If you’re not already, try the latest 1.1.7 firmware I posted last month on the wiki (https://freeflysystems.atlassian.net/wiki/spaces/FRP/pages/269320215/Arc200+Firmware+Updates). I improved the hall-sensor autotune robustness. There was an issue previously where you’d get that non-standard orientation message if the motor wasn’t spinning at just the right speed (which was usually in the 500-1000rpm range). It should work well at most speeds now and give a correct error message informing you if it’s an invalid calibration speed. Try 1000rpm with the new firmware.
2 Likes
Will there be an update in the future that includes a more standard method of motor detection that this community is more used to? At the moment, we’re doing detection by flipping another board over on its back and spinning up the wheels like a dyno, then setting the Freefly equipped board’s wheels on it.
On VESC based systems, which most of us are using, sensor and motor attribute detection is done by having the esc spin up the motors using some basic algorithm then getting the feedback it needs from that. I’m not going to pretend i understand exactly how it works, because detecting motors by running the motors seems like carts and horses in the wrong order but they do it that way and it works well.
1 Like
You should not spin up your motors externally. Something is very wrong if that is required, and it won’t result in good measurements. Arc200 also spins up the motor itself for detection (granted through a considerably different technique than VESC, but a similar user experience).
I recall seeing a video a while ago of someone doing this, not sure if that was you. You need to keep turning up the slider in the FOC tune pop-up until the motor starts spinning. Specifically, you’ll have to command a speed higher than RPM Closed Loop 1 which is automatically calculated in a prior autotune step. Once you command a speed this high, the motor should ramp-start and spin on its own then you click the start on the FOC autotune. This may be a bit confusing because the start button does not spin up the motor, rather you spin up the motor yourself with the slide then click start afterwards (at some point I’ll clarify this in the GUI workflow).
2 Likes
We are. 
2 Likes
That probably was us. None of this is intuitive lol
1 Like
@mmaner, @longhairedboy : Thanks for the informations 
Do you consider the Arc200 a good enough product to be purchased ?
Is the Arc200 hardware going to be updated soon or is it considered stable?
If that’s the case, I prefer to wait a more mature one
By the what is the period of your HW release cycle ?
Kind regards,
Boubak
1 Like
It’s definitely reliable and performs very well. It’s one of those weird situations where you have way more capacity than you need, but that just makes the load percentage that much less. I think it would be difficult to kill one of these.
I can see the Arc 200 being the top tier controller for esk8, e-surfboard, carts, etc.
3 Likes
Resistance causes heat. Heat kills controllers dead. If you want it to live, run it at 60% of it’s limits. We’re playing with skin and blood here when things go wrong.
3 Likes
Gospel… !st Moe, Chapter 2
2 Likes
The hardware is considered stable and there are no plans to rev it at this time. The software is of course constantly evolving and that will continue to be true.
Being a new product line, I don’t have any information to share at this point on new hardware design release cycles.
1 Like
@mmaner and @christensent : Thanks for your answers
That looks great I am pretty conviced 
@christensent : A last question, I’m curious : Since when the hardware design has been frozen and tested by Freefly from now ?
Kind regards,
Boubak
1 Like
If the rain will go away, Im gonna get a nice long test run on this board before I tear it down again to do some mechanical upgrades.
@christensent, @tabbytabb is there a built in method for getting telemetry? I seem to remember something about bluetooth, but I just spent an entire 4.5 seconds and didn’t see a BT broadcast.
@boubak The hardware has not changed since we began shipping, everyone has the same hardware and is the same thing you’d get if you ordered today. Testing was completed before we started shipping.
@mmaner The only way to get telemetry is through the telemetry tab of the Windows GUI. You can connect to the GUI over either USB or BTLE. There is also logging where you can download data logs from previous drive runs over USB, but at the moment that is disabled in the latest firmware due to a software bug that we are working on (hoping to get logging re-enabled in the next firmware release).
1 Like
Cool, I’ll just measure speeds using a gps app.
1 Like
I am going to do a build log on this board pretty soon, just want to get some substantial ride time in before I do. I rode for about 6 miles this weekend, just downtown and in the neighborhood behind my house. There’s no long stretches to let loose on, but I hit 27 mph pretty easy in a quarter mile flat straight. The torque is sick, even sensorless I don’t have any trouble starting from stand still. I am using a very liberal acceleration curve as this thing will throw you quick. There’s just so much raw power available.
The only 2 things I would suggest need improvements are the Programming UI and connectors. Ill outline the issues I ran into with both.
Programming UI / Software
It’s very complicated. Coming from VESC configs it’s not very intuitive, which is not necessarily a bad thing, but on it’s on it prohibitivly complicated. I found myself referring back to the youtube tutorial 3 and 4 times before taking the next step, just to make sure I understood what was going on.
My biggest issue was related to resolution. I tried 3 different notebooks and on none of them could I see all of the config windows. As there is no scroll bar on the sub-windows I had to switch back and forth between landscape and portrait to be able to click buttons, it was pretty cumbersome.
I like that these controllers are so configurable, but maybe a frontend where you can enter basic values and then punch a couple of buttons to configure might have some utility.
Connectors
The connectors are robust and will likely never break during normal use. That being said, the 8mm bullets are fairly difficult to work around because of the size. Is that large of a bullet connector necessary? The sensor cable should probably be a more common form factor, that would save some space and be easier to connect to motors since its going to be a solder connection anyways.
I broke the USB cable on one of the Arc 200’s because I was swapping back and forth so much, checking values and testing changes. Either a USB type 3 or JST type connector would be more reliable I think.
The size of the connectors and the amount of lead length on each connector requires a lot of space and they need to be secured when the install is complete. I’m not sure how to resolve that other than maybe a single dongle that all the connectors are on and/or maybe shorten the leads substantially.
The space requirements for the Arc 200 are substantially higher than that of a FocBox or VESC 4.12, so much so that to fit a 12s4p I had to build my pack as a double stack and get a custom depth Alter Wedge enclosure from @psychotiller so that everything fit.
4 Likes
Thanks for the feedback. These are things that we are definitely working towards improving as goals for the product!
If you download the latest GUI (1.1.8), it now has scrolling and window-resize-support on everything to support as low as a 720p monitor.
2 Likes
That is awesome. That is literally the only thing that made this difficult. Thanks for taking care of that.
3 Likes
Want to ask about the sound it makes? How would you compare to other commutation strategies?
Any new testing for these @longhairedboy @mmaner @Kug3lis ? Would these be good on a off road build you guys think?
1 Like