Spirit System

I have a problem, that Spirit sometimes does not initialize right, after update to 3.2.2. It makes a normal serwo twitch, and then - one second later - it twitches to the side and stays there. Always same angle. See photo. I wasn't able to make video yet...

I think it's software update, bacause it happens on two of my helis, both were fine before the update.

It happens around once per 10 times, and restart typically helps.

Helis are
- Goblin 570, Spirit 1 (red), FrSky x6r receiver, ACCST D16
- Goblin Kraken 700, Spirit 1 Pro, FrSky x8r receiver, ACCST D16

Radio is FrSky Horus x10s with ISRM update.

Here's what I've changed recently:
- Spirit update (probably from 3.1.0)
- ISRM module installed in Horus (gives access capability)

I think it's not related to failsafe because I have failsafes set right.

Any idea?

Hello,

Can you verify Diagnostic tab? What values you can see there? Are all channels in the middle?
It look like it is related with receiver data.

ZeXx86 wrote:Hello,

Can you verify Diagnostic tab? What values you can see there? Are all channels in the middle?
It look like it is related with receiver data.

Sure, I'll check tomorrow and report.

ZeXx86 wrote:Hello,

Can you verify Diagnostic tab? What values you can see there? Are all channels in the middle?
It look like it is related with receiver data.

I checked today, and you're right - my channels are not in the middle. Because my spring tension is light, sticks go down a little bit, if radio stays vertical. So it has some negative pitch and negative ele. But it's at zeros when taking off and flying. (I am aware of some risks with such setup and do care that my channels are centered before takeoff, I have a very special setup with my radio...).

But, I tried to then reproduce the problem, to make sure it'is found. I trimmed thr/pitch and elevator a lot, to have even more negative values. But Spirit is still starting fine.

Any thoughts?

I am not sure I understand your setup.
But it is very important that sticks are centered during initialization, not only before takeoff.
Pitch stick is not important but aileron, elevator, rudder must be always in the center during initialization and take off.
Channel center is calculated automatically each powerup.

ZeXx86 wrote:I am not sure I understand your setup.
But it is very important that sticks are centered during initialization, not only before takeoff.
Pitch stick is not important but aileron, elevator, rudder must be always in the center during initialization and take off.
Channel center is calculated automatically each powerup.

My setup is complex, I have modified my radio and flying style heavily. In short, my sricks don't have "hard stops" in the middles. They go smoothly through the center.

I will have to cover it with logical switches then, so everything is zero initially....

This also explains why I was seeing some forward tilt while taking off, although channels were centered. They weren't centered for Spirit.

This also explains why there isn't much prc eision on diagnostics screen, no fraction etc. so you can't set it up precisely. It just doesn't matter, because Spirits calculates centers at startup anyway right?

Yes, that is very correct.
Feeling center for pilot in the air is very important, because you will be never able to really hover or fly stright.

This behavior is identical from version 1. So firmware 3.2.2 is not changing anything there.

ZeXx86 wrote:Yes, that is very correct.
Feeling center for pilot in the air is very important, because you will be never able to really hover or fly stright.

This behavior is identical from version 1. So firmware 3.2.2 is not changing anything there.

I changed my setup for all the helis. I have now a "throttle cut" switch, which is always engaged, when connecting the battery. Besides handling throttle, it also forces pitch, rudder, ele and ail to zero. So I always have zero positions when initializing Spirit.

But the problem still happens.

I think it's related to some tiny movements helicopter has while Spirit is initializing. Maybe new software is more sensitive then? I do connect battery during initialization sometimes (I connect backup power first, which initializes Spirit, and then main pack), but I'm very careful to not move the model in any way. If there is any movement, it's smaller then the one made by wind on windy days. And never had issues with older softwares. I'll keep observing this..

I have some new insight on this topic.

First, I replaced backup 2s Lipo with a 1st buffer 50F. And the problem is still there. So this excludes that connecting the second power source could be problematic.

Then, I also replaced Kosmik 200 with Tribunus 200. Difference is, that Tribunus reports THR errors with sound (THR signal missing or THR too high when connecting).

And what I can seem is that when Spirit fails to initalize, there is some problem with THR too. And this happens early, even before Spirit makes the usual servo movement at the end of initialization.

So if Spirit initialization fails, it looks like this:
- I connect the battery
- Servos are moved to the center immediately
- Esc arms, confirms voltage with sound
- Esc starts to beep with THR error every second - then I know sth goes wrong with Spirit
- after some time Spirit makes the usual movement with servos (ESC still beeping)
- one second later Spirit goes weird and moves serwos in weird direction

The log shows only 'cycling ring' messages.

Unfortunately I don't know if Tribuns reports 'THR signal missing' or 'THR too high' (damn Morse code in Tribunus...), but I suspect the first one.

Any idea?

Have you tried to update to 3.3.1 firmware? There is improved compatibility with FrSky latests updates in receivers.