I'm new to the Spirit FBL world. I've been using V-Bar for a while. I picked up a Spirit RS unit and am encountering some unexpected behavior. Every time I connect the main battery the startup looks normal, but I always get a message the the receiver rebooted within the first 5 seconds. Is this normal for this type of FBL? None of my normal Jeti receivers do this.

Are there any known issues that would cause the RS to stop responding to inputs? I've been flying with the RS for 3 weeks. At the end of my 2nd flight this weekend I pulled into a hover to land and the heli just started drifting until it tipped over. No low RF alarms, just no response to stick inputs. It was only ~15' (5m) away. Post crash control checks showed unexpected servo response (not linear with the sticks). Checking the Spirit logs on the radio showed sporadic reports of corrupted packets. Attempts to access the log from a PC failed with a "corrupted data" indication.

I confirmed the anti-static kit on the tail is grounded to the motor mount, and re-flashed the firmware to 3.5.0. I'm still getting the RX reset but the logging is now readable, the servo response seems to be better, and no corrupted packet reports. The TX is a DC-24.

Can someone with an RS please let me know if the reboot is normal?

In case it matters, the RS is powered by a Castle 2.0 BEC set to 8v with a 6s lipo as the source. Both BEC outputs are connected to the RS. I have a telem uni cable connected to a Jive ESC. The Jive BEC output is isolated from the system.

As you would expect, I'm a bit gun-shy of this unit until I figure out what is going on. An unresponsive heli in a 3' hover is not a good thing.

Thoughts / suggestions?

Hello,

it seems that something is connected at the E2 pin. Please make sure to not plug anything there.
No reboot message should be present. This is very abnormal behavior.

Please note Castle BEC powered from 6S will give only fraction of current that is advertised by manufacturer. So using JIVE BEC could be actually better.

_________________
Spirit System developer

I have the 2nd power connector from the BEC connected to E1. E2 is empty as requested in the manual. Just to check, I removed the connector at E1 and tried again but received the same receiver reboot message. All ports are occupied except SYS, E2, and SAT2.

I tried the Jive BEC initially but the HV servos I have don't work with less than 6v. The Castle BEC is rated for 8-9A with a 6s source.

One observation: The "Spirit" device appears on the radio device list immediately after the reboot occurs. Prior to that I only see "REX-RS".

It seems the reboot is happening when Spirit unit will signalize successfull initialization. So it is enough for servos to move faster. It look like BEC voltage is falling under 3V.
Are you using some cable extensions for BEC cables?
I can recommend to use a different servo connectors maybe. Some users reported that this resolved problem with BEC voltage drops.

_________________
Spirit System developer

After some experimentation I confirmed that the reboot does not occur if I add a capacitor to the circuit, currently 4700uF. Wires are 22AWG and do not have any extensions to the original leads from the BEC.

Would this under-voltage explain the lack of controllability I experienced at the end of the flight? The servos weren't really moving much at that point.

I ran the BEC test and got a steady low voltage alarm (~3.9v) even with the capacitor.

Is there a recommended BEC to use for a 6s setup? The Castle doesn't seem to up to the task.

Something like the Western Robotics Hercules Super BEC PRO https://www.amainhobbies.com/western-ro ... sp/p265789 should work, but it comes with a few downsides. Like the CC BEC Pro, installing this WR BEC adds bulk and makes wiring messy. With later generation ESCs featuring robust internal BECs, consumer demand for external BECs continue to diminish. Unlike the BEC Pro with selectable voltages in 0.1V increments, this WR BEC has only 4 voltage presets (5.2V, 6V, 7.4V and 8.2V) which means you have only two preset voltages (7.4V and 8.2V) to choose from.

Other than trying a more modern ESC with a more robust internal BEC, consider replacing the small glitch busting 4700µF capacitor with a super capacitor buffer unit.
25Fx3 https://www.amainhobbies.com/xguard-rc- ... 9/p1406624
50Fx3 https://www.amainhobbies.com/xguard-rc- ... 8/p1406625

depending on your location: Chargery or MTTEC Keto could be great BEC choices too.

as you are getting that low voltage alarms even with a cap: it seems the BEC is absolutely breaking down... even a supercap buffer will not be for sure a safe choice i guess. with some successive demanding manoevers even the caps could discharge too low and break down.

_________________
cheers
Michael

From our experience Castle BECs have the lowest performance at the market. Their rating is far from reality. In any case these circuits are years old, designed for servos that were available years ago, so it was sufficient.
However for todays servo it does not work well.
Any faster movement could result in voltage drop under 3V which is not enough for any electronics. So it could fail/crash even during pitch pumps.
If I can recommend then look for recently released BECs/ESCs. These are usually rated for 10A or more.

_________________
Spirit System developer

I can only agree, made bad experience with a Castle Creations 2.0 in a Voodoo 400 helicopter and a CC 10A in an F3F glider.

In both cases I could measure short output voltage drops down to 0V with ~2ms duration, followed by receiver reboot.

At the helicopter the issue occurred during a spirit BEC test on the desk. I strongly recommend to run the BEC test before going to the field.

My measurements did not show excessiv servo currents. The measured current peaks were below 10A and were below the rated CC limits. For unknown reasons the CC devices apear sensitive to short current spikes, possibly short enough not be captured by my scope.
I tried with different samples of the same type, not all of them showed the voltage shutdown with receiver reboot. Apparently the CC devices are different in this respect.

This is a capture of the BEC voltage (blue trace) and servo current (yellow line) during a spirit BEC test.
voltage scale = 2V/div
current scale = 10A/div



tail servo: 1 x Savox SB-2272
swashplate servos: 3x Savox SV-1257