80 to 90% battery level takeoff is triggering outrage

[Tommymonkey]

Guys, this post is very interesting DJI Forum|Battery Critically Low Voltage Warning

 

[Oso]

All you guys should listen to Ianwood, his advice will solve your problems and bring back your confidence in your Phantom 3 over time.

I just got back from CES where I had two separate conversations with knowledgeable DJI employees, one from each of their two booths at the show. Each of their booths had one technical person there, and I made it a point to seek out that person. I asked for the most knowledgeable person to discuss Phantom 3 firmware. Below is what I learned, and this info was echoed by both employees in these separate conversations.

Regarding reports of sudden power loss during flight, battery shut downs are only a problem with the 1.5 firmware, period. 1.4 did not have the problem, nor any firmware before 1.4. It's important to know 1.6 solved this critical problem. That's why 1.5 was replaced so quickly with 1.6. So if you're using 1.5, DON'T FLY again until you either go back to 1.4, or move up to 1.6. After these conversations I deem 1.5 as the kiss of death, so just don't fly it!

The problem is solely in the 1.5 Phantom firmware, NOT the battery firmware. The 1.5 Phantom firmware was misreading the voltage levels from the battery, simply as that. However the display was showing the right voltage and % level. This misreading of the battery below 3.0V by the firmware is what triggered the battery protection which shut down the motors. Again, this was not a problem in the battery firmware, it was solely a Phantom firmware problem. If you recall there was no battery update in 1.5, however the shutdown problems started with 1.5, which supports the the notion that the Phantom 1.5 firmware is what started this issue. In 1.5, although you may see 60% battery level in your display, and you really do have 60% battery (a true reading), the firmware does not see that level, it mistakenly sees a lower voltage. The firmware is flawed, way off. It misinterpreted battery level, lower than reality, hence battery protection kicks in and turns off the power when your battery gauge says it's at 40 to 60%, then falling from the sky. Not good.

Normally when you land at 20% your batteries are not close to 3.0 V, I think the batteries read somewhere around 3.3V - 3.4V if I recall right. In 1.5 the firmware, when you see a reasonable battery level (I'll use 50 or 60% as an example), the firm sees 3.0V, which isn't reality, a bogus reading. Yes, this is a major error in battery level detection in 1.5 FW, one that will cost DJI a fair amount in warranty claims, I would assume. In this particular case, I can see why DJI would be motivated to MANDATE getting off 1.5, given the catastrophic consequences being reported. Anyone flying 1.5 is on thin ice, IMHO.

So what's the battery firmware update for in 1.6? This new battery firmware is where the acceleration decrease and vertical speed decreases comes from. They put a governor to limit maximum current draw. Sudden high current drain can fluctuate voltage on a low battery which might contribute to a problem if someone ran the battery down to 10% and then gave it full throttle, instead of babying the throttle. 1.6 firmware should help with flight times for anyone that has a habit of full throttle flying. It's my opinion that DJI didn't have to put a governor in the battery since they fixed the root problem in the Phantom 1.5 firmware, by releasing 1.6. Given what I learned, I believe the 1.4 battery firmware should work fine with 1.6 Phantom firmware, and you'll have your power back, and the drifting will be fixed too from the new 1.6 Phantom firmware (which was common complaint in 1.4 Phantom firmware). I particularly don't like losing the acceleration and vertical speed I use to have with 1.4 (I'm spoiled), so I'm going to do an experiment using 1.4 battery firmware with my 1.6 Phantom firmware and report back, to prove this point. I happen to have one battery with 1.4 firmware that I'm going to use for a quick test. All my other 4 batteries have 1.6 firmware.

The Phantom 1.6 firmware definitely fixes the battery shutdown issues. Nobody has has reported a problem to DJI with 1.6 , specifically with battery shutdown being an issue. With 1.6 you can fly partial discharge batteries like you use to do with 1.4 with confidence, just land before 20% for good margin and you'll be fine.

The auto-discharging of the battery also has nothing to do with this battery shutdown issue. If your battery goes into auto-discharge and you power up the craft with it, the discharging automatically stops. The Phantom firmware knows nothing about the discharge routine, it only reads the voltage at power up (correctly now with 1.6) and reports this to your screen. With 1.6, what the Phantom firmware sees as the voltage is now the same as what you see on the screen as the voltage, which is also interpolated into percentage left on the charge.

Cold weather has nothing to do with the battery shut down problem, however cold batteries will reduce your flight time, which is logical because Li-Po's are known to perform worse when cold, typically 10-15% less flight time. It is a good idea to keep the batteries warm before launch to maximize flight times, however a cold battery isn't a factor in battery shutdown, the 1.5 firmware was the sole cause, simply from misinterpreting the "true voltage" of the battery.

On a related subject...... regarding going back to 1.4 from 1.6, one for these tech guys (on this subject I didn't verify with the other guy) said you can go back to 1.4, however you have to go back to 1.5 first. Then after you have 1.5 loaded successfully you can go back to 1.4. He said this would work. That does not mean you can go back to 1.32 after you get to 1.4. Then he said try it if you feel compelled, he wasn't 100% sure about 1.32, he said it might work, but was pretty sure it won't.

FYI, going back to 1.32 I think you lose way-points, POI, and Follow me, etc. Those important features (Important to some) came in 1.4, if I remember correctly. For most people I don't think you'll want to lose these features. I realize 1.4 had the drifting issue, but I was able to tolerate that, however I don't get much wind here in So. Calif. That "drift fix" could be more important to others, and that's fixed in 1.6.

DON'T USE 1.5 !

This is a terrific overview of how the Phantom firmware was misreading the voltage levels from the battery. Thanks sir.

 

[John N Kronyak]

Thank you Old Timer. If you saw dinosaurs flyin' you got a bit on me. LOL I'm 59. We have seen more tech change in our time to know how things really work. our generation is the one that brought these high tech toys to fruition. Too much guessing and not enough research on these posts. And your right. If folks have nothing better to do than to try to make others feel stupid, they should find another venue to ventilate. These are high tech toys that still need a bit of educating to control and bugs in software to sort out.

 

[Phill2]

I'm a noob to drones but found that my battery warning comes on at 30% and it auto lands at 20% but I read of others running their batteries down to 7 or 8%.
10% doesn't seem much to get the bird back to me if a long way off.
Is there a way to change this?

 

[John N Kronyak]

Circles are no fun.

Please tell us some more about this. How did you prove it was in auto discharge at the time?

Save your breath. these folks just don't get it.

 

[Oso]

Guys, this post is very interesting DJI Forum|Battery Critically Low Voltage Warning

Thanks for posting this link. The video was indeed very interesting. Also, I totally want that guys set up.

 

[skyeboysteve]

I'm a noob to drones but found that my battery warning comes on at 30% and it auto lands at 20% but I read of others running their batteries down to 7 or 8%.
10% doesn't seem much to get the bird back to me if a long way off.
Is there a way to change this?

The 30% battery warning (set by default in the go app) is an entirely different animal to the critically low voltage warning.

Anyway, the point seems to be the reason for having a fully charged battery is so that the cold voltage droop won't trigger the critically low voltage alarm auto land, before battery has had a chance to warm up above 20C, when the droop is less as the battery drains.

 

[John Locke]

g`day all ,,,, sorry to jump in but does this happen to just the P3 as i have a P2v+and have never had any dramas at all . have logged more than 450 flights some with very low take off voltage of 40%
I am thinking of upgrading to the P3 but not real happy about all the falling out the sky type posts ?

ps after 450 flights im still on my first set of props ,,, never crashed

am i doing something wrong ?

This battery shutdown flaw is only in the P3A and P3P that's running 1.5 firmware. All other firmware doesn't have the power shutdown flaw. Just avoid 1.5 like the plague.

 

[John Locke]

FWIW, and perhaps it was just luck on my part, I had my P3 start auto-landing while on v1.5--twice (coincidentally I flew my most scenic flights over rough terrain on this firmware version--so I'm feeling lucky I still have it). In both cases I pressed the return home button on the RC and it cancelled the auto land. I immediately flew it home and landed without further incident. I believe the low voltage warning was triggered by my high acceleration inputs and perhaps I was able to cancel the auto land because the low voltage warning was triggered momentarily and the voltage "recovered" from lower power settings during auto land.

Anyway, just something to consider should your P3 enter auto land--you may be able to cancel it. Then again simply updating your firmware should save you from this specific problem anyway.

If you're still on 1.5 you need to get off that asap, seriously. There's a huge flaw in 1.5 reading the wrong voltage of the battery, lower than reality, which triggers auto land. In severe voltage mis-readings the power will be shut down mid-flight. This situation can be aggravated with hard accelerations, which pulls the voltage down (just like you explained).

 

[skyeboysteve]

If you're still on 1.5 you need to get off that asap, seriously. There's a huge flaw in 1.5 reading the wrong voltage of the battery, lower than reality, which triggers auto land. In severe voltage mis-readings the power will be shut down mid-flight. This situation can be aggravated with hard accelerations, which pulls the voltage down (just like you explained).

Yes for sure--get off 1.5. But note that the test that Tommy linked to was done with v1.6 where he demonstrated how high loads on "cold" batteries will still give you critically low voltage auto land if the battery was partially charged and not much warmer than 20C.

 

[Airpixtoledo]

This doesn't happen to most people.

I was one of those people. My battery was at 55%, went up to 200 feet, Pantom shut down, dropped like a rock. I recovered it. Sent it to DJI for repair, got it back in two weeks fixed and with a new battery. Sold it the next day.

 

[GadgetGuy]

I was one of those people. My battery was at 55%, went up to 200 feet, Pantom shut down, dropped like a rock. I recovered it. Sent it to DJI for repair, got it back in two weeks fixed and with a new battery. Sold it the next day.

I am one of those people as well. After 190 successful flights on my P3P, one of my batteries displayed 85% remaining, with a lowest cell voltage of 4.05V before launch. I flew it straight up to 240 feet, and the P3P dropped like a rock and crashed onto concrete 15 seconds into the flight, due to an in flight battery shut off, with a "Low Voltage Protection" displayed as Record 01 in the Battery History afterwards. Aircraft Firmware version 1.4.1. After the the crash, the battery still read 82% remaining with a lowest cell voltage of 4.07V and 100% battery life remaining. Battery had not been topped off before flying, and had been left idle for 2 weeks with a 10 day time to discharge, but I only planned a 2 minute test flight, so with 87% and 4 green lights, I figured I was fine. Wrong! Intro screen on newest P3P says to fly with minimum 50% battery and App displayed 87%. Analysis of the iPad Air2 flight log .txt file reveals that the takeoff battery temperature was 70° F and confirms the same battery percentage readouts of the DJI GO app. However, the individual battery cell readings are shown as significantly less, and Cell 4 reached 3.02V just before the shutoff.

Clearly, the failure to top off the battery, especially after it had begun discharging, and the reliance upon the displayed Percentage remaining battery in the app and the app displayed cell voltages, with the battery in that condition, were the cause of the crash. My crash reveals that these crashes can even happen at 70° F (20° C is 68° F), and with an app readout of 87% remaining battery, and with firmware prior to 1.5. If you have multiple batteries, there is no way from within the DJI GO app to tell which batteries have entered the discharge condition, if they are all still displaying 4 green lights. The ONLY way to completely avoid these in flight shutoffs is to top off every battery before use. Multiple flights on a freshly topped off battery are probably OK, as the topping off rebalances all the cells and makes sure the readouts in the app are accurate. I can't speak to 1.6 being a panacea for these problems, but the in flight battery shut offs are not exclusively limited to 1.5, and the discharging condition of the batteries is clearly a factor, as this is the only time in over 200 flights where any battery was also in a discharge condition and not topped off before flight. Safe flying!

 

[GadgetGuy]

All you guys should listen to Ianwood, his advice will solve your problems and bring back your confidence in your Phantom 3 over time.

I just got back from CES where I had two separate conversations with knowledgeable DJI employees, one from each of their two booths at the show. Each of their booths had one technical person there, and I made it a point to seek out that person. I asked for the most knowledgeable person to discuss Phantom 3 firmware. Below is what I learned, and this info was echoed by both employees in these separate conversations.

Regarding reports of sudden power loss during flight, battery shut downs are only a problem with the 1.5 firmware, period. 1.4 did not have the problem, nor any firmware before 1.4. It's important to know 1.6 solved this critical problem. That's why 1.5 was replaced so quickly with 1.6. So if you're using 1.5, DON'T FLY again until you either go back to 1.4, or move up to 1.6. After these conversations I deem 1.5 as the kiss of death, so just don't fly it!

The problem is solely in the 1.5 Phantom firmware, NOT the battery firmware. The 1.5 Phantom firmware was misreading the voltage levels from the battery, simply as that. However the display was showing the right voltage and % level. This misreading of the battery below 3.0V by the firmware is what triggered the battery protection which shut down the motors. Again, this was not a problem in the battery firmware, it was solely a Phantom firmware problem. If you recall there was no battery update in 1.5, however the shutdown problems started with 1.5, which supports the the notion that the Phantom 1.5 firmware is what started this issue. In 1.5, although you may see 60% battery level in your display, and you really do have 60% battery (a true reading), the firmware does not see that level, it mistakenly sees a lower voltage. The firmware is flawed, way off. It misinterpreted battery level, lower than reality, hence battery protection kicks in and turns off the power when your battery gauge says it's at 40 to 60%, then falling from the sky. Not good.

Normally when you land at 20% your batteries are not close to 3.0 V, I think the batteries read somewhere around 3.3V - 3.4V if I recall right. In 1.5 the firmware, when you see a reasonable battery level (I'll use 50 or 60% as an example), the firm sees 3.0V, which isn't reality, a bogus reading. Yes, this is a major error in battery level detection in 1.5 FW, one that will cost DJI a fair amount in warranty claims, I would assume. In this particular case, I can see why DJI would be motivated to MANDATE getting off 1.5, given the catastrophic consequences being reported. Anyone flying 1.5 is on thin ice, IMHO.

So what's the battery firmware update for in 1.6? This new battery firmware is where the acceleration decrease and vertical speed decreases comes from. They put a governor to limit maximum current draw. Sudden high current drain can fluctuate voltage on a low battery which might contribute to a problem if someone ran the battery down to 10% and then gave it full throttle, instead of babying the throttle. 1.6 firmware should help with flight times for anyone that has a habit of full throttle flying. It's my opinion that DJI didn't have to put a governor in the battery since they fixed the root problem in the Phantom 1.5 firmware, by releasing 1.6. Given what I learned, I believe the 1.4 battery firmware should work fine with 1.6 Phantom firmware, and you'll have your power back, and the drifting will be fixed too from the new 1.6 Phantom firmware (which was common complaint in 1.4 Phantom firmware). I particularly don't like losing the acceleration and vertical speed I use to have with 1.4 (I'm spoiled), so I'm going to do an experiment using 1.4 battery firmware with my 1.6 Phantom firmware and report back, to prove this point. I happen to have one battery with 1.4 firmware that I'm going to use for a quick test. All my other 4 batteries have 1.6 firmware.

The Phantom 1.6 firmware definitely fixes the battery shutdown issues. Nobody has has reported a problem to DJI with 1.6 , specifically with battery shutdown being an issue. With 1.6 you can fly partial discharge batteries like you use to do with 1.4 with confidence, just land before 20% for good margin and you'll be fine.

The auto-discharging of the battery also has nothing to do with this battery shutdown issue. If your battery goes into auto-discharge and you power up the craft with it, the discharging automatically stops. The Phantom firmware knows nothing about the discharge routine, it only reads the voltage at power up (correctly now with 1.6) and reports this to your screen. With 1.6, what the Phantom firmware sees as the voltage is now the same as what you see on the screen as the voltage, which is also interpolated into percentage left on the charge.

Cold weather has nothing to do with the battery shut down problem, however cold batteries will reduce your flight time, which is logical because Li-Po's are known to perform worse when cold, typically 10-15% less flight time. It is a good idea to keep the batteries warm before launch to maximize flight times, however a cold battery isn't a factor in battery shutdown, the 1.5 firmware was the sole cause, simply from misinterpreting the "true voltage" of the battery.

On a related subject...... regarding going back to 1.4 from 1.6, one for these tech guys (on this subject I didn't verify with the other guy) said you can go back to 1.4, however you have to go back to 1.5 first. Then after you have 1.5 loaded successfully you can go back to 1.4. He said this would work. That does not mean you can go back to 1.32 after you get to 1.4. Then he said try it if you feel compelled, he wasn't 100% sure about 1.32, he said it might work, but was pretty sure it won't.

FYI, going back to 1.32 I think you lose way-points, POI, and Follow me, etc. Those important features (Important to some) came in 1.4, if I remember correctly. For most people I don't think you'll want to lose these features. I realize 1.4 had the drifting issue, but I was able to tolerate that, however I don't get much wind here in So. Calif. That "drift fix" could be more important to others, and that's fixed in 1.6.

DON'T USE 1.5 !

My in flight battery shutoff and ensuing P3P crash occurred on Aircraft Firmware 1.4.1 only after I failed to top off a discharging battery that displayed 87% and 4.07 V in DJI GO upon take off. That's real raw data that contradicts several of your assertions. See my post above.

 

[GadgetGuy]

This battery shutdown flaw is only in the P3A and P3P that's running 1.5 firmware. All other firmware doesn't have the power shutdown flaw. Just avoid 1.5 like the plague.

I experienced it with 1.4.1, so it is not exclusive to 1.5. See above.

 

[GadgetGuy]

If you're still on 1.5 you need to get off that asap, seriously. There's a huge flaw in 1.5 reading the wrong voltage of the battery, lower than reality, which triggers auto land. In severe voltage mis-readings the power will be shut down mid-flight. This situation can be aggravated with hard accelerations, which pulls the voltage down (just like you explained).

It can also happen on 1.4.1 under the triggering conditions. See above.

 

[Phill2]

The 30% battery warning (set by default in the go app) is an entirely different animal to the critically low voltage warning.

Anyway, the point seems to be the reason for having a fully charged battery is so that the cold voltage droop won't trigger the critically low voltage alarm auto land, before battery has had a chance to warm up above 20C, when the droop is less as the battery drains.

Ambient temp here is never below 20 C

 

[Mark The Droner]

So both the drops in the sky above were with 1.6?

Were the batteries in an auto-discharge state prior to turning them on?

 

[GadgetGuy]

So both the drops in the sky above were with 1.6?

Were the batteries in an auto-discharge state prior to turning them on?

If you are referring to my crash, described above, as one of the "drops in the sky above", mine was not on 1.6 but on 1.4.1, but the battery was past the 10 day auto-discharge, and was certainly at some level of auto-discharge prior to being used in the aircraft, which I believe was the real problem, which is cured by topping off all stored batteries immediately before use.

 

[Mark The Droner]

That's what I figured (the auto-discharge). There are just so many consistencies relating drops with auto-discharge. Maybe 1.6 solves the problem, but I don't think it's correct to conclude the problem has nothing to do with the auto-discharge.

 

[GimbleOut]

Seems part of the issue beyond firmware errors is the misunderstanding of the % charge reading. The % charge is interpolated from voltage readings. However, voltage readings are just a proxy for the energy actually in the battery. You will only really know that by readings of voltage under load. That's why checking individual cells under load matters.

Seems to me a battery sitting for some time could indicate adequate voltage and appear ready to fly but in reality not enough energy to supply the loads of flying. Problem

However a battery discharged under loads and carefully watched will have a more accurate % charged reading and can be used.

It's been said.

Charge before flying
Check cells under load
Never FW 1.5

Multiple flights on batteries flown same day are fine...but keep checking the cells

Happy Flying


Sent from my iPhone using PhantomPilots mobile app