the V3 with RC motors:
Is the amperage of my throttle output perhaps too low?

Last night I bench tested the setup with bare minimum: CA, throttle, HV-160, motor and one battery (6s1p, 5.8 AH). Worked perfectly.
Today I went to ride and it doesn't work. Everything lights up and I get the "turn on" beep from the HV-160 (but not the "throttle is connected" tweet noise). I rotate the throttle and see the CA is recognizing the input. Visual inspection of the connections - everything looks fine.

The only differences between the successful test last night and today are:
1) The bile was upside-down last night, right side up today
2) The battery. Today I'm using 8s1p, 5.0 AH. Could it be the lower AH rating (I think these batts have lower C rating too) that is causing my loss of function?

By the way, I'm getting +5v for the throttle from the throttle input pad in the CA, as shown in this picture. http://endless-sphere.com/forums/viewtopic.php?f=4&t=37964&start=270#p573801

It was a marginal crimp that worked on the bench but not on the bike.

 

[hjns]

Hi there,

I don't run an ebike with an RC motor, but maybe I can ask the right questions:

• An 8S1P 5ah lipo should still be able to deliver 25C aka 125A. So that should not be the problem with your ESC.
• The ESC seems not to receive a throttle input. Can you measure a positive throttle signal at the ESC, after it got transmitted from the CAv3?
• Did you program the ESC for a specific lipo input? I don't have experience with Castle ESCs, but most ESCs need to be programmed for the specific voltage input that you give. 8S may be too high for the specific settings.

 

[cruzxia]

Thanks lizardboy I got it working, I had it wired correctly, It was the boot sequence that was the issue. Blank screen on the CA, is when the boot loader is active.

Cruzxia

 

[MattyCiii]

Way late in the game but...
I have a BionX Press Switch for Magura brakes. It's very close to the threshold for e-brake activation sensitivity for the CA. I actually have to wiggle the brake lever around a bit when I boot up the CA in order for the CA to consider the brake "open". Once in a while when I ride, I hit a bump and the CA thinks I've hit the brake.

I suppose I could wire a resistor in series to the connector and drop the sensitivity (though I'm probably way off on that. And I'd be hard press to guess a value, and whether to go series or parallel, etc.)

What would be great is if this could be parameterized, and set - just like the throttle. So, you would go into a setup menu and see what the CA senses in both the "brake at rest" and "brake activated", and you would set the point in the brake pull where the CA cuts the throttle.

Would such a thing be possible? I hope such a feature would apply to all types of throttle, not just the niche product I decided to buy...

 

[teklektik]

Way late in the game but...
I have a BionX Press Switch for Magura brakes. It's very close to the threshold for e-brake activation sensitivity for the CA. I actually have to wiggle the brake lever around a bit when I boot up the CA in order for the CA to consider the brake "open". Once in a while when I ride, I hit a bump and the CA thinks I've hit the brake.

I suppose I could wire a resistor in series to the connector and drop the sensitivity (though I'm probably way off on that. And I'd be hard press to guess a value, and whether to go series or parallel, etc.)

From this thread about the Magura Brake Pressure Switch, this unit appears to provide a simple switch closure - which makes sense. It does not seem that CA parameterization is the answer.

The switch you are discussing uses a conventional barrel connector offering only two connections like the one in the thread - looks like the same unit...


Is this a new installation, or did it work previously when connected to your controller? If so, what controller do you have?
Try connecting a meter on the leads to determine what is happening there electrically.

 

[MattyCiii]

Thanks for the reply.

When I first installed my V3 CA it worked perfectly. But my CA installation has been on-again, off-again, as I tried changes to wire routing, etc., and now it's behaving as I said above. Originally I cut off the barrel connector and soldered on JSTs, to make the brake switch mate with the stock V3 CA. Since I've found other problems with my CA, I just now did some tests with a continuity tester.

What I've found is if I adjust the brake handle throw/wear adjust knob to full open, the brake switch operates as expected. So, this is definitely just a case of an over sensitive switch.

 

[naix]

hi justin.

I am testing your product CA V3 now . it's fun

By the way.　I want "torque + pas ctrl "mode too.
ex)
when starting , use torque sensor mainly,
when cruising , use pas sensor mainly.

<<I want to start safety with a high response pedaling like a japanese ebike>>
In the composition of only PAS, the motor drive with a high response without delay is very dangerous.
because It malfunctions by Pedal position adjusting.
But in Pas+Torque sensor with speed mapping , I think that we can ride safety.

The torque sensor can realize a high response　for starting dash without rotating pedal.
(of course, need torque threshold setting and reduce power with speed mapping)

The Pas sensor can't realize a high response. because it need delay or need to rotate pedal when starting.
But the Pas sensor can realize easy riding when cruising.

so , it use torque sensor mainly when starting and use pas sensor mainly when cruising.
It's good solution ,Don't you think?

naix

 

[shorza]

Hi Justin, I have just installed the CAv3, and think it's a great piece of equipment.

The perfect position and angle for mine mean the screen and button are upside down. How hard would it be to program an 'upside down' mode?

 

[amberwolf]

IIRC there should be enough wire length and whatnot for everything except maybe the button connector to install the CA/display PCBs upside down, then just make a little jumper cable to the buttons. In the jumper, just swap the two button wires and now they are reversed, too.

 

[justin_le]

Justin, I have some further data which may explain the problem I'm having, and also seems curiously related to what lizardboy has described regarding his current readings being doubled.
Now, I just did a Zero Cal in the 16 code, and I see the following:

What's up with that?

Hey Joe, that is very observant and indicates that somethings has gone funky at the hardware level. When there is no current through the shunt, then both amplifier circuits inside the CA should be outputting 2.5V which is what gets displayed on the screen. If you are seeing a value of 1.18V when you hit "zero amps" then either one of the op-amps is damaged, there is a bad connection in one of the sense leads to the shunt (white or blue wires), or you actually had a substantial amount of current flowing when you hit the zero amps routine. Is there a chance you can try reflashing with the B15 firmware and confirm that with that as well, you also get 1.18 and 2.4V?

I have similar voltage offsets with 1.18 and 2.43

OK very interesting, this is almost for sure the source of all your issues too. The first reading (1.18V) is from the low x10 gain amplifier, while the 2nd voltage (2.43V) is from the high x100 gain amp. Normally if the 10x amp is at 1.18V, then the 100x amps should be saturated at 0V. While if the 100x amp is at 2.43V, we'd expect the 10x amp to be at 2.493V We'll probably need to have you ship these back for us to look at and try to figure out what caused this. It might be related to the beta PCB's having exposed via's and hence being more likely to have incidental shorts form on the board.

Is anyone else seeing something other than 2.48-2.50V after doing a current zero?

-Justin

 

[justin_le]

Not sure if this has been discussed but I was wondering if it would possible to control the ebrakes with CA V3 but using a legacy throttle. I have my throttle wired directly to controller at the moment and I was wondering if it is possible to wire the ebrakes into the CA V3 instead of the controller. This would save some wires going to controller and a cleaner install. I'm not sure if the CA V3 can handle this since I have the throttle features disabled. I tested my legacy throttle and the ebrakes by hooking up an ebrake lever to the CA V3 and it did cut power to the motor. Is there something I could do other than wiring both the throttle and ebrakes directly to the CA V3?

Hi Degull, you can certainly have your throttle go directly to the controller and have the ebrake cutoff go just to the CA, and then use the CA's throttle over-ride line either to activate the ebrake input of your controller or pull the throttle signal of your controller down. No issues there, and no need for the CA to have a throttle attached in order to take advantage of the new ebrake input.

Justin

 

[justin_le]

Is there any plan to offer a constant-current cruise control capability when running a current throttle through the CA, one that allows the operator to "dial in" a fixed power level of assist?

I haven't played with cruise control yet but on the agenda is to do the same style of "auto-cruise" that exists on some motor controllers. So if you hold the throttle steady for a fixed amount of time, then you can release the throttle and the CA would preserve that input until you either engaged the ebrakes or touched the throttle again. In that way if you had the CA setup as speed throttle, you would have proper cruise control, if you had it setup as a pass-thru throttle, you would have voltage cruise control, and if you had it setup as a current throttle you would have a power cruise control.

Is that more or less what you are asking?

-Justin

 

[justin_le]

It seems that when the ebrake is attached to the CA and is applied, To (Throttle Out) is forced to ThrO->MinOutput.

It was originally like that, but in the more recent firmwares I've changed it so that when the ebrakes are pressed, the throttle output goes all the way down to 0V rather than just to MinOutput. That way, in principle you can tell from the controller end if the user has simply released the throttle / exceeded a limit (Throttle Out = MinOutput), or if they've engaged the ebrakes (Throttle Out = 0V).

As a possible downside, if your controller is programmed/jumpered for regen when ebraking, that functionality will be lost if the ebrake is fed to the CA.

Correct, but with the update above you could wire up a comparator circuit that compares the throttle line with about a 0.5V threshold, and send the output of that to the ebrake line. Then whenever the throttle signal drops below 0.5V, it will engage the regen. We haven't tested this out yet on a bike yet but in principle it should work OK.

-Justin

 

[mrbill]

Is there any plan to offer a constant-current cruise control capability when running a current throttle through the CA, one that allows the operator to "dial in" a fixed power level of assist?

I haven't played with cruise control yet but on the agenda is to do the same style of "auto-cruise" that exists on some motor controllers. So if you hold the throttle steady for a fixed amount of time, then you can release the throttle and the CA would preserve that input until you either engaged the ebrakes or touched the throttle again. In that way if you had the CA setup as speed throttle, you would have proper cruise control, if you had it setup as a pass-thru throttle, you would have voltage cruise control, and if you had it setup as a current throttle you would have a power cruise control.

Is that more or less what you are asking?
-Justin

Yes. Excellent. Will you be adding this to the the CA V3, or is it on the agenda for some future version?

Other related features are:

1) A "Resume" capability. It would be nice to return to the same cruise setting after, say, tapping the brakes. Sort of like the "resume" feature on an automobile cruise control.
2) A way to incrementally adjust the current cruise setting up or down, again, as seen in automobiles. Lyen's version of Infineon controllers are available with a separate cruise control that functions much like those in an automobile, minus the "resume" capability.

http://imageshack.us/photo/my-images/708/cruisecontrol.jpg/

Unfortunately, the Infineon controller throttle and cruise control cannot be used in current mode, just voltage/speed mode.

 

[justin_le]

Hi Justin, I have just installed the CAv3, and think it's a great piece of equipment.

The perfect position and angle for mine mean the screen and button are upside down. How hard would it be to program an 'upside down' mode?

There is no way to flip the orientation of the character LCD screens in software, so you would have to remove the circuitboard and rotate it 180 degrees as Amberwolf suggested. Unfortunately though the LCD bezel is not symmetric so it doesn't fit very well in the enclosure this way.

-Justin