Hi everyone - i hope someone has some input on what to do here,
I seem to have some issues running cruise control w/ dual-VESC 6 and the new Wand!
After updating firmware on both VESC, running the entire motor set-up, and pairing with the Wand - it’s seems like only the master VESC, is able to run with the new cruise control setting on the wand, whereas the slave-VESC just kinda keeps accelerating up to Max RPM (slower or faster depending on the CC-speed setting)! The master keeps CC speed on the motor just fine, and adds additional torque if the load on the wheel is increased - the slave vesc-motor just slowly stops the more load added to the wheel while running in CC mode - and it's only a load applied with the hand so far.
It’s as if the Slave-Vesc doesn’t recognizes the specific input sent over the CAN from the Master Vesc in regards of V, ERPM, and amps! I’ve tried to reconfigure and reinstall firmware and reconfigure on both VESC's now for the past 5 hours, but nothing I’ve done has made the Slave work properly with the Master.
Anyone has any ideas of what could be happening?
Cheers!
I do not have a solution to your problem, unfortunately. However, could you please be so kind and let us know whether you are able to adjust cruise control with the throttle and/or by pressing button#1 on the WAND, as shown here > https://www.youtube.com/watch?v=OMTkhU69pJo&feature=youtu.be&t=322 ?
Aye, holding down left button increases the speed in CC, just like in the video. It works from a stand-still as well! But at the moment it only seems to apply for the motor connected to the master VESC! The smart reverse-break works really well on both motors/VESC
Thanks a lot for checking. CC for me only works through the throttle but not by holding down the left button, as shown in the video. However, I have not observed that only the master-VESC but not the slave-VESC is adjusting to a set CC value correctly. But perhaps I have not looked carefully enough either, since I have only been riding with the WAND one or two times so far. The guys from Trampa are perhaps able to help (sorry for being of no use; I am still new to this)?
Hm ... just checked again. What I observed is that when you set CC to a certain value and then try to slow down the tyre connected to the master-VESC (at least I assume it is the master-VESC), it will adjust the torque on this tyre accordingly; however, if you do the same to the other tyre that is connected to the slave-VESC (at least I assume it is the slave-VESC), the torque will not increase. In fact, in CC mode, you can simply stop the tyre connected to the slave-VESC from spinning with your hand. Moreover, if you then slow down the tyre connected to the master-VESC again with your hand, it will again increase the torque and the previously stopped tyre will start spinning again and also increase its torque according to how much pressure you apply to the tyre connected to the master-VESC. So yes, I can confirm what you said, I think.
I am wondering, however, whether this really is an issue? When riding in CC uphill for example, the master-VESC will increase the torque on both tyres to keep a set speed, no? Thus, as long as the master-VESC is able to control a set speed when in CC mode, it should not be a problem, I guess. Or did I miss something here?
Yes kind of - but the slave will always rev the motor up to max speed, regardless of the CC setting, just with pull and torque depending on the speed setting in CC. So for instance a 5 km/h CC will still rev the slave up to 50 km/h speed, while holding up the back of the board, whereas the Master will keep running a steady 5 km/h. It seems like the Master motor takes the heaviest load until it reaches a certain speed for some reason! Going downhill is sort of the same - you can feel how it only applies the break on one wheel. Normal breakin works just fine though!
I've just been down in my parking lot with a couple of friends who runs the same set-up as me, also had their wands. None of em had the problem we're describing!
Picking up the rear end on one of theirs and push CC would cause both wheels to spin equally, with an equal amount of pull and torque on both wheels! CC downhill would also be a lot more smooth and didn't feel as draggish to one side!
I dunno if I've missed something in the App settings which induces this behavior in the control :-)
But hopefully Benjamin has a quick fix in the backhand with his sharp eye, and god-like programming skills :)
Benjamin will look into this topic. The 3.61FW has a small bug for Wand reverse function. You can try the 3.60 FW if you like. FW 3.62 will be released soon.
The Wand with the different CC has an older version of FW on. It must have slipped through on the FW update. Some days before shipping we decided to go for a different CC implementation.
We re-flashed all Wands with the latest FW. Probably one or two PCBs slipped through that update. You can follow this instruction to update the Wand:
https://vesc-project.com/node/1114
Thanks for the reply Frank :)
Updating FW on the Wand, will it work both in regards to Hans Meier's issue with the Wand having the forward pitch as the CC, and will it help with the slave VESC/motor issue as well? Or do you know if that will be addressed in the 3.62 FW update?
Do you have a link at hand for the Vesc-tool 1.18 containing 3.60 as well? The latest version i have on my PC besides 1.19 is the 1.11 version of the Vesc-tool! :)
Cheers
Benjamin bangs on a major VESC-Tool update and will probably finish it over the next weekend.
Till then VESC Tool 1.19 and FW 3.61 will stay the most recent FW.
Unfortunately FW 3.61 does not allow REVERSE mode with the Wand remote. Only Smart Reverse works. I hope everyone has the patience to wait till next week for a FIX.
We will look into the CC. Please note that a bench test is not giving you any results. You actually need to to ride the board!
This is a link to the VESC-Tool 1.18 for Windows, allowing reverse functionality in combination with the Wand. https://send.firefox.com/download/11dd27828ce1bcfd/#4VEK2W1TnGscdOzWFL0AGQ 1
I'll see if i can specify the problem better - I'm not a great writer, but i'll try to create the best image i can :-)
To begin with i updated the FW on the NRF dongle, and FW on both my VESC6's to 3.61. While i had the board on the worktable, i did discover something was slightly of with the the way the slave was handling input from the wand compared to to the master (ie. master vesc ran by a fixed RPM - slave would always spin up to max RPM, slower or faster depending on the speed set in CC)
This initially lead to me doing a short bench-tests before i took it out on the streets. I tested load on my workbench by applying equal pressure and resistance with my hands to the wheels at the same time, while it ran in CC set between 6 up to 12 km/h, just to see if there were any differences in torque and speed. Here i discovered that the slave motor would always stop spinning, even with slight resistance applied, and at the same time, the master would just apply more and more pull the more resistance added - still while applying resistance with my hands on both wheels at the same time!
I then brought the board outside - have to add here, I drive with springs only on the trucks, no trampa-dampas, since i like the wavy, quick responsive, more snowboardish, bowlskating feel more. And i think this might was the reason why i've been able to feel the differences i'm describing below so much!
On straight road the while applying light pitch with the thumb, i would immediately feel a slight push and drag towards the right side, and it looked like the right side of the truck were kind of trying to get ahead of the left side (master controls right motor for me) which i had to compensate for, just slightly. There were no differences in drag while applying the break through thumb pitching, and in general would feel just as even and smooth as it's ever done with my old maytech controller! On normal pitch as well going forward, this unevenness in pull/drag would seem to totally even out itself, if just applying enough pitch in the beginning of the acceleration ( between 75 - 100 % ), but would again be ever so feelable when reaching the desired cruise speed and lessening on the throttle. While testing a bit up and down on one very steep hill we have in our city, I'd start to feel the slight push towards the right side again, unless applying up in the higher end of the throttle curve, which would even this out. Both motors are Trampa 6364 154 kv and are set at 55 max amps each, breaking -25. Because i like speed, and the Lamborghini like acceleration :)
I then wen't on to mess a bit more around with the sexy functions of the wand. The smart-reverse was working as perfect as i would imagine, absolutely no drag to either side when coming to a hold on the fairly steep hill, and very easy to control with thumb-pitching. Also on the workbench when i came home later, it worked perfectly - each motor running in reverse, with perfect equal RPM and appliance to torque even when applying a good load of resistance by hand.
Coming to the cruise control part i did testing as low as 6 km/h up to 35 km/h: Starting flat-ground accelerating smoothly while holding the left button of the wand, I'd feel a just ever so tiny push again towards the right side, and looking down on the truck it kinda looked like it was trying to overtake the other side again - same look and feel as when driving a one motor set-up with just springs in the trucks, no dampas. It's fairly uncomfortable! This became really really pronounced when going up-hill, and there was a great amount of pulling ahead in the truck section of which the master was trucking.
The super sketchy part started, when i was driving on flatground in CC coming to a downhill part of the road. The masters side would start to apply some breakin to keep the speed constant, while the slave still seemed to go forward at as much power it whatever was going with! This would induce a fair amount of speed-wobles even going as slow 20 km/h.
All throughout the afternoons Tuesday and Wednsday, I've been driving around my parking which has a steep ramp and loads of space, with 2 buddies who has nearly identical set-ups. One had the same issue as I with the CC and acceleration, though it was harder to feel, because of his trampa-dampers in the trucks. The other guy didn't have the CC problem, but could for some reason not use his reverse function - while this function does work for me and the other one.
I've had a hard time collecting any numeric for what the heck is going on while road testing, since my Metr app haven't been able to connect to my board since i updated the VESC firmware the previous time back in April.
As well in the vesc-tool while benchtesting i can only see RT data for the VESC/motor currently connected at the time, which has made it somewhat hard to analyse what's really going on.
Connecting the old Maytech remote again instead of the Wand, makes everything run as smoothly as it ever did though.
I hope this info is slightly more usefull, than the other hastely typed messages I've come with! :-)
Cheers! And sorry for the wall of tekst!
Please try out the new VESC-Tool 1.21.
Did you check whether the issue is solved with the VESC-Tool 1.21 and FW 3.62 (yes, I am lazy and I should go check myself, I know ;-))?
The CC is hard to check on the bench BTW. the only thing that would give you a clue would be motor temp doing an extended ride.
Just updated to FW 3.62. You can still stop the wheel that is connected to the slave VESC with your hand when in cruise control, and only the wheel that is connected to the master VESC increases its torque relatively to the pressure applied. Thus, I do not see a change compared to FW 3.61 on the bench. Perhaps MaddiQ could tell us, however, if there is a change when riding?
The behavior you see with the second wheel is expected. The cruise control works by running PID speed control on the master VESC and setting the current of the slave VESC to the same values as the master VESC. This works well when both wheels are on the ground and loaded, but if you have them in the air they can run at entirely different speeds. The reason that cruise control is implemented like that is to avoid having the motors fight each other in any situation (e.g. when turning, after a wheel slip, on weird setups). It even works well if you have different motors with different gearing ratios, which I had on one board for a while.
You can try some logging like this (note that you don't need to run a second app such as gps keeper any more)
and compare [8] Current motors to [41] V1 current. 8 is the sum of all motor currents, and 41 is the current for the first VESC. If you have two VESCs, 41 should be half of 8 if everything works as expected, which has been the case for me so far.
Thanks a lot for shedding some light on this (and everything else :-)), Benjamin.