CA3 problem oscillating throttle output voltage

[John in CR]

I'm trying to use the CA3 throttle functionality for the first time and I'm running into an issue of an oscillating throttle output voltage from the CA3 with the throttle at zero. The voltage goes high enough to spin the wheel, so I can't go for any kind of test spin. On both of my available CA3's the firmware version (beta 21) does not include the Throttle out 0 throttle threshold, which brings up another problem. For some crazy reason the CA3 utility will not connect to my CA3's, so I can update the firmware. I check the com port in admin tools and it has the right port number and my XP computer says it's functioning properly, but I can neither read what's in the CA nor update the firmware in hope that a newer firmware version has a setting to get rid of this oscillating voltage issue.

I couldn't find anywhere that says I need a diode or resistor on the throttle line with a CA3 like used to be required between he controller and the throttle override wire on a CA2...Maybe that's my problem, since I have the CA3 going directly to the controller throttle wire (actually 2 controllers' throttle wires, since I have dual controllers).

 

[amberwolf]

With teh CA3 you don't need diodes/etc in the throttle line.

When you say it's oscillating, do you mean that on the CA screen itself, (left button once from the main screen, IIRC) where you can monitor the throttle voltage in and out, you can see the voltage output changing, with no throttle connected to the throttle input of the CA, and no controller connected to the output of the CA?


If so, I'd say the first thing to get working is the CA-to-computer thing, so you can update the firmware (or at least change it to a different one, and also so you can save and send setups to experiment with things without having to go into the menus all the time).

When connecting the CA to the computer, disconnect the CA from everything except the computer and the power source of the CA, to eliminate interfereince.


If the problem isn't that the CA shows the voltage changing on it's monitoring screen, then the CA isn't creating this voltage, and it might be from the controllers being connected in parallel at the throttle--sometimes this causes a ground loop.

 

[John in CR]

With teh CA3 you don't need diodes/etc in the throttle line.

Thanks for confirmation of that AW, though I'm still considering a diode near each controller on the throttle line to block voltage differences between the 2 controllers, since other than batt +/- it is their only common connection. Is their anything wrong with that? I think it's a good idea, because many of my dual controller rigs have had issue with one controller getting a throttle fault during operation that requires either letting off the throttle to zero or cycling the key on/off to get going with full throttle.

When you say it's oscillating, do you mean that on the CA screen itself, (left button once from the main screen, IIRC) where you can monitor the throttle voltage in and out, you can see the voltage output changing, with no throttle connected to the throttle input of the CA, and no controller connected to the output of the CA?

I see it on the CA3 screen with about the same timing as the throttle ramp up and down setting, but only when the throttle is connected to the CA3. It goes from about zero to 1.4V, and near the upper end starts to spin the wheel. Since I plan to use it on the power throttle setting, tomorrow I'll try throttle ramp up and down timing of 0. The .20sec/v I set was to go conservative for initial test rides, but now that I think about it, all my bikes except one bike with a throttle tamer have instant ramp with direct throttle connections to the controllers. Current ramp up times are handled by the controllers.

If so, I'd say the first thing to get working is the CA-to-computer thing, so you can update the firmware (or at least change it to a different one, and also so you can save and send setups to experiment with things without having to go into the menus all the time).
When connecting the CA to the computer, disconnect the CA from everything except the computer and the power source of the CA, to eliminate interfereince.

It's one of those frustrating computer things that eats hours and makes me want to pull the plug from the wall and chunk the whole shebang out the 3rd floor window. There's no interference, it's just each of the CA3's, a 20s battery, the TTL serial wire from ebikes.ca, and multiple proven to work USB ports on my computer. I'm not doing it on the CA of a working bike until I get this figured out.

 

[amberwolf]

Thanks for confirmation of that AW, though I'm still considering a diode near each controller on the throttle line to block voltage differences between the 2 controllers, since other than batt +/- it is their only common connection. Is their anything wrong with that?

The diodes won't necessarily have *exactly* the same voltage drop, so each controller may have slightly different response to the same output from the CA. Plus you have to measure the voltage drop you do get, and then add that to the output voltage you want from the CA for zero and full throttle, so it'll actually get the correct voltage at teh controllers.

I used a diode on a test setup of the trike so I could parallel the throttle and the CA's throttle output (controled only via PAS), so I could have a PAS with a speed limiter (switchable to three levels via presets) that would let me just pedal along and have the bike just go 20MPH, or 10MPH, or 5MPH, and not worry about it (without using any form of "cruise cotnrol"), but still have a non-limited throttle override for those few moments here and there that require it. I had issues trying to get the voltage worked out so the CA would be ablet o correctly control the motor. I am not sure I ever got it quite rigth, because it seemed like every change I made to the levels to get it right changed the loss across the diode, so it was always too little or too much. IIRC I eventually just lived with it being higher than needed, but not high enough to cause the controller to error out.

(at this time I am now running that motor's throttle direclty thru the CA, only, along with PAS, and use the other motor's throttle completely separate...when I eventually get around to rewiring eveyrthing to have a throttle bypass, I'm going to use op-amps to isolate everything from everything else, with adjustable gains so I can hand-tune it with potentiometers ratehr than dicking around in the CA menus trying to get it right over and over and over....)

I see it on the CA3 screen with about the same timing as the throttle ramp up and down setting, but only when the throttle is connected to the CA3. It goes from about zero to 1.4V, and near the upper end starts to spin the wheel.

If it *only* happens when the throttle is connected, then it kind of sounds like an electrical issue (ground loop?). Does it happen when the throttle is connected to teh CA input, but with the CA output *not* connected to anything?

Is the throttle *only* connected to the CA, not anything else, for any of it's wiring? (including grounds)

If the throttle input and output levels are setup for the specific throttle and controller ranges you've got there, then the CA shouldn't output anything higher than whatever you've set for the "off" output level, until the input receives a level higher than it's "off" (minimum) setting.

There's no interference, it's just each of the CA3's, a 20s battery, the TTL serial wire from ebikes.ca, and multiple proven to work USB ports on my computer.

Can you try it with a lower voltage battery, like 10-14s? The CA should be good for at least 100v, but perhaps there's something going on with the regulator at 20s that might not happen at lower voltages, causing a serial problem.

 

[John in CR]

Thanks AW. Slight differences in voltage drop won't make any difference on my setup, just half of the motor would be making slightly less torque. I could always use 2 of the "ideal diodes" I've been sitting on for years and have no voltage drop.

On the bright side, I solved my problem. The CA Setup Utility Program didn't work perfectly on my XP machine. The beta version I had in my CA3's had a compatibility issue and the reset setting to default would freeze up on my machine, but using my son's PC updated it like a champ.

Road test is coming soon, and I'm looking forward to seeing the top speed with my 16s Chevy Volt module. The no load of 107kph is great, especially considering that's at 56V. Full charge will be 66V, bringing no load up to 126. That's with the smallest 13" tire I've ever been able to find at a 19.1" diameter, so my latest HubMonster creation will have great acceleration and hill climbing performance. I may have to mod the pair of 24fet Greentimes for more current. They have 4110's and are mounted for great airflow, so they have a lot more to give than the 70-80A limit they're set for, but I want to go stock first as well as see how the CA3 throttle control and current limiting work out.