Posts: 21,420 Threads: 597 Likes Received: 9,040 in 6,691 posts Likes Given: 1,428 Joined: Jun 2018 Reputation: 797 Thanks for your efforts in trying to get to the bottom of this issue.
It's interesting that the slowest of the 3 cards resolved the issue and the two faster ones caused problems. I certainly wouldn't have expected that. Have you tried any independent data read/write speed tests on those 3 cards?
I would also be interested to know if your spare air unit exhibits exactly the same behaviour with the same 3 cards. That would then prove if the issue is definitely is down to the cards, or if it's your original air unit that has some kind of fault. ![[-]](https://intofpv.com/images/collapse.png) • Posts: 465 Threads: 51 Likes Received: 187 in 119 posts Likes Given: 88 Joined: Aug 2018 Reputation: 6 (06-Oct-2020, 01:28 AM)SnowLeopardFPV Wrote: Thanks for your efforts in trying to get to the bottom of this issue.
It's interesting that the slowest of the 3 cards resolved the issue and the two faster ones caused problems. I certainly wouldn't have expected that. Have you tried any independent data read/write speed tests on those 3 cards?
I would also be interested to know if your spare air unit exhibits exactly the same behaviour with the same 3 cards. That would then prove if the issue is definitely is down to the cards, or if it's your original air unit that has some kind of fault. I was totally surprised, as well. And the jury is still out, having flown only 6 LiPos successfully so far. But I've never been able to fly even 2 LiPos without seeing the problem in the past. If the current SD card continues to work well, I won't be very motivated to swap in the other air unit. This quad has been nothing but trouble (it still doesn't fly very well), and I'm fed up with debugging it. If I eventually build the second air unit into a new quad, I'll check it with the various SD cards at that time. • Posts: 304 Threads: 45 Likes Received: 55 in 46 posts Likes Given: 10 Joined: Aug 2020 Reputation: 2 That is very strange. The card with the slower speeds worked while the faster ones didn't. • Posts: 465 Threads: 51 Likes Received: 187 in 119 posts Likes Given: 88 Joined: Aug 2018 Reputation: 6 (06-Oct-2020, 07:18 AM)AlanB1976 Wrote: That is very strange. The card with the slower speeds worked while the faster ones didn't. It is strange, but there could be an explanation. I found this in the Betaflight documentation for Blackbox: Quote:Your choice of microSDHC card is very important to the performance of the system. The OpenLog relies on being able to make many small writes to the card with minimal delay, which not every card is good at. A faster SD-card speed rating is not a guarantee of better performance.
microSDHC cards known to have poor performance
- Generic 4GB Class 4 microSDHC card - the rate of missing frames is about 1%, and is concentrated around the most interesting parts of the log!
- Sandisk Ultra 32GB (unlike the smaller 16GB version, this version has poor write latency)
microSDHC cards known to have good performance
- Transcend 16GB Class 10 UHS-I microSDHC (typical error rate < 0.1%)
- Sandisk Extreme 16GB Class 10 UHS-I microSDHC (typical error rate < 0.1%)
- Sandisk Ultra 16GB (it performs only half as well as the Extreme in theory, but still very good)
• Posts: 304 Threads: 45 Likes Received: 55 in 46 posts Likes Given: 10 Joined: Aug 2020 Reputation: 2 Hmmm.... Makes sense. Although I would have thought the video recording would have fallen in to the bucket of needing the constant high write speeds and not the small writes.
I've receive my GPS now so once I get time to wire it in, will let you know how I get on. Just need a break from kids and rain now • Posts: 2 Threads: 1 Likes Received: 0 in 0 posts Likes Given: 0 Joined: Oct 2020 Reputation: 0 (06-Oct-2020, 04:26 PM)SJChannel Wrote: It is strange, but there could be an explanation. I found this in the Betaflight documentation for Blackbox: Hello, I have just registered because I have some issues with OSD but in my case it has dissapeared. Could you check if you have continuity with ground in any of the TX/RX pads of the Air Unit? I read it in oscarliang blog related to OSD issues in Vista units and just connecting the DJI cable to one of my Air Units unit, I have confirmed that the internal connection RX (white wire) is connected to internal GND (Brown wire connection), so problably internally it has a diode or something broken. I have other unit not showing the continuity, as I far as a I know from other bus protocol, HI and LO wires must be ground isolated. I think you could also check it. Kind regards, Alfredo • Posts: 465 Threads: 51 Likes Received: 187 in 119 posts Likes Given: 88 Joined: Aug 2018 Reputation: 6 (13-Oct-2020, 05:35 PM)Alfredrone Wrote: Hello,
I have just registered because I have some issues with OSD but in my case it has dissapeared. Could you check if you have continuity with ground in any of the TX/RX pads of the Air Unit? I read it in oscarliang blog related to OSD issues in Vista units and just connecting the DJI cable to one of my Air Units unit, I have confirmed that the internal connection RX (white wire) is connected to internal GND (Brown wire connection), so problably internally it has a diode or something broken. I have other unit not showing the continuity, as I far as a I know from other bus protocol, HI and LO wires must be ground isolated. I think you could also check it.
Kind regards,
Alfredo I checked it on my air unit, and I didn't see anything to worry about. With my multimeter set on the "diode test" function, it showed about 1 diode drop (0.6 to 0.7 volts) from some of the signals to ground, and a higher amount when I reversed the test probes. There were not any direct shorts from the signals to ground. • Posts: 304 Threads: 45 Likes Received: 55 in 46 posts Likes Given: 10 Joined: Aug 2020 Reputation: 2 My GPS is up and running. I haven't noticed any freezing as yet. I'm using the Sandisk Extreme 32GB.
The GPS osd I have is speed/distance/alt/arrow • Posts: 465 Threads: 51 Likes Received: 187 in 119 posts Likes Given: 88 Joined: Aug 2018 Reputation: 6 (17-Oct-2020, 08:52 AM)AlanB1976 Wrote: My GPS is up and running. I haven't noticed any freezing as yet. I'm using the Sandisk Extreme 32GB.
The GPS osd I have is speed/distance/alt/arrow Alan, are you using the DJI remote, or do you have a separate receiver in your quad? I did some more flights with mine, and it is still freezing up sometimes. Next I am going to try installing a Crossfire receiver and using my Taranis radio. I like it better than the DJI remote anyway. And at least it will warn me when my battery voltage is getting low. • Posts: 304 Threads: 45 Likes Received: 55 in 46 posts Likes Given: 10 Joined: Aug 2020 Reputation: 2 I'm using the Radiomaster tx16s and crossfire receiver • Posts: 465 Threads: 51 Likes Received: 187 in 119 posts Likes Given: 88 Joined: Aug 2018 Reputation: 6 I have an update on this problem. I installed a Crossfire receiver in the quad, so the air unit is no longer handling the RC part of the system. It's great to be using my Taranis again, which I can count on to tell me when the battery is getting low. I flew 8 packs a couple of days ago and didn't see any OSD stuttering, but the goggles kept alerting me that the SD card in the air unit was too slow. So I guess using the SanDisk Ultra card isn't the fix, after all. I'm back to using the recommended SanDisk Extreme again. Next time I fly the quad, I'll find out if the stuttering is still gone. • Posts: 18 Threads: 0 Likes Received: 2 in 2 posts Likes Given: 0 Joined: Oct 2020 Reputation: 0 I have 5 Air units and 1 Vista. 4 Air units use BEC's as part of the FC board from brands GepRC and iFlight. 1 Air unit runs direct from 4s. The Vista runs direct from 3s. All of them experience an issue where the mAh counter will freeze and then unfreeze itself up to 30 seconds later averaging 10 seconds. Sometimes it will freeze till power cycle. The issue was common on all firmwares since the OSD was added but is now infrequent since the latest v01.00.06.00 firmware. There seems to be no common trigger it just occurs. A look in the blackbox confirms the mAh counter was running correctly the whole time so it's something to do with how it communicates. • Posts: 465 Threads: 51 Likes Received: 187 in 119 posts Likes Given: 88 Joined: Aug 2018 Reputation: 6 (27-Jan-2021, 01:17 AM)ChangeAgent Wrote: I have 5 Air units and 1 Vista. 4 Air units use BEC's as part of the FC board from brands GepRC and iFlight. 1 Air unit runs direct from 4s. The Vista runs direct from 3s. All of them experience an issue where the mAh counter will freeze and then unfreeze itself up to 30 seconds later averaging 10 seconds. Sometimes it will freeze till power cycle. The issue was common on all firmwares since the OSD was added but is now infrequent since the latest v01.00.06.00 firmware. There seems to be no common trigger it just occurs. A look in the blackbox confirms the mAh counter was running correctly the whole time so it's something to do with how it communicates. That matches my experience with my first Air Unit. My second one hasn't exhibited the freezes so far, but I haven't flown it a lot yet. Both Air Units too often stop recording with the message "SD card slow", even though I'm using their exact recommended SD card and I've tried a bunch of different ones. DJI seems determined not to admit there's a problem. • Posts: 21,420 Threads: 597 Likes Received: 9,040 in 6,691 posts Likes Given: 1,428 Joined: Jun 2018 Reputation: 797 (27-Jan-2021, 02:33 AM)SJChannel Wrote: Both Air Units too often stop recording with the message "SD card slow", even though I'm using their exact recommended SD card and I've tried a bunch of different ones. DJI seems determined not to admit there's a problem. This has been a bug bear for a lot of people since the last few firmware updates. I have a feeling that DJI don't really care TBH. They've used the FPV community to prove that their FPV product works well enough so it could be incorporated into their own DJI FPV Drone, and my guess is that going forward they will now put most of their time and effort into improving that and just leave the standalone FPV products to stagnate with unfixed issues. I really hope I'm wrong but I guess we shall see in the coming months. • Posts: 1,530 Threads: 49 Likes Received: 721 in 532 posts Likes Given: 481 Joined: Oct 2020 Reputation: 93 Are they connected to dedicated UARTs or are you using softserial? Do any other OSD elements freeze, or just the mAh counter?
Some have reported that a capacitor across the power inlet leads to the Vista/AU prevent these random freezing issues as well. • |
