Since it is imperative when planning auto missions to know your flight times I have been testing with stabilize/loiter hovers and measuring air time while I familiarize myself with the Arducopter platform.
I now have 17 full flights with respect to battery failsafe at 10.5 volts. I bought 3 batteries with the IRIS.
On these batteries the best hover time I can get is 10.5 to 11 minutes.
Yesterday I bought a more expensive Thunder Power RC G6 Pro Performance 3850mah 45C battery and a HI-Tech X1 MF balance charger. On the first flight with that battery I got 13 minutes hover/loiter. I hope to try it again today. Maybe it will get better after a couple of cycles. I only charged it with 1.9 amps for the first charge.
BTW - The $74 I spent on the charger was worth it. It charges the IRIS 3500 packs in an hour at 2 amps(well below 1C). The stock charger was taking 2.5 hours. it also displays the individual cell voltages and has a storage charge feature.
All battery tests were done with no payload. No legs, no camera, just like it comes out of the box.
I am curious to know what others are getting. The IRIS was advertised 10 to 15 minutes flight time. Assuming that it means 15 minutes with no payload and 10 minutes with legs and gimbal I would obviously like to be at 15 minutes with no payload. Doesn't seem possible as delivered.
Good for you Justin. Very few are getting these results. I think there are batches of their motors that are drastically different than other motors. I am always using new props, pulling about 23 amps and get nowhere near what you are getting. If all of the Iris weights are equal then motor efficiency or ESC performance is the only other explanation.
Do your motors get really hot? Like you can't touch them for more than 1 or 2 seconds?
Nope, the motors are not noticeably hot. The batteries themselves get a little warm and lumpy immediately after flight, but not as bad as when I had the voltage threshold at default settings. I'm not sure why I got lucky here, but I'll take it. My only other observations are 1) my vibration log is crazy tight, 2) I never engage a GPS mode when testing, and 3) wind is <3mph when testing and I'm very light on flight adjustments. I don't see why any of these conditions would account for substantially better flight times, but there it is.
Could you upload a flash log? I should be able see the difference between our experiences.
Hover throttle out / in
The 4 RCOU chan values
My vibration readings are excellent as well.
Thanx for the info
If you do end up looking at the log, I'd appreciate an "official" flight time from you as I was using my stopwatch and I'm not yet proficient with log analysis. I'd assume you'd time throttle up to throttle down rather than arm to disarm, but I couldn't figure out how to read that.
There is a major difference in how much current Justin's Iris is consuming. His Iris is averaging about 20 amps at hover with a range of 18.8 to 21.
Mine draws 23 to 25 amps.
Hotel and Christian, do you see the same difference?
I would have to dig it out of the transport container in a while and pull logs,
I have grounded our exemplar due to safety concerns vs business insurance given all that has been posted here/our internal SQA of the APM software and I don't often get to a multi-mile wide area that is flyaway safe.
The last such test resulted in a near disaster fortunately in a multi-mile wide test area(Ag facility in northern cali).
If the 3DR supplied lipo wasn't absolute crap it could have reached a major highway and caused issues there(for once glad of the crap lipo).
The numbers published at:
indicate one way the thrust curve can be made safe is to go to 4S(3min) flighttimes(definite overheat on motor/power system with present prop and past the motor limit but hover is at 34%)
(hmm a 10x3 prop at 4S will give a 49% hover but no vendor, 10x3.8 pushes the powersystem into bad areas)
or 3 bladed props and 3S with an APC slowfly shape:
and keeps the hover throttle at a fairly safe 54%
however these solutions BOTH suffer from practicality.. I have NOT found a vendor for 3 bladed props with a hub thin enough to fit the 3DR prop bolt(longer replacement prop bolts would be quite welcome and the alternative) nor have I located a vendor with a 10x3 pitch prop.
I've never flown IRIS with tall legs and gimbal on a 3S lipo with 2 blade props.
With 3S lipo I've only flown her with 10x5x3 props (Graupner). Current draw was around 19A. Hover throttle was higher than 70% and the thing didn't feel safe to operate.
With 4S and 9x4.7x2 props (APC slowfly), tall legs and gimbal I get about 13A of current draw. So far this seems to be the best combo for the tall legs + gimbal config.
With 3S and 10x4.7x2 props (APC slowfly), short legs and no gimbal, I get about 13-14A of current draw.
Very interesting remarks. Can you expand a bit on the safety concerns with regards to the APM software. So far I was under the impression that the IRIS hardware needed quite some work to bring it out of beta state into something usable. But I never questioned the software. Now you got me a little concerned. :-)
With regards to the prop vs. 3S/4S configs. I've done some tests and are still planning to do some more:
I have ordered 10x3.8x2 (APC SlowFly) propellers to try that combo and see how it feels. I also bought an IR thermometer gun to measure motor temperature after test flights. I went through these configs in eCalc and did see the risk of running the motors hot. Therefore the IR gun. I'm also fortunate enough to fly at sea level.
Also very important: I had to swap the propeller mounts on the stock IRIS, as the stock prop bolts were too short. I'm using this one now: https://store.3drobotics.com/products/propeller-fastener-kit
Hope that helps.
Mark, where are you located at? What's the elevation of your location?
I'm located at sea level and judging from the Youtube video of Murat, he must be as well.
If you are located at a higher altitude (+ higher temperatures to make it even worse) your performance will suffer quite a bit. Have a look here: http://diydrones.com/group/iris/forum/topics/iris-with-gimbal-and-t...
In e.g. Denver you won't be able to get IRIS off the ground safely.
Graupner is quite a bit different than APC slow fly profile is what I am gathering and still different than these 3 bladed monsters from GLB/HK etc.
The safety concerns come from a number of patches to the original software and a lack of SDLC/SQA/QA methodology and the internal design/structuring of the software as well.
AT least some of the incidents in this forum NOT related to hardware/mounting relate to this instead including the indoor accident mentioned 24 hours ago..
have had a few of the propeller fastener kits shipped from 3DR all were too short for the 3 bladed from GLB(GLB was fairly well balanced however :)
RTL activation indoors due to no true manual flight mode state in effect but while non GPS flight mode selected.(ie Finite state machines with allowed states defined in state tables is ONE way to fix this from the design end)
Its NOT a problem easily fixed by QA/SQA IMHO but instead relates to the software development methodology selected and the design patterns used to develop software for the FC/APM.
BTW where did you get the 10x5x3 graupner.. seems to be rare as hens teeth ?
btw my fix for the software is to port OpenPilot into the pixhawk environment if I keep on this path ...
easier just to order OP flight controllers , did I mention OP is developed to the precepts of an SDLC?(only one I have found like this so far where I can actually trust in bug tracking for the FC firmware).