Spirit System

This is very disappointing I have to say. This is the SECOND time my Spirit unit has hot started on me.

First time was probably a month or two months ago. I was tuning the heli with the WiFi module. I was kneeling next to the heli. As soon as I hit SAVE the heli hot started, full throttle. It gave me a bruise, a cut and destroyed my phone I was using to program. To avoid a chicken dance (it was an electric Goblin 500) I put my foot on the head and ripped the canopy open to unplug the battery. Thankfully it just fell on its side or the damage would have been worse.

Now fast forward to a few days ago. This time it was a different heli, with a different Spirit II unit, without WiFi module. I was doing tuning with my laptop, the Windows software, and the USB cable, so very different scenario.

I was tuning tail parameters to get rid of tail wag and to make the tail more locked in. I was doing a ton of tuning that day on all my helis. This was at least the third time this same day I was connecting to the 570. One thing I noticed was that when I was doing tuning, I had the warning message that I was going to be defaulted to Bank 0 and that bank switching was disabled. When I had the hot start, that warning did NOT pop up. But I never thought anything about it. I changed one parameter, set the Rudder Delay to 0. Went to click SAVE, and the heli hot started right away. Thankfully it just spun one time, then the motor stopped. There was no damage, it was sitting on a table. All that happened was the USB cable wrapped on the tail one time, that was it.

But this then leads me to believe the issue is when there is some sort of software bad initialization or error, that the unit just goes crazy and hot starts. Maybe the unit is rebooting and sending full throttle in error. I do have fail safe configured and works fine if I turn off the radio, then the motor stops.

Now I am afraid to keep tuning, and I had shelved my WiFi modules thinking that made me safe, now I am not sure!

First adive NEVER work with the blade at the worshop with power batteries! Or use external 2S battery...

Hello,

it is very sad. Good thing is we have never heard about similar issue within 10 years. So I guess there is something wrong in your setup and/or it can be avoided.
Unfortunately we will need much more information about your setup to be at least able to somehow reproduce it or try to find the reason.

If Failsafe was configured properly and tested, then the only issue that could be technically responsible is out of synchronization at the receiver signal line.
This is very dependent on what protocol you are using, because some protocols are safe and with some this could potentially happen.
For example SRXL2, EXBUS, PWM, PPM are perfectly safe. In case of SRXL and Spektrum DSM2/X it is less safe but still this scenario should not occur.
Only with Futaba SBUS the problem could happen in a certain situation. This also depends on framerate, modulation you are using. Additionally there are several implementations of this protocol, for example from FrSky.

It has no relation to Wifi module at all I guess. Except this could cause some additional interferences, at least with FrSky radio, depending on position.

So my question is what receiver and protocol is used?

Tomas,

Thanks for the reply. Part of my setup process is to check failsafe. Once the configuration is done, I remove main and tail blades, spool it up and turn off my radio. That way I check my fail safe is correctly setup.

I use Lemon DSMX RX diversity satellites, two on each helicopter both the 500 and 570 have the same setup.

Two hot starts sure are Dangerous. Hope you get some answers here toad. Keep us posted.

Thank you for the details.

Are you able to reproduce the issue, at least to some degree?
Also what is your radio and frame rate? These two things could play also a role.

We have some Lemon satellite for testing and we already found they are not behaving perfectly identical as original ones.
There could be many firmware versions and hardware revisions so it is possible one might not behave as expected. This could be also direct reason for what you are observing.

The key is to reproduce the issue somehow. Until now we were unable to do so yet But for sure Save button will be likely needed to trigger it.

I have a Spektrum DX9 Black Edition running 2.05 firmware. Frame rate is 11ms DSMX.

For now the helis seem to be running fine. Since I have a long winter hiatus in the hobby and dedicate to simming, I can do testing to try to reproduce it. For what's left of my flying season I will not attempt to do any more tuning. Is there any way to capture logs or something to provide to you if indeed I am able to reproduce it?

What I run is LEMON RX DSMP (DSMX COMPATIBLE) DIVERSITY SATELLITE. I run two on each heli, very similar placement. One is on the bottom with antennas pointing out (left and right) and another is on the side frame, antennas pointing up and down.

The link to the product is here: https://lemon-rx.com/index.php?route=pr ... ption=true

Do you think that the 4651T could be more stable? I was thinking of upgrading to that one to be able to use integration. It works really well and I think it's an amazing feature, as I use it with a 4649T on my Goblin 380.

As a quick update I re-ran my failsafe tests and will be releasing them on YouTube, just to show the process of turning off the radio and seeing the motor stop.

Tomas,

Here is the video:

https://www.youtube.com/watch?v=KDmmSejWgws

Please note after disconnect, when entering failsafe, the swash tilts hard and the tail also pushes hard. This happens on all three models I tested, Goblin 380 500 and 570. I will be uploading to my YouTube one video each day, today in the late afternoon US time the 500 will pop up and tomorrow the 570. Is the swash/tail movement normal or can that be configured?

Hello,

thank you very much. The failsage engagement looks good, but normally swashplate should be level, not tilted.
This depends on how the channels are saved for swashplate.
You can check diagnostic tab to see what features are engaged and how channel values are looking.

Important is that motor is stopped.

I recommend to try reproduce the issue in the same way as originally. By clicking at the save button.