Near crash after Warning:Propulsions output is limited
- Thread starter EmptyHassan
- Start date
- Joined
- Oct 30, 2015
- Messages
- 325
- Reaction score
- 236
- Age
- 53
Battery or cell voltage is not linear during use!
It's important to keep that in mind because you can go bellow 3.6 or even lower and still have enough juice to get back home and land safely (considering it's in good shape of course). Just notice how voltage decreases fast during hard/fast moves, and quickly increases back when flying easy or slow. After landing it can go up half a volt or more. I've landed bellow 3.65V and it returned back to 3.78V a few seconds after landing.
Also, battery percentage is not a direct relation to total or cell voltage. It's not just a matter of how good (or bad) the algorithm used, there's just a lot of other factors involved. It's an estimate, that's why it's important to keep an eye in voltage and TIME.
We don't have advanced telemetry in most other RC models, we rely on TIMED FLIGHTS (or runs, etc.) and VOLTAGE taken right after motor kill to draw more accurate conclusions about general battery condition. My take on DJI and P3's battery management is they get it pretty OK overall, I now rely on voltage to time my flights and also speed, etc.
It's important to keep that in mind because you can go bellow 3.6 or even lower and still have enough juice to get back home and land safely (considering it's in good shape of course). Just notice how voltage decreases fast during hard/fast moves, and quickly increases back when flying easy or slow. After landing it can go up half a volt or more. I've landed bellow 3.65V and it returned back to 3.78V a few seconds after landing.
Also, battery percentage is not a direct relation to total or cell voltage. It's not just a matter of how good (or bad) the algorithm used, there's just a lot of other factors involved. It's an estimate, that's why it's important to keep an eye in voltage and TIME.
We don't have advanced telemetry in most other RC models, we rely on TIMED FLIGHTS (or runs, etc.) and VOLTAGE taken right after motor kill to draw more accurate conclusions about general battery condition. My take on DJI and P3's battery management is they get it pretty OK overall, I now rely on voltage to time my flights and also speed, etc.
- Joined
- Oct 30, 2015
- Messages
- 325
- Reaction score
- 236
- Age
- 53
Based on what I've seen personally with P3 batteries, I'd back that up. I've seen some pretty good batteries from friends get unbalanced cells after this procedure. Not a killed battery but minor deviations on otherwise level cells.
I've never performed "deep discharge" on any of my LiPOs, ever - certainly not the way DJI tells us to do it. I've cycled LiPOs but with the use of some really fancy and capable computerized chargers, which are also able to balance and measure IR during procedures.
I know DJI tells that's good and the right thing to do and all, but I for myself would rather take my chances by not following them on this specific if I'm dealing with a perfect, functional battery.
- Joined
- Oct 18, 2015
- Messages
- 1,617
- Reaction score
- 747
- Age
- 70
Empty - you were lucky, no way around that. In the future, aside from starting on a full battery, I'd avoid takeoff and landing around groups of people. The last thing you need is some 10 year old off his Adderall thinking it would be a good idea to check out the P3.
If the DJI recommendation for battery maintenance ruins a battery, I would deem that a sign of an unhealthy battery that should not be flown anyway. The battery basically gets put out of it's misery. I too won't do the 8% cycling if the battery doesn't show signs of deviation in healthy drones, but if I constantly see a major deviation I'll cycle the battery if it's due at 20 charges. That said, only once has it helped out of 2 batteries that I've cycled. The other battery wasn't damaged any more, but it didn't help either.
Last edited:
I agree, it's best to watch voltage, not percentage. Msinger's flight log view shows in spades the DJI "gas gauge" algorithm seems to be unreliable. At the same percentage of 30% or 50% the voltage can vary 1 whole volt (total voltage). Look at the numbers of my two flights, then look at the 40% comparisons to EmptyHassan's flight below. Considering the voltage under load ranges from 17 to 14 volts, 1 volt variance at the same percent level between two flight in the same model bird is a lot. I truly think the DJI gas gauge algorithm needs improvement.
Attachments
Last edited:
- Joined
- Mar 13, 2015
- Messages
- 1,064
- Reaction score
- 309
Every 20 cycles I go to 2% then to 100% no harm done so far.
Original poster looks like you should make sure you take off with 100% battery every time. Also, I'd practice a little more flying it. That was a pretty hectic return. I realize it was trying to auto land but even then you were wasting a lot of energy yawing back and forth and sitting idle several times.
Original poster looks like you should make sure you take off with 100% battery every time. Also, I'd practice a little more flying it. That was a pretty hectic return. I realize it was trying to auto land but even then you were wasting a lot of energy yawing back and forth and sitting idle several times.
Lesson learned
Does that mean you can't fly multiple flights on the same battery charge? Or is there a certain battery charge threshold that you shouldn't take off under?
After charging to 100% I took the P3 out for a flight today and it performed perfectly without any alerts or errors. Analyzing the log on the phantomhelp tool shows intermittent cell deviation of up to 0.145. In total, out of 16 minutes of flight time there was significant deviation in the cell voltage for maybe 7 seconds. Is this normal?
Does that mean you can't fly multiple flights on the same battery charge? Or is there a certain battery charge threshold that you shouldn't take off under?
After charging to 100% I took the P3 out for a flight today and it performed perfectly without any alerts or errors. Analyzing the log on the phantomhelp tool shows intermittent cell deviation of up to 0.145. In total, out of 16 minutes of flight time there was significant deviation in the cell voltage for maybe 7 seconds. Is this normal?
- Joined
- Mar 13, 2015
- Messages
- 1,064
- Reaction score
- 309
I wouldn't do multiple flights on the same battery especially if the batteries are stored cold beforehand but I wouldn't do it either way.
Go to Healthydrones and upload your flight logs from your tablet/phone. The site is free with a limit of 50 flight uploads. It has a power section that shows your battery deviations along with an explanation of what deviations are and what to look for.
SOC (state of charge) is notoriously difficult to calculate with lithium cells. A lot of factors come I to play here, the main one being that the power curve is it linear so not directly linked to the cells current voltage.
I agree, dji do need to work on their SOC algorithms! For now, I only monitor the cell voltages, the % means nothing to me.
- Joined
- Oct 30, 2015
- Messages
- 325
- Reaction score
- 236
- Age
- 53
In keeping with the idea of ease of use and maintenance for the majority of P3 pilots (i.e. first time flyers and non-hobbyists) - which is IMHO the ultimate and main proposal of DJI with the P3 and the reason of its success - the % gauge makes total sense, and is a useful tool. It's certainly not as precise as voltage measurement but it gives an instant, intuitive feedback on remaining battery capacity and thus flight time.
I'm not advocating for % reliance, but rather reference, which are distinct concepts. Don't push it, stay within limits, and you're fine. At least most of the time. IMHO percentage does't have to be as accurate as voltage. And unlike V, it's linear, after all the charge is being "consumed" (or drawn) when the P3 is flying, whereas voltage is affected by how the battery is being drawn at X moment.
After some time most everyone get used to the voltage numbers and how that relates to battery state at a certain point. You get to "know" your battery just like hobbyists do after a while with all the voltmeter measuring and all. But % gets you there and through without much thinkering, and that's good for novices.
I'm not advocating for % reliance, but rather reference, which are distinct concepts. Don't push it, stay within limits, and you're fine. At least most of the time. IMHO percentage does't have to be as accurate as voltage. And unlike V, it's linear, after all the charge is being "consumed" (or drawn) when the P3 is flying, whereas voltage is affected by how the battery is being drawn at X moment.
After some time most everyone get used to the voltage numbers and how that relates to battery state at a certain point. You get to "know" your battery just like hobbyists do after a while with all the voltmeter measuring and all. But % gets you there and through without much thinkering, and that's good for novices.
Wow! Scary landing in that condition in such a small area with so many people around. You got guts simply flying it in that situation!
Good you got it back!
Sent from my iPhone using PhantomPilots mobile app
Good you got it back!
Sent from my iPhone using PhantomPilots mobile app
I was wondering if anyone with enough knowledge could tell me what I did wrong to get the propulsion output limited error. I have been getting it on every flight. I've seen it as high as 49% battery level and at lower levels which is more understandable. Every time Ive gotten the message, I am moving full speed.
HealthyDrones.com - Innovative flight data analysis that matters
HealthyDrones.com - Innovative flight data analysis that matters
msinger
Approved Vendor
- Joined
- Oct 30, 2014
- Messages
- 31,962
- Reaction score
- 10,787
- Location
- Harrisburg, PA (US)
- Website
- phantomhelp.com
msinger
Approved Vendor
- Joined
- Oct 30, 2014
- Messages
- 31,962
- Reaction score
- 10,787
- Location
- Harrisburg, PA (US)
- Website
- phantomhelp.com
I think that messages comes up when you reach 3.6V on your lowest cell. If you haven't done this already, go into the battery page and check the box that displays the battery voltage right below the battery percentage in the upper right corner. I've notice several things happen at 3.6V (which is the level of your lowest cell of the 4 in the pack). First, the battery number font color goes from green to orange, that's a warning sign. You should head to home or be nearby. In the battery page you'll see the battery level animations go to orange also. At this point I've noticed I no longer have full power. This is when I stop using full power and mellow out to maximize my flight time. I start thinking about slowly heading back to home, grabbing any last minute photos of video that seem worthy.
Once the numbers go orange, my top vertical ascent speed goes from 11mph to about 7-8mph. The craft also doesn't accelerate horizontally as quick, however top speed is still over 30mph, but it takes longer to get there. This is the result of the 1.6 firmware that includes a power governor, which limits the ability to apply full power. On the last half of a battery charge, voltage fluctuations (during full throttle) are greater than with a fully charged battery. These high voltage fluctuations is what has been known to trigger battery protection to kick in, which starts with auto land, and in worse cases, mid-flight power shutdown. To mitigate the fluctuations DJI apparently put in these safeguards, limiting available power.
As others have been saying, it's much safer to monitor your battery voltage than the percentage. I believe the percentage calculation algorithm is flawed in the firmware. This is further complicated when a craft is started up with a partially depleted charge, and the firmware doesn't know how much of a charge is truly left in the battery, it's only guessing IMO. I think that's why we see such a huge deviation of percent vs. voltage levels from one flight to another, such as EmptyHassan's flight that started on a partial charge versus my flight that started on a full charge. Apples and oranges, over a full volt difference at 40%. I think that huge delta is caused by starting a flight on a partially charged battery, almost certain. Notice Hassan's total voltage is 13.481, which equates to 3.37V/cell. That's very low, too low, which is likely why auto-land was triggered. I always start heading back at 14.4V, or 3.6/cell.
Last edited:
