Answer from Holger on the sequencing. Procedure in the case of failsafe is: 1 go to CH-altitude (defined in the sample to 30m) and stay on for 5 seconds did position (PH) 2 fly home (Usually the speed is 7-8m/sec) 3 at homepoint go down: 3.1 4m/sek descend until altitude of 20m 3.2 2m/sek descend until altitude of 10m 3.3 Use the 'landing-speed setting' for the load-10m for descending above This procedure may take 60 second (what you set as failsafe time). After this time the MK uses the "gas emergency" for the set time: 4 motor to emergency gas for the selected time (in my example 85% for 9 seconds hoover gas) If the RC signal comes back within the procedure, the failsafe is canceled at the MK is back under RC control. Regards, Holger So in a situation where your fly back and land time is greater than 60 seconds (set in the Failsafe CH Time) the copter will switch from using the new descent rate value to the Emergency Gas setting. So here is an example. You are flying the Par 5 Pink Dogwood at Augusta. 575 yards (525m) but a bit longer overall since you are starting behind the tee box and want to pass the green to show the back side of the green. So let's call it a total from power on/home point out to 620 yards (566m). And we have planned to fly it as a straight out from behind the tee. Well straight out in the center of the fairway taking into account that the Pink is a dogleg left hole. Way out there at the end of the outbound leg the TX battery decides it has had enough and shuts down triggering the fail safe event. 0 - 5 seconds - copter climbs to CH altitude of 50m 6 - XX seconds - copter starts coming home at 7m/sec Total time to get home 81 seconds from 566m. NOT GOOD! Ok so the plan isn't going to work very well. Reality would be: 0 - 5 seconds - copter climbs to CH altitude of 50m 6 - 60 seconds - Copter heads back but only goes about 385m Copter goes into Emergency Gas and sets down about a third of the way down the fairway, likely in the trees since it is a left dogleg and the direct flight path home is over the trees. Lesson learned, adjust the furtherest distance you fly out so that you have the time to get home. Here's the example using the same scenario. 0 - 5 seconds - Climb to CH altitude of 50m 6 - 32.5 seconds - Amount of time available to fly home using 50m CH altitude - 192m 32.6 - 55 seconds - Enter set descent rate of 1m/s From 50m First 30m/7.5 seconds, next 10m/5 seconds/ and final 10m/10 seconds. 55 - 60 seconds - safety margin so that the ship doesn't go into Emergency Gas So the safe play would be that the target is only 192m from the takeoff/home position. An increase in CH altitude with shorten the CH distance and a decrease in CH Altitude will increase the safe distance. In the case of the Pink Dogwood hole the best plan would be to take off from some midpoint on the hole and from a position, likely on the right side so that the CH altitude can be kept down at 25m which would increase the distance to about 236m.
Gary great work! Remember Holger says CH shouldn't be any higher than 30 meters? I think I'm starting to see why Holger is telling us have a come home altitude of 30 meters is so that the copter can't accelerate faster than 4 m/s because if it accelerates to 6 m/s (which mine did) the copter would not slow down enough to reach 2 m/s at the 20 meter mark and hit the ground very hard and most likely do extensive damage. The problem is that that in many circumstances you need a CH of higher than 30 meters. Case in point, a par 5 that slopes down hill or even a par 3 that has a significant drop the copter will slam into the hill, fairway or tree's upon its CH route. If you even use my street for example its a long straight road to a dead end. I live near the dead end. Half way up the road it goes up hill and by the time it levels off its about 15 meters higher. I use this road to test my FPV range. If I go 300 meters down the road, the road is 15 meters higher than its start position. Many trees on my street are 30 meters high. I need a CH of "at least" 45 meters so that it clears all the trees on the way back and that might not be quite enough safety margin. I would say 50 would be more comforting. So what do we do when we need more than 30 meters for CH altitude? Now I can't prove my theory on why Holger wants CH of no higher than 30m but I don't feel good about testing it. Let's say we want to test this, if at 20 meters its falling faster than 2 m/s, by the time you react to this by turning the radio back on the copter will probably be no higher than 15 meters and there won't be enough time to turn the radio back on and for the copter to sync up with the radio and pull up before it hits the ground. I'm seeing about 1.5 seconds before the radio turns back on and the copter recognizes the signal, and it will need at least another 0.5 second for corrective action to avoid the crash so I'm estimating a 2 full seconds. If its falling faster than 2 m/s at 20 meters (remember mine was at 6 m/s and climbing at 30 meters if the test proves the theory I think the copter is in for some big damage. But maybe I'm totally wrong about this. I believe we need a guinea pig for this project. I would volunteer but I just had a free fall at 300 feet that I'm just starting to recover from. The guinea pig would need to weigh at least 11 pounds since the majority of our pigs weigh at least 11 pounds. 12 pounds would probably be better. I can't see many of us having combinations less than 11 pounds. For the test we don't need a gimbal on it just about 5-7 pounds of dead weight attached to the bottom of the CS8. If anyone want to help piece together a test pig I have 8 motors, props, a battery tray, stand offs, batteries, a GR16, and a couple booms and a few other odds and ends that I can donate. I also have a PSD that will work with 3 new BL's. But we would need a FCB with the new ACC, two center hub plates, and a NAV/GPS setup. Then again if you really think about it, how much longer will the majority of us being running MK electronics once SYNAPSE comes out? Until then its probably just easier having a spare Graupner battery in your pocket.
[EDIT] Sorry...didn't see Gary/Holger's response before I posted. Looks like you got the answer!] I looked for that information before when I was working on the Rather Good Beginners' Guide® for MK Tool....and I couldn't find the answer. The source code is a little hard to reverse engineer as it's all dealing with real time events -- when I get a chance I'll email to Ingo Busker and ask. The documentation just makes reference to descending "at a predetermined rate" which could be Emergency Gas, or it could be "Advanced High Speed Plummet." Andy.
Little update for everyone. When I upgraded my N91 on Tuesday I really didn't get much of a chance to extensive fly with the firmware because its been raining a lot. Yesterday was my first chance I had to really fly. For me, I'm having problems with stability in PH. It's doing a mini toilet bowl effect. Also come home was skating around a lot and took 4 times as long to get home. Reading the install directions on the website they want you to change a few settings in the Navigation tab. I did that before the install and it said that if the PH starts moving around to reduce the GPS ACC which I did. I went from 40 to 30 to 20 to 10 and it didn't make much of a difference for me. I also tried a GPS D of 100 and that did improve come home but it was still skating around in PH. Because I have a shoot tomorrow I can't really be playing around with me much more today so I reinstalled N90J and its back to normal rock solid. It was a bit deceiving at first after I installed N91 because from a quick hover it was fine. It was only when I started going 30 to 40 meters up did it start moving around. Even then when I was trying it at lower altitudes around 15 meters it was still moving around about 10 meters. Not saying that anyone else will experience this as I might have not followed something completely but I did copy Gary's settings to the T. Except I used QC's recommended Z-ACC setting of 50. Just a heads up for anyone else that tries the new beta firmware.
Did you check the Kp Index for the GPS system, Dave? Usually I go to n3kl.org, but that appears to be down. So check http://www.swpc.noaa.gov/rt_plots/kp_3d.html Correct for the Time being UTC -- you're five hours behind, I think? KP=4 is a bit marginal.... Andy.
Andy I use http://n3kl.org/sun/noaa.html but its been down today. I flew with N91 last night and early this morning and had the same results. I then switched back to 90J within an 1 hour of my last flight using N.91 and it was very solid like I'm used it flying. I think the issues for me are the GPS settings. I would have liked to try the 90J GPS settings with the N91 firmware but I just ran out of time. I knew it would be raining a lot today and I needed to make sure the copter was good for a shoot tomorrow. If I was to describe what I was seeing it was more like a mini toilet bowl affect that had a radius of about 5-8 meters. Also I just couldn't fly straight with it and come home didn't come home straight like it normally does.
Has anyone else tried the new Beta N.91? I'm curious to see someone that's using the MK PDB has tried it and if the recommended GPS settings are working for them.
Hello Dave, Adjustment of PID values in MKTool will typically not remedy this "toilet bowl effect" you are referring to above. The usual causes of this effect are that the compass is not calibrated or you are getting EMI from the main battery leads. Greetings, Adam
Thanks for the reply Adam. Maybe I'm not describing it right. It might not be as of a toilet bowl effect as much as the copter is moving around "a lot" in PH and definitely not stable in CH. Like I mentioned above I went straight back to 90J firmware and the copter is solid as a rock like it was before. I did before/after comparisons within an hour of each other and at the same location. I have over 160 hours on my copter, and I never seen this type of in-flight behavior before I switched to 0.91L. The one thing I didn't do "after" installing 0.91L was to recalibrate the compass. I don't think that would be an issue because I don't think Gary did that and its working for him. The GPX data showed the compass values all good. I am definetely not getting EMI from my battery leads as I twist and heat shrink them. My battery leads also run to the back of the copter toward boom 5 and they not excessively long. I have the GPS and Nav out on boom 1 like most people here. I have a shoot tomorrow and I'll have more time to play around with it. I'll try to have my wife video tape the copters behavior with 90J and then with Beta 91. I was thinking that maybe I just didn't do the install right and I"m missing one stupid setting somewhere.
Dave I think I am on 0.91j but it might be 0.91k or l. No way to check as I am on the backwoods of Wyoming, at least presently. No I did not redo the compass calibration. I'm confused by your reference to 'N91'. Are you saying there is a release of 0.91n?
Sorry I was just referring to the Beta 0.91, which is officially listed in my download folder as FC_0_91_NC_031 Beta that's why I had 91N in my mind. The actual version is 91L.
I got my FC back from Riley at QC (thanks, guys!) and finally had some time to get the copter put back together again. I flashed it with Beta 0.91 (NC, too), and carefully went through all the contingencies that others have shared. First flight was completely uneventful, with the copter handling beautifully. flew a pair of 8000 batteries testing everything, and it was solid as a rock. Thanks for all the advance assistance, folks!
Steve thats great. IF its not too much of a bother, would you mind posting a screen shot of all your settings? If you can't do them all if you could just do the Altitude and both of the nav pages I would appreciate it
I reloaded the 0.91L firmware and appropriate navigation firmware and retested. I used all the same parameters and flew it on 3 sets of batteries. It appears to be very close to 90J as far as PH. Just a hair more of skating around but I"m talking a couple feet. After 30 minutes of flying I feel much better and have to chalk up my issues to either a bad install, a botched setting somewhere that I didn't catch, or just stupidity. I"m leaning toward the later.
Believe it or not, I just accepted all the default settings. I think Riley did something with the FC board, but that was probably nuked when I upgraded to the Beta firmware. I think all I had to do was turn on Altitude Control. I will check the next time I have her on the bench. After almost 10 flights and 12 batteries, the only glitch I had was that my FPV link went down on my last flight. Of course it was right after I'd offered the DP a "ride" on the copter by way of wearing the Zeiss Cinemizers for a flight. Sigh.
Michael this was posted earlier in this thread. "The zip file contains two FC hex files. A MEGA644 and a MEGA1284. If you have a black board it has the MEGA1284 chip."
Just got the new ACC board installed, upgraded FC from 0.88n to 0.90j. Also updated NC to 0.30h. Did some test flights yesterday and the new altitude hold was amazing with spring loaded throttle! Just wow, it was so much nicer than before. So everything went nice and smooth until I tried PH and got quite significant amount of toilet bowling. I didn't have any problems with the old firmware and my gps/navi is lifted almost 50mm from the motor boom so should I just recalibrate the compass? I did not calibrate the compass after the FC update 0.88n-->0.90j.