Just checking if anyone has upgraded from 2.00 to 2.01? http://forum.mikrokopter.de/topic-44207.html One of the notable changes mentioned on the forum is ... "5. GPS Parts of the old GPS routines were re-activated so that the behavior in PH is similar to the old version. Some users felt that the behavior was slightly better at 0.90. Note: In older Versions, the parameter GPS-ACC was zero by default. In this version this parameter is important and must not be zero. Try to increase that if there are still problems in PH." Other changes are ... 7. Other changes: 7.1 - Do not start if MK is moving or the sensors are giving wrong signals 7.2 - Flights with vario altitude-control possible up to 3000m from ground - in older versions it was limited until 950m. 7.3 - Compensation of the barometric altitude error in high altitudes (barometric altitude formular) 7.4 - Reduce Nick/Roll when switching off. That will reduce Beim Ausschalten Nick und Roll-Vorgaben reduzieren. So the MK will not easily turn over if the Nick/Roll stick is used to switch off. 7.5 - virtual menu: showing of the single BL-Currents 7.6 - HOTT & Jeti bugfix: sometimes the min/max-temnperatures were wrong 7.7 - (since NC2.00f) Bugfix: in locations with big magnetic declinations, the MK had an Heading error when flying Waypoints and dPH 7.8 - (since NC2.00f) If the MK starts without GPS-Fix, ComingHome was disabled. Now the Position for CH is the first GPS-Fix in flight -> in most cases that is close to the starting point and better than not coming back 7.9 - The reset-command via the serial interface is ignored if the motors are running 7.10 - Maximum Altitude: Hott Speaks: "Max. Altitude"
Sebastian not yet. I have a new FC 2.5 board to test and will install it on that board. For folks that have been chasing temp issues note item 7.6 HOTT & Jeti bugfix: sometimes the min/max-temnperatures were wrong
I installed this version and flew it a few times the other day. Not enough data to determine if it makes a significant difference. It returned all my settings to defaults, so it's hard to tell. I'm planning to carefully re-do my previous configs (PIDs, etc.) and test, maybe today.
OK, much more extensive info now. I went through all my old settings and restored them to what they were before, and the copter's flying beautifully. PH is still far from perfect, but it does seem better. It still wanders a bit if invoked in any way other than more-or-less North. Which is really weird. But I haven't re-calibrated my compass, so I'll try that. I flew around a bunch today, and cut together some highlights. At 1:30 or so of this video, you can see how AH/PH is working. All in all, I'm quite happy, and looking forward to spending more time in the air now. All of this video is completely unstabilized, by the way. While the MōVI is lovely, the Radians (still) do a really nice job.
Thanks for the feedback Steve ... We're on shoot this week, so i've reverted to 0.90 for reliability for the shoot. But I'll get 2.01 a go over the coming weekend. Let us know after the GPS calibration if it makes any difference to the PH. Looks pretty good already!
I installed it on my MK Hexa XL flight test bed. Seems to work OK -- AH is solid. PH had a bit of wandering to it (Kp Index is very low)....not specifically "toilet bowling," but the worrisome thing was that Come Home did not activate at all. I was about 30 meters out from "home" (having verified that I had GPS lock). I have the GPX file and can see CH enabled for 10 seconds, but the GPS sticks do not do anything and the aircraft looks lit was still in PH. I'll contact MK and ask if this is a known problem. Andy
I found out why Come Home did not activate (thanks to an email from Holger). When you activate Come Home, if the RC sticks are not centered (and this can be either because the sticks are not physically centered or because you have used Trim on the transmitter!), then Come Home will not engage -- it simply goes into Manual flight mode. In other words, Come Home will appear to fail and it appears to fail "live" in that the copter will revert to complete manual control even if it was in Position Hold before you enabled Come Home. There is no visual indication that this has happened other than the copter does not come home -- so watch out! In the GPX file, you will see that the GPS Sticks status changes from D (Position Hold) to "m" -- meaning manual flight mode, but that's not immediately visible, of course. If you have a Smart On Screen display you will see that the Come Home status is not activated and you're now on your own in terms of trying to bring the copter home. This could be a Really Bad Thing depending on why you decided to engage Come Home. The definition of "centered" is that the sticks must be below the GPS Stick Threshold set on the Navi-Control window of MK Tool. The default is 8. In my case, I had significant left roll trim on the MX-20. I will recalibrate the ACC and test again. This seems like a "fail live" situation to me that could well result in the loss of the aircraft so be sure to test that Come Home is working properly before you need to rely on it. Andy.
Say what!!?!? You have got to be kidding. So if I take off, everything good to go, do any type of trim or don't have the sticks dead center it performs a DJI come home to china maneuver? Did you point out to Holger not only the stupidity of this but that such a change with no disclosure what so ever over the same functionality since the beginning of MK is going to lead to lots of failures?
I'm not sure it's "any kind of trim" -- it's "trim that exceeds the GPS Sticks Threshold (default value of 8)." But you're right, it does seem distinctly sub-optimal behavior (Note British Understatement.) On my MK Hex Flight Test Bed, I've re-calibrated the ACC, and I've re-balanced the copter and I'm just heading out to try another flight. There seems to be several strategies we have available: 1. Increase the GPS Sticks Threshold. I'm not sure what the downside or unintended consequences of doing this might be but I'll email Holger and ask. 2. Ask Holger to change the behavior so that at a minimum the copter does not not revert to manual mode but stays in Position Hold -- which would be sooo much safer than manual mode! You could then sort out orientation and treat it as a Come Home failure.... Or do you think we'd want a different action if the sticks are not centered "so I'm not going to Come Home, so there...." Andy.
WHAT!! Surely this is a huge step backwards! What could possibly be the advantage of this "upgraded" feature ?!? I'm sure CH has been a safety net for many out there who have lost orientation etc. have hit CH until the craft is in clearer sight, then disengaged CH to take control again after safely parking it in PH. To be in the crap and then to hit CH and you're in even more crap .... makes no sense. The whole threshold business sounds like aerial russian roulette ... If I hit CH will it : A) fall out of the sky ...B) shoot to the moon ...C) fly away to china ...D) do nothing ....E) actually come home ...Well is seems like v1.09 will have to remain on-board for now
Wow, I never thought DJI would look so appealing. This whole "WWHD" (what will Holger do) nonsense is so infuriating. Or will Freefly save us all with a home run when they ship their Synapse? [collective sigh]
I certainly never realized that the GPS Stick Threshold was used like this. For example http://www.mikrokopter.de/ucwiki/en/MK-Parameter/Navi-Ctrl only describes it being used in the context of Position Hold. I've emailed to Holger asking whether this is standard behavior for prior versions of firmware and also suggesting that a better strategy would be (presuming that the is a technical underpinning to the idea of not going into CH if the sticks are not centered) to simply stay in PH so that the pilot could safely resolve the orientation of the copter and try and bring it home. The current scheme (of reverting to manual mode) is like an autopilot that you think has engaged but is silently sulking. Nobody ends up flying the copter! The autopilot isn't (because the sticks aren't centered) and you aren't (because you think it's in Come Home). I'll keep you posted. Andy.
I'd say this qualifies as P1, if I recall the proper classification of bugs. P1: Critical - The defect results in the failure of the complete software system, of a subsystem, or of a software unit (program or module) within the system. from: http://geekswithblogs.net/ppothuraju/archive/2005/01/22/20664.aspx
I just did another test flight. I recalibrated the ACC again, put a small weight on one boom to statically balance the MK Hexa XL better (actually it was a Mondo Stinger). Flies nicely now. No trim required and Come Home works fine -- but very aggressively. I mean it scared me -- my flight test bed Hexa XL is very light (I get 22 minutes out of a QC 6200 LiPo! That light! ). When I engaged CH, the copter's nose came up 15 degrees (based on GPX file) and it came back towards me like a bat out of hell -- actually 22 meters/second (again, that's what the GPX file shows as Ground Speed -- that's 51 miles per hour). Did I mention it scared the s&*t out of me? I need to find a way to get a slightly more genteel Come Home rather than the rip the eyeballs out of the sky Come Home. Anyway, AH was very stable. PH was good -- slight wander of about one meter or so -- not toilet bowling as such. If I had the copter flying towards me moderately quickly, and let up on the sticks and enabled PH, then it would toilet bowl in a logarithmic spiral, settling down into gentle wandering in about two orbits. So, apart from the major Gotcha with Come Home described above, it all looks pretty good. Andy.
I"m sure happy with 90J. I know in 90J it doesn't bahave like this as I had my trim at 20% by accident and I had hit come home and it came home fine. Did I mention I am happy with 90J.