Question for those who fly a P3P with HDMI module

[Waynosan]

Hey guys, in the second half of year my P3P video downlink started to become severely corrupt after flying for 10 minutes or less. This is what I had posted about this before on another thread:

..... I began experiencing a lot of noise on the live-stream from my P3P. This happened after it warmed up for a few minutes and occurred even when the unit was within a few meters of the TX. Was about to give up and send the unit in for repair when I decided to put the original USB module back on the TX. Cured the problem so it turns out at the HDMI module that I have went bad. Something to try if you're having this issue.

Well, I end up sending the TX w/HDMI module in to DJI for repair since I like sharing the video downlink with others. When it came back it was doing the same thing. So I sent it back again. This time they sent me a different (appears new) TX w/HDMI module which I received last night. After re-binding the TX to my P3P, I tested the HDMI out functionality using my TV, which worked fine. I left it running for about 10 minutes and there were no apparent video issues. At this time I should have probably tested it a few more times but the TX firmware was version 1.6.0 and the app was telling me an update is needed. I updated the TX to 1.8.0 and BAMMO, the severe corruption returned after a few minutes. Luckily I found out how to downgrade the TX firmware and I downgraded it back to 1.5.8, which I tested three times for >10 minutes each time and didn't see any video issues. I now flashed it back to 1.6.0 and am testing that but so far so good.

My new theory is that when the DJI engineers made tweaks to version 1.8.0 TX firmware, they did so only testing it with the stock output module (or they didn't test it long enough on a HDMI equipped TX), and the code does not play nicely with the HDMI module. In other words, this was not an HDMI output module hardware failure like I originally suspected.

For the few of you that have this configuration I setup a poll to try and validate my theory.

PS: If you want to see what this video corruption looks like I have some videos here where I swap between the HDMI and stock output module:
Box | Simple Online Collaboration: Online File Storage, FTP Replacement, Team Workspaces

 

[Quest-So]

I've installed 2 days ago on my RC such a board (preparing for goggles). As you can see in my signature the RC runs with v1.8.0 and I see no issues with the video stream. By contrary, I'm positively surprised that the HDMI is outputing the video even if the DJI GO app is not running. Probably in some cases I will fly without using the tablet, just goggles attached to this output.

 

[Waynosan]

I've installed 2 days ago on my RC such a board (preparing for goggles). As you can see in my signature the RC runs with v1.8.0 and I see no issues with the video stream. By contrary, I'm positively surprised that the HDMI is outputing the video even if the DJI GO app is not running. Probably in some cases I will fly without using the tablet, just goggles attached to this output.

Appreciate you replying, thank you. I wonder if the firmware on the quad has something to do with it then, becuase I believe I'm running a later version than what you specified in your signature. The other difference is that your Phantom is advanced, and I have the pro. This problem is being really difficult to diagnose.

If you get the chance, could you do the following:

1. Connect your TX to your tablet or phone as usual
2. Place the phantom about 1 meter away from the TX
3. Start everything up and set a timer for 10 minutes
4. After 10 minutes, come back and a) wave your hands in front of the camera and b) move the phantom around so the camera looks at a couple of different scenes
5. Let me know if video looked OK

Three out of three times when I did this yesterday I saw severe video corruption when I had TX firmware 1.8.0 installed. Corruption happened zero out of three times on the 1.6.0 and 1.5.8 firmware versions. It also did not happen with TX firmware 1.8.0 when I had the stock output module installed (this was on my previous two TXs, I have not tried this on the one I just got back.)

If nothing else, it will give me some piece-of-mind that you will not have this happen to you when you're actually flying your phantom.

Thanks again for the help and if anyone else chimes in on the poll please let me know in the thread if you have a P3A or P3P.

Have a good one!

 

[Quest-So]

Here we are. I've made the requested test. No issues at all.
Conditions were as follows:
DJI Go was updated to the latest v3.1.2.
Drone was placed out on the balcony (outer temp -1°C / 30°F), RC was in the house 7-8 m from the aircraft (2 walls btw them).
A small 16" Tv set, placed nearby the RC & tablet, has shown the HDMI output.

Hand waving and changing the scene (rotating the drone left right +/- 45° were done 3 times after 10, 30, 50 min from the drone start (motors steady off). All ok.
Then, with the drone still on, I've moved it into the house/living room and placed it for 10 min in the front of a big TV set (I've chosen a dynamic programme with a lot of scene changes). All ok, no macro blocks on rapid scene changes, no nothing. All ok.

What sw versions you're using on the AC ? Could you downgrade your P3P back to v1.8.80 (which in my opinion is very stable, I'm running it since oct 2016), or at least v1.9.60 ? (I've noticed other pilots are complaining about the latest v1.10.90).

 

[Waynosan]

Here we are. I've made the requested test. No issues at all.
....

What sw versions you're using on the AC ? Could you downgrade your P3P back to v1.8.80 (which in my opinion is very stable, I'm running it since oct 2016), or at least v1.9.60 ? (I've noticed other pilots are complaining about the latest v1.10.90).

Wow, you really committed yourself to the task! Glad your is running well; definitely not having the problem I'm having or you would have seen it for sure.

I'm running FW 1.10.90 on the P3P. I just took a look at their changelog and 1.8.80 seems like would be a good version to drop down to if I have to go that route. I'll try flying this weekend to see if there are issues in flight. If the 1.6.0 on the TX works OK then I think I'd rather stick with this combination. There does not seem to be a changelog for the TX FW aside for what changed in the latest version.

Thanks for your help, I figure whatever info I can get I'll forward to DJI to help them sort it out.

 

[Hadyan Dharmawan]

"..........I now flashed it back to 1.6.0 and am testing that but so far so good."

Firmware 1.8.8 the most stable until now....I think

I'm using HDMI Module on my P3A TX and it work with no issue till now, P3A (fw 1.8.8), TX GL300C (fw 1.6.0), DJI GO 3.0.0 Android

 

[Varming]

And it isn't possible to downgrade the AC from 1.10.90

Sent from my SM-G935F using PhantomPilots mobile app

 

[Quest-So]

And it isn't possible to downgrade the AC from 1.10.90

Sent from my SM-G935F using PhantomPilots mobile app

True.
But, I've read this morning a thread someone says it's possible to install an older firmware but previously you have to erase the SD card that's inside the AC's body, underneath the ESC board.

 

[Waynosan]

Well darn,
It being the weekend and the weather not too horrid I was able to try flying the P3P in my backyard yesterday. I should have performed a range-test first because after the AC got about 20m away, the "low transmitter signal" warning started to flash. By the time it got to the back of the yard (maybe 50m?) it lost signal completely and RTH failsafe was initiated. I regained control when it was about 10m or so above me and had enough time to guide it back down onto the deck. At first I thought it may have been a FW incompatibility issue so I flashed everything to the latest but the range remained the same (all tests at this point were done on the ground with the props off). I noticed that when I put my hand over the left TX antenna the signal would drop out completely. When I did the same with the right there was no change in TX signal strength. I think someone forgot to attach the right antenna inside the TX. I wonder if there are actually two TX circuits in there or just one as running a TX without an antenna connected usually results in a fried TX. I sent an email to DJI asking if I can open the TX to check this w/o voiding the warranty. Waiting for a reply. If I have to send it in then this will by the 3rd time. Urgh.

 

[Quest-So]

Waynosan, why do you mention just 2 antennas ? There are 4, one in each leg. Test how your hand influences the signal strength by touching in sequence all 4 legs. The back-left leg contains too the compass.

 

[Waynosan]

Waynosan, why do you mention just 2 antennas ? There are 4, one in each leg. Test how your hand influences the signal strength by touching in sequence all 4 legs. The back-left leg contains too the compass.

I meant the antennas on the transmitter (TX), not the aircraft.

 

[Waynosan]

Finally got another TX back from DJI. This one does not have range issues so the problem was definitely with the last one that they sent me. For S&G I tried it with the v1.8.0 firmware but the video artifact started acting up just like on the other ones. going back to v1.6.0 cleared it up like it did before also. Hopefully they'll fix this in the future or never prevent me from rolling back to v.1.6.0.

 

[Quest-So]

You haven't wrote yet what RC model you had (GL300 A, B, C ?)
What was the 1st one you initially had, then the 2nd one DJI has sent you back, then the 3rd present one.

 

[Waynosan]

You haven't wrote yet what RC model you had (GL300 A, B, C ?)
What was the 1st one you initially had, then the 2nd one DJI has sent you back, then the 3rd present one.

Initial one was GL300B, the next was C and C presently. Someone on RC groups P3 owners thread said it's a known issue and DJI is working on it. Said to stay on 1.6.0 for the time being so I have renewed faith that they'll fix it, even though it seems like I'm the only one having this particular problem on this site (based on my poll).