how can i down grade 1.8 to 1.6 ?
- Thread starter guy1a
- Start date
DesJardins
Member
What are the crash symptoms?
guy1a
Member
1. We both use DL .. i just got back to use the frisky module with er9x and now the naza is acting great.
2. The crash is caused as if the link is down for two seconds. The if in manual mode it flips. If in ATT / GPS mode it stops in midair. Then goes on forward.
https://www.youtube.com/watch?v=BwxbE8AXH1I&feature=youtube_gdata_player
Sent from my GT-I9100 using Tapatalk 2
2. The crash is caused as if the link is down for two seconds. The if in manual mode it flips. If in ATT / GPS mode it stops in midair. Then goes on forward.
https://www.youtube.com/watch?v=BwxbE8AXH1I&feature=youtube_gdata_player
Sent from my GT-I9100 using Tapatalk 2
guy1a
Member
After you select 'upgrade' it opens a new window and counts steps .. login. Authenticate .. login fails .. in 1.8 it connects.
Please don't harm yours,... !
sgs II
I've just experienced this flip behaviour.
Ouch!.
It seems that both motors on one side cut together for a moment. I've had single motor/rotor failures before, and the model drops on one corner. This doesn't it drops on one side. Seems to be the side it is possible moving slightly towards. A few feet later, it starts recovering, and straightens up, and tries to recover from the drop. All motors are working when it does this. I was high enough that the model recovered, I then flew back to base, and landed. Rules out just about any hardware problems. My radio is programmed to trigger the Naza failsafe on signal failure, and set throttle to about 33%. Battery was showing 78%, and this data was coming back fine, with altitude, location etc.. All of these merrily carried on unbroken, so the radio link looks to have been fine.
I'm running cobra motors on a DJI450 frame, with Turnigy ESC's.
For me, if this had happened just a few feet lower, the model would have been badly damaged.
Normally the unit will sit and hold position in GPS mode really well, while it flies happily in Atti or manual mode with just a bit more control input, depending on the conditions.
I've flown my Brother's Phantom for ages, and never seen anything like this.
It's as if some part of the PID algorithm, momentarily overflows or wraps, in some specific combination of control/motion parameters.
Best Wishes
Ouch!.
It seems that both motors on one side cut together for a moment. I've had single motor/rotor failures before, and the model drops on one corner. This doesn't it drops on one side. Seems to be the side it is possible moving slightly towards. A few feet later, it starts recovering, and straightens up, and tries to recover from the drop. All motors are working when it does this. I was high enough that the model recovered, I then flew back to base, and landed. Rules out just about any hardware problems. My radio is programmed to trigger the Naza failsafe on signal failure, and set throttle to about 33%. Battery was showing 78%, and this data was coming back fine, with altitude, location etc.. All of these merrily carried on unbroken, so the radio link looks to have been fine.
I'm running cobra motors on a DJI450 frame, with Turnigy ESC's.
For me, if this had happened just a few feet lower, the model would have been badly damaged.
Normally the unit will sit and hold position in GPS mode really well, while it flies happily in Atti or manual mode with just a bit more control input, depending on the conditions.
I've flown my Brother's Phantom for ages, and never seen anything like this.
It's as if some part of the PID algorithm, momentarily overflows or wraps, in some specific combination of control/motion parameters.
Best Wishes
gtranquilla
RadioActive
According to message 4... the flip only happens is manual mode...... only hesitates in other modes. If in manual mode, then only the three gyros are providing upright stability..... if it is PID controller then for the NAZA you only can adjust the proportional and reset gains I believe....... i.e., P = proportional, I = Integral gain. In the case of DJI products those are P = Basic gain and I = Attitude gain. D = Derivative gain which the Naza does not utilize. Anything is possible if there is a database buffer overfill.... but I don't know how that could be the case here. One would have to examine the DJI source code which they will likely not release. And it would be extremely difficult to open the object code and work backwards to recreate the source code.
Mine was in atti mode at that point.
The only thing I'd changed, was I'd done an IMU recalibrate, and was enjoying how accurately the system had hovered in GPS mode. Then came out to atti mode and went for a fly, then had the flip. I've seen but can't track down some threads talking about a recalibrate resulting in the flip....
However I'd never experienced this with older firmware (which of course doesn't allow you to recalibrate the IMU).
I do have a strong 'suspicion' that there is a hiccup somewhere in the latest DJI code. Problem is that I _know_ from my own flying, that it is so very rare that it is one of those 'almost impossible to trace' events, since it is rare. However it does seem to have very specific symptoms. A total cut on both motors on one side, for perhaps half a second, perhaps when doing a slow slew or the system is correcting for a slew.
Going to see how easy it is to 'go back' to older code, and then not fiddle for a while.
Best Wishes
The only thing I'd changed, was I'd done an IMU recalibrate, and was enjoying how accurately the system had hovered in GPS mode. Then came out to atti mode and went for a fly, then had the flip. I've seen but can't track down some threads talking about a recalibrate resulting in the flip....
However I'd never experienced this with older firmware (which of course doesn't allow you to recalibrate the IMU).
I do have a strong 'suspicion' that there is a hiccup somewhere in the latest DJI code. Problem is that I _know_ from my own flying, that it is so very rare that it is one of those 'almost impossible to trace' events, since it is rare. However it does seem to have very specific symptoms. A total cut on both motors on one side, for perhaps half a second, perhaps when doing a slow slew or the system is correcting for a slew.
Going to see how easy it is to 'go back' to older code, and then not fiddle for a while.
Best Wishes