So I've been using Spirit 2 since april this year on my Blade fusion 270, and have collected over 1300 flights with it so far. The one and only reason I gave this unit a try is because it was supposed to have one of the most reliable rescues on the market. I'm into helis just over 2 years but I took it by the storm, flying a lot, and wanting to learn 3d, so a reliable rescue comes real handy. Here's a video of my current flying style: https://www.youtube.com/watch?v=MMHtBJgbsTI
And Spirit's rescue lived up to it's name, I was super happy with it's performance just until recently. Always perfect rescues unless vibrations on the model were really high (after some fast & sloppy repairs), but in that case, rescue climbout was merely like 20° from vertical at worst, so still pretty good.
But in the last few days I've literally had it 2 times want to shoot my heli into the ground, both resulting in a crash after I Throttle-hold it quickly after weird rescue behaviour.
So i started to wonder if some sensor is starting to go bad (after all it has 1300 flights & about 15 crashes on it), but then unit wouldn't pass the startup self-checks I assume, and there would be some error logged in log file?
Some info: rescue type i use is "normal with acro delay" 0.6s, with 75% rescue pitch - so leveling to the closest flat orientation (upright/inverted), 0.6s climbout & then roll to normal orientation+climbout until rescue is disabled.
Vibrations on the bench are within normal limits - same as ever before.
I was running version 2.8 and worked great for me until last weekend where that first awfull suicide rescue was performed. So i even switched to the latest FW 3.2.1, and although the heli seems to fly even better, rescue failed miserably once again on this FW.
Now a little bit detail from this last failed rescue: Since i had new FW updated on Spirit i was cautious the 1st flight. As always 1st thing testing rescue on the ground before flight: OK. So took it up, 1st check how it flies in all modes, then went on with rescue from upright hower: OK. Rescue from inverted hower: OK. Rescue form a funnel: OK, rescue from backwards hurricane: OK, rescue mid-pirofliping: OK. So i though it's all fine.
Now here comes the second flight: prior flight as always checking rescue on the ground: OK. Then flying it and towards of the end of the flight, just before finishing some ailern tic-tocs, figured I'd try rescue 1 more time, and there comes the horror - from like 45° sideways upside position, heli turns like 45˘sidways towards the ground and strats shooting for the ground, so quick TH from me and dig into dirt (i don't fly very high usually so dirt nap happens fast). Went to pick it up, and when placed heli upright on the ground, i checked the rescue again: OK - swash mopves directly up. No error messages saved in the LOG again.
Then at home after repairs, i powerd up, tried to move heli all kinds of directions, shook it, turning wildly upside down and back for like half a minute, then place it on skids imidiately, testing rescue, and again it worked as it should.
Also the interesting part is that on this last occasion rescue worked fine for the 1st flight (heli was just out of a warm car), and failed towards end of the second flight, when unit had probably cooled down to outside tmepreature (just over freezing, 2°C).
So the only thing that i can think of that changed from all the time before when rescue worked flawlessly and now that it doesn't is the outside temperature. Only the last few days it's around freezing when i fly outside. It wasn't as cold ever before since i started to use Spirit in april. And i've read somewhere here on this forum a topic on this, some user had problem with rescue when he was out in less then 10°C...supposedly this should be solved with FW 3.1 and newer, but maybe the problem still persist in really low temperatures...is there a chance accelerometers get stuck for split second in cold, giving out wrong data, or data processing slows down bcs of cold?
Advice please, I've started loosing trust in this unit should i just ditch using rescue and fly more conservatively or refrain from flying at all for the next 3 cold months, or somehow warm unit between flights, or could there be something else (I hope this behaviour is not permanent)?

Edit: All the swasplate movement vs stick input aswell as all compensations work correctly afcourse. Using KST DS215 for cyclic and Spektrum H3065 for tail servos, and Spektrum 4649t RX. I also added picture to show the unit's position on heli and unit's setup file for my heli.

Hello,

please check this page:

http://manual.spirit-system.com/index.p ... ise_Rescue

We are testing rescue for many years also in -30°C, so it must work properly.

But temperature is playing role indeed. It is advised to not cool down the electronics for many reasons because then water vapor is condensing in the electronics.

No, sensors are not stuck in low temperature.

Stabi Correction can help in your case, but if it is temperature dependent waiting after initialization can also help if electronics was cold.

_________________
Spirit System developer

Thank you for answering.

Yes I've already read several times the manual page of Imprecise rescue. I always tend to look around for answers myself first, before starting to ask questions..

Now I also went out today for 7 flights, being carefully a bit higher then usual to be able to recover from bad situations myself w/o use of a rescue. Temperature was exactly 0°C (according to my car). So the good thing is, I figured i'm not crazy, as from last two occasions I was still a bit unsure of myself (as things happen lightning fast)...what if i just disengaged and Throtle-hold too quickly..but i figured out today that rescue indeed is going crazy.
So results were the following:
- on the ground before the flight - always OK
- beginning of the flight upright hover - always OK
- beginning of the flight inverted hover - always OK
- after a little flying around, flips, rolls, hurricanes, funnels - always OK
- after more demanding manouvers - piroflips, tictocs, piro tictocs - rescue always goes BAD, performing up to 45° sideways climbouts or even worse, and remains that way or worsens till the end of the flight.
- one flight after doing a few right-rudder piroflip repetitions - rescue goes CRAZY - shooting heli almost straight towards the ground (from an upside hover), so i disengaged it quickly and regain control. Knowing what to expect i tried it few more times before the end of the flight from a safe height and it stayed same horror until the end of the flight, and also after the landing, swash was moving down and sidways upon rescue activation instead up (obviously showing sensors must be reading that heli is in flat inverted orientetion - while in reality it's flat upright).

Again no error message stored in the log after coming home and hooking it to computer.

So that leaves me totaly confused and still hoping for some kind of solution. Will try with enabled Stabi correction, but still i don't believe this will help much since it all worked great till like a week ago, always flawless rescue performance throughout entire flights and every manouve i'm capable of doing atm, felt like nothing could get in its way. Now out of the blue it feels as usless as Ikon's It will rescue from basic flight and basic 3d manouvers, but thats not what I need it for. I need it "in the heat of the battle".
Still hoping it has to do something with the cold and possible condensing on electronics as you mentioned, and can only wait for some warmer days to confirm or deny this.

Please if you have any further ideas with more info that i provided from todays flights, I'd be happy to read some further response.

Your observation is perfectly logic and can be perfectly explained.

I believe there are at least two possible solutions to the problem.
Yes, it should work perfectly after any maneuver, including 3D flying. We are developing Rescue system much longer than competition for many years so I can tell you everything.

In any case, for every sensor, big difference at the beginning and at the end in temperature is always causing unwanted drifts.
Electronics is warming up for around one minute to around 30°C from 0°C.

Drifts are also caused by few other things so it is good to avoid them.
If you will leave model for a several seconds after initialization it can improve the situation and make it perform in the same way as in hot days. Longer it will be at rest after initialization, the better for precision. This is recommended by all gyro manufacturers.

It is all dependent on model condition, maneuvers that were flown previously (before rescue is engaged), temperature difference and especially time. At the end of the flight probability for any imprecision is always the highest.

But Stabi Correction is using additional sensor to fix a drifting issue. And it does not matter why it started to drift or how long model is flying. It can be used on 95% of models but with a models that are disturbing such additional sensors result could be even worse even at the beginning.

You can always try rescue safely according our instructions. No need to risk in the air.

_________________
Spirit System developer

OK, thank you for your suggestions. Will try with Stabi correction and leaving model on the ground for longer times after initialisation next time when i go fly, and will report back if there are any improvements.
Snowfall forecasted for next 2 days, so no flying at least till firday though

So i put another 7 flight session in. Weather 4°C this time and high humidity.
Stabi Correction enabled this time.
Waited for around a minute after initialization with heli sitting on the gorund before spooling up on 5 out of 7 flights.
Results aren't very good.
There was not any difference if I left heli sit on the ground for a while after initalization.
Swash always move straight up on the groun before flight after activating rescue.
After take-off and just hovering or flying circuits all upright, rescue was good stright up.
After any kind of 3d, rescue started a drift forward. As 3d gets harder, the more forward to forward-sideways drift the rescue develops, to the point its slanted more then 45°.
But to the contrary of the last time i was testing, it was consistent in drift direction - always forward and a tiny bit to the left.
That being said, i made another interesting observation...since i don't trust this rescue anymore, i always tried it when hovering upright tail-in first. So as i described the drift was forward, means heli drifted away from me. Then one time after a try tail-in, i bring it back and hower it left side facing me...and watch this: upon rescue activation it drifted away from me again, so not forward but directly to it's right side. Then i switched side facing me and it drifted away from me again, this time to it's left side.
It's feels like somewhere in the middle of one manouver, it detected that ground is sideways toward my direction, and like that data got stored till the rest of the flight...

Anyways, not much progress made, I still can't believe rescue on this unit worked flawlessly for over 1000 flights, now all of a sudden it became uselles.
If nothing else I'll have to wait for warmer days to see if problem will still presist, and if it does, something inside unit must be going bad bcs of use in that freezing humid air last two weeks. But to find that out, I might have to wait for spring...
Is there a possibility to return the unit at some point as a damaged one (even though there is no visibla damage) if I decide to do so, and get a new one 50% discounted, you have some sort of that kind of option, don't you?

Best regards, Tomaz

Hi Tomaz,

I am following your thread, it´s really interesting.

Question: Could you reproduce your testing results with a bigger heli ? I am not sure whether a 700 would behave similar.

Best Regards,
Eugen

Hey, no unfortunately not..this one is the biggest i currently own and fly...day after day

But as i already mentioned: rescue have been pretty much perfect for 1000+ flights before, same heli, same setup, same unit, sometimes with far worse vibrations then what they are in current state, but still it worked flawlessly any time and in midle of any manouver. I probably literally used it like 1000 times with complete confidence...and now it became just as usless as thereis with many other FBL unit brands, so i still believe something bad must be going on in my unit, because the only thing that is really different from before is weather (temperature mostly).

Hello,

thank you for the details.

If there will be no other choice we can verify your unit. Certainly it is not normal.
It would be nice to make list of anything that changed from time it worked.
Was it really only temperature?

Even different cable wiring can make a difference.
Also please do not forget helicopter is full of rotating parts that are wearing with each flight.
So such changes can happen on its own.

Vibrations can be very low, but if there is some at high frequency that is causing big disturbances it can be enough. For example from motor bearing. But it should be all visible in the analysis plot.

No, we are not giving discounts in such way. Especially while not knowing why it is happening yet.

Now it rather seems that temperature is not playing role.

It can be just bouncing/free connector in the frame or any small difference.

_________________
Spirit System developer

Hey, thank you for reply.
What do you mean by bouncing/free connector in the frame?