Do not load firmware V4.22

RTRyder

Merlin of Multirotors
I did the upgrade last night and first test flight this morning the CS had bad wobbles and jitters, wandering all over the place once in the air. I reverted back to 4.2 and it's back to it's normal smooth flying self. Several other folks over on RCG ran into the same problem so I recommend staying away from 4.22 until DJI has a chance to figure out what went wrong with this release.

Two features added to 4.22, it allows selection of immediate motor stop vs the 3 second shutdown in ATTI and GPS mode, and the best feature, the ability to export and import the parameters. One feature lost, the ability to select manual arm/disarm vs. armed at power up. I really like the import/export, too bad the release doesn't work well but once they get it fixed you will be able to get the WKM setup for different conditions and then save the file. Choose the appropriate setup for the weather on any given day, load it up and go fly, much nicer than having to burn a charged pack figuring out where to set the gains because the weather is different than last time you flew!

Ken
 

GGoodrum

Member
Wow, glad I waited. Thanks for the info.

Hopefully they will issue a new version, that fixes what they broke, and also takes care of the camera mount twitching/hunting/drifting problem. I know they are working on the latter. I received a number of PMs from Sidney, trying to nail down the conditions of when it happens. Anyway, the programmers are working on it, so hopefully they will include this with the the 4.22 fix.

-- Gary
 

RTRyder

Merlin of Multirotors
Wow, glad I waited. Thanks for the info.

Hopefully they will issue a new version, that fixes what they broke, and also takes care of the camera mount twitching/hunting/drifting problem. I know they are working on the latter. I received a number of PMs from Sidney, trying to nail down the conditions of when it happens. Anyway, the programmers are working on it, so hopefully they will include this with the the 4.22 fix.

-- Gary

That reminds me, after I loaded 4.22 my normally steady AV130 started drifting and wandering around while sitting on the bench, something it never did before. It reminded me of how the Coptersky mount on the quad moves around while on the ground and then seems to work Ok once in the air. After I reset back to 4.2 it stopped moving while powered up on the bench, weird...

Ken
 

GGoodrum

Member
Now that is very weird. Did you let it run until the LED started flashing only twice? On mine, it is fine until it finds the 5th sat, regardless of what mode it is in, and then it starts the drifting/wandering. Very repeatable. You don't even need to use a mount. You can simply plug in a servo, pretty much any servo, wait for the LED to start flashing twice, and you will see this hunting/wandering. I only have one WK-M, so it isn't possible for me to check to see if this might just be mine, or if maybe it might be a load issue. If you have time, do you think you can try this test on one, or both of your setups?

I'm really having a hard time understanding how this is such an issue on my setup, and it just doesn't seem to be the case with yours. I just watched your quad video again, with the CopterSky mount, and I just don't see the drifting/wandering problem around the neutral points, like I see in my direct drive setup. Did you say yours does do the drifting/wandering on the ground? I watched my last flight again (http://youtu.be/GdEEJQcn79Q), and you can not only see the effects of the hunting/drifting, but you can hear the damn servos chattering away. This was before I tweaked the gains a bit more, to clean up the roll axis reactions, but the jello still goes away, once the platform angle is changing. It is only when the platform itself is level that I get this annoying problem.

Anyway, if you could at least verify for me if you get the same sort of drifting/wandering with a servo simply plugged in, I'd appreciate it. If you don't get this, after you reloaded 4.2 back in, it might mean that there might be something like a constant, or memory location that isn't getting initialized/loaded properly. In any case, I'm going to try reloading 4.2, and I'm going to try doing a compass calibration again.

Thanks -- Gary
 

GGoodrum

Member
I just did some more tests, and found that the drifting/wandering scales up with the gains. What I mean is that if you increase the gain, like to the max of 100, you will see that angular "jumps" it makes, between steps, is a lot more. The net result is, for instance, if I set the gain on an axis that has a 2:1 reduction, so that I get the correct angular movement that will keep the camera level, the twitches will be the same as my direct drive mount with half the gain setting. I guess all I can do, at this point, is dial the gains way down, so that the amplitude of the twitch is drastically reduced, and then try flying straight and smooth. Then, when DJI comes up with a fix, I'll boost the gains back up so that I can eliminate platform movement from the equation.

Also, it turns out I had 4.0 loaded still, so I went ahead and loaded in 4.2. I even did a new compass calibration. This had absolutely no effect. I also tried every servo I have here and oddly, the twitches are a bit less with "normal" resolution servos, but then then the movements were not near as smooth.

Anyway, I think I am done, for now. I will try setting the gains back down to around 5 or 6, so I get very little stabilization at all, and try flying this afternoon.

-- Gary
 

Gunter

Draganflyer X4
Also, don't be tempted to download the latest version of the assistant software, it won't work with the old firmware! It took me 1/2 hr to find the old version and recalibrate everything!

Gunter.
 

buhojean

Member
Hi, new here.

how to get back to the old firmware?

Edit.
Ok thank I see in RCG.
 
Last edited by a moderator:

andymotor

New Member
Hi, I also updated 4.22, and I found all the parameters back to defualt parameters, so I configured parameters again, and fly my baby, it looks good, :02.47-tranquillity:
 

nisouck

Member
Do not understand, updated to 4.22 with new assistant 1.12, and all is as before.
Now, it is possible to export parameters and later to import them.
 

RTRyder

Merlin of Multirotors
Do not understand, updated to 4.22 with new assistant 1.12, and all is as before.
Now, it is possible to export parameters and later to import them.

Most likely because you're not flying an 825mm hex with an RTF weight of roughly 3.5 kilos. My small quad that weighs about 1 kilo with GoPro camera mount, WKM, and 2 2S5000 battery packs flys nice on any version of the DJI software, the big heavy lift machines are a different story. I WISH it flew well on 4.22 but it clearly does not, very easy to see on the video I recorded with 4.22 and afterward with the older 4.2 version, it's much steadier on 4.2.

Ken
 

nisouck

Member
Most likely because you're not flying an 825mm hex with an RTF weight of roughly 3.5 kilos. My small quad that weighs about 1 kilo with GoPro camera mount, WKM, and 2 2S5000 battery packs flys nice on any version of the DJI software, the big heavy lift machines are a different story. I WISH it flew well on 4.22 but it clearly does not, very easy to see on the video I recorded with 4.22 and afterward with the older 4.2 version, it's much steadier on 4.2.

Ken

My Hexa is about 700 mm motor to motor, TOW is about 3 Kg ......


View attachment 1624
 

Attachments

  • DSCF0903.jpg
    DSCF0903.jpg
    148.7 KB · Views: 324
Last edited by a moderator:

vislaw

Member
Appreciated seeing this thread as I had just loaded 4.22 myself and was able to go back to 4.2 before going outside.

THX
 

If you are going to use 4.22 firmware have in mind that you need to go down in your gain settings by 15-30%. I lowered my gains and it flies more or less the same.
Sidneyw on rcgroups said that there is a gain difference between all firmware updates. This is the first time I noticed the difference in 4 updates.
 

RTRyder

Merlin of Multirotors
If you are going to use 4.22 firmware have in mind that you need to go down in your gain settings by 15-30%. I lowered my gains and it flies more or less the same.
Sidneyw on rcgroups said that there is a gain difference between all firmware updates. This is the first time I noticed the difference in 4 updates.

I have upgraded my first WKM as many times as there have been firmware updates since it was first released, I have never had to change gains after any of them to get the hex to fly right again. First thing I noticed when I tried to fly with 4.22 was that it was acting like the gains were too high.

That this release requires lowering the gains makes me doubt the statement made on RCG... "Firmware has nothing to do with it since we did not change anything other than what we stated; import/export and arming." If that was the case then why does it need the gains lowered to fly right? If as also stated "Firmware version 4.22 is needed for View and Single Waypoint" then they changed a lot more than just adding export and the arming.

Something doesn't add up here, I'll stay with 4.2 until they get their story straight and fix the problem...

Ken

P.S. found the following in the release notes for 4.22 so Sidney's first statement is obviously BS...

Changes:
Two types for cutting off motor: Immediately and Intelligent
Optimize the performance of taking off when throttle is at center


What’s new?
Configuration import and export function
 
Last edited by a moderator:


Tahoe Ed

Active Member
Sidney has removed the 4.22 software from the DJI website until they can do further testing. This version included the ability to use the data link and one click waypoint. That may have been part of the problem along with the fact that you cannot use Heli for JR/Spektrum Transmitters only ACRO. I am switching mine over now. I will miss the TH but I will adapt.

Edit: The issue with the Spektrum/JR transmitters in Heli mode is arming. If you don't use that feature and prefer TH then you are good to go.
 
Last edited by a moderator:



RTRyder

Merlin of Multirotors
Early reports are the re-release of 4.22 works quite well. I haven't tried it yet as it's raining here but I will give it a try as soon as I figure out why one of the motors won't spin up on my WKM quad, I suspect one of the wires broke where they go into the motor casing. If it works ok on the quad then I'll load it up on the CS6 and see how well it works on a large multi.

Ken
 

Top