I have been doing some testing (outside the US) that involve very high altitudes - 1800 meters or more. The climb takes no more power than coming down. Even though going up takes more power per minute, coming down is slower, so the total power consumed when coming down is easily equal to the power going up.
I was thinking that it would be more energy-efficient to shut off the motors at the highest point, and letting the quad free-fall and then turning the motors on again at 30 meters or so to right itself and slow down for landing. Has anyone tried this? I have been told that dji can do this.
I'm using APM and PixHawk controllers.
I've done that with my tiny $98 plastic quad, it recovers quickly, It'd probably take a larger, heavier copter longer to recover and you may want to make sure your battery is well secured or the copter my throw it out when you throttle back up, and it may put undue stress on the copter's arms, but it might work.
What mode were you flying in? You mentioned that you wanted to go straight up and straight down, if you were in Loiter or Alt Hold, you are limited by your max vertical speed parameter. When I'm in stabilize mode, I descend like a rock at minium (not zero) throttle. I have never ascended to 1800m, but from 120m I would estimate 5-7 sec to the ground. It really sounds like your descent is being limited.
I was flying in AUTO mode the entire time. The max descent (in Mission Planner) is 500cm/sec. I set mine to 450 and it wobbles quite a bit on the way down.
The wobble can be reduced by forward movement. In auto mode, I would create a descending spline circle. However, to increase your speed coming down, you will need a different mode. Before thinking of turning off motors, I would try descending in stabilize, using telemetry like an instrument flight. Be wary, coming down fast you may need more than 30m to slow down.
Depending on your quad size it may be possible but when stabilization is turned back on it will stress the hell out of the quad. A lot of variables here to say it will be safe, unless its a strong acro quad with carbon fiber or composite props I wouldn't even think about it..
With the latest generation of ESC's with active freewheeling enabled, "right side up" descents are very fast, much, much faster than without it enabled. You may even be putting power back into your battery with active freewheeling. Simonk has it and so does blheli.
Like other have said, doing a controlled dive is much safer but since you are flying beyond visual sometimes you may dive the wrong direction. Depending on your setup, even a dive may cause failure. Been there, done that! Good luck
In a remote part of a country outside the US, away from commercial and GA air routes, I have taken a Phantom 2 with gimbal and GoPro up over a mile AGL on numerous occasions, for AP reasons. I descend in what DJI calls manual mode, which is essentially idling the motors, unless I apply forward left stick. I will let it free fall up to 10 m/sec descent rate, and the quad holds attitude well. Sometimes I will give just a bit of throttle to drop descent rate to 6 m/sec. I usually start spooling up around 200 meters, but will stay in manual mode until somewhere below 100 meters. I trigger manual mode with the momentary switch SH on my Futaba 14SG radio, so if I experience brain lock, it's likely to come out of manual mode. Regarding VRS, I've wondered how much dirty air is actually beneath my blades when idling down. In any case, I'm ready with right stick forward if I sense VRS. Throughout the descent, I am in the googles watching altitude, vertical speed and pack voltage. Could one get down faster and on fewer mAH with motors off? Sure. Or in a dive? Sure. But idling down, or throttle just above an idle, avoids any motor/ESC risks, restart risks, major tumbling, etc. Tumbling not great with 3 axis gimbal powered, and camera on board. I also chose my descent location to be over soft vegetation, just in case.
I'll have to try it sometime. I have never flown a 'factory built' craft, so I can't accurately predict how any of my quads/hexes will perform (I have 8-10 of them), but I'll try some day soon.
Didn't you say it was a one time event, never to be repeated?
I've done this with my Blade 350 QX. But not at the altitude you're talking about. I was maybe 250 feet up, cut throttle, and try to wait until the last moment to fire them up and recover. The 350 QX had some decent punch out and would only take about 15, maybe 10 feet from free fall speed to achieve hover and hold altitude again.
With the QX, bringing the throttle all the way down didn't fully stop the props. They would maintain idle, but it would still go completely silent and drop extremely fast. I think maintaining at least some throttle would be wise. It will use a little bit more power, but it will also help maintain orientation, prop direction. That means less power will be spent flipping it back over or getting the props back up to speed. Plus the quad will still technically be in your control. Tumbling thousands of feet while flying blind (you said you couldn't see it when it went up that high), and expecting to eventually make visual contact with it and then step in and have all systems respond is dangerous and careless. If you quad is in the air, you should be piloting it. Even if it's a "controlled free fall", but tumbling blind is not piloting. Go 5 or 10% throttle at least.
I'll admit there was something cool about the sound of angry bees suddenly disappearing and then swarming back in at the last minute to save the quad. But I never felt like it wasn't in my control.
I was thinking of going over to Nevada and getting FAA approval to try for an altitude record. I wanted to use virtually 100% of my batteries to go up. Although I could use a parachute, that would be extra weight.
If you are going for an altitude record, does it need to land? I mean, can you just go up as high as possible until batteries fail and then let it come crashing down? I'm not necessarily suggesting that, as it involves some potential dangers. But in an unpopulated area, like some remote parts of Nevada, it might be alright. Of course, it would cost you the cost of your aircraft, but to break a record may be worth it.
I would think that at record breaking heights, a parachute, even if deployed at lower altitudes would still be high enough to potentially, and likely, drift in the winds. Since there is a very good chance it would be lost, might as well not add the weight of the shoot and let it come crashing down. Again, not actually recommending it, I'm just bringing it up for discussion.
When the Model Rocketry clubs come to the Dry Lake Beds here in Nevada, they get special permission from the FAA to do record height flights with their high power rockets. That would be a perfect time to hook-up with such a club and go for a legitimate altitude record!