Disclaimer: I am a software engineer by trade and have had a small smattering of electronics training, however, I have not written any stabilization software. My rant is based on an outside observation and using educated guesses as to how these units work.
First my story goes like this; I have a 60" powered 3D plane that was originally setup with a non-stabilized FrSky receiver. Because the plane is super light, a moth fart will make a noticeable twitch, so others suggested using a stabilize to make it feel, "heavier". I bought an Archer SR6 to throw in there. Mind you my radio was already setup and working, throws, directions, etc. had long since been set.
The first thing you need to do is "calibrate" the receiver by going through a series of "set it like this and wait" motions. I am not sure what chip they use in their receivers, but what I do know is the chip itself does not need calibration. On a side project I have been playing with some 9 DoF (Degrees of Freedom) chips, which contain an accelerometer, gyro, and magnetometer, all having 3 axis of sensing hence the 9 degrees. I don't think these receivers need the compass, so I assume they are using 6 DoF which contain accelerometers and gyros. Here is a data sheet from one and you will notice, they are very clear about output values in relation to the "dot" on the chip.
Given the chips all know what direction they are moving why does FrSky need us to calibrate? Well, I am guessing that is so they are free to mount the chip any direction they want, and the calibration process lets them determine which way is "up". This process though a bit annoying, is a minor issue and I don't really care about having to do it.
Next, you have to mount the receiver, and this is where it starts to get annoying. Rather than letting me just tell the software the direction the receiver is facing, I am forced into a very limited set of directions. Why? This is lame. My aircraft is large and I am able to mount the receiver the way they require me to, but this certainly could pose issues for smaller aircraft.
Next, I setup my radio with a switch for channel 10, the gyro mode switch, and a knob for channel 9, gyro gain, and I ran the self check. I then enable the receiver using the SXR Enable button,
and then I check the direction of correction, and during this I notice the direction of correction is wrong for both the elevator and the ailerons. No biggie right? There is a bunch of "Inverted" switches in the settings, so I reverse both surfaces, and viola, they are now doing the proper correct. Job done right? Nope. Now both surfaces apply proper correction to motion, however when I move the sticks on my transmitter, the surfaces move the wrong direction!
Why? I cannot think of any reason, the "invert" switches cannot just change the direction of correction. At this point I was confused as I thought I had the receiver mounted wrong or ? so I spent time trying to find a picture to make sure my "pins to the rear, label up" was indeed referred to as "Level". In OpenTx I believe there is a wizard that shows pictures of the various positions, not so in Ethos. I did have it mounted correctly and at this point, I gave up and just disabled it.
A couple days later, I decided to man up and solve this stupid thing. I double checked everything I had done to this point and decided I had not messed anything up. So, I then decided that since I had to invert the surfaces at the receiver level, I would reverse them on the transmitter, which should make them respond correctly to stick inputs and the motion correction would remain correct. Of course when I reverse the channel output, guess what else happens. Yep, the outputs were not at exactly 1500µs so both ailerons and the elevator need to be recentered. Arggggg. Ok no biggie, this is again another annoyance of their system but not a deal breaker. So, with stabilization in the off position, I recenter my surfaces, adjust my throws so they are once again even and correct, and then I run the self-check, which should pick up the new center and end positions. Whew, finally done and ready to fly, right?
I flip my stabilizer switch from off, to stabilize. Both ailerons move up a bit, and the elevator moved down. OMG!!!! Why? I am sure there is something I didn't do quite right, but this is the lamest thing yet. There is zero need to move controls unless the stabilizer detects motion.
I QUIT!!!!
I doubt FrSky will ever "fix" these issues, so I will never again use one of their stabilized receivers. I will buy one of the $15 stand alone units used in quads.
I really like FrSky products, I have had an X9D, the X9D 2019, an X20S, and I just got myself on the list for the X20 Pro AW. Their receivers, as receivers have always been rock solid for me, I have used a number of their Neuron ESCs, but their stabilization is...I am just going to say it....horrific! Never again.
(one final note, I have set one of these up before (and used it)).
First my story goes like this; I have a 60" powered 3D plane that was originally setup with a non-stabilized FrSky receiver. Because the plane is super light, a moth fart will make a noticeable twitch, so others suggested using a stabilize to make it feel, "heavier". I bought an Archer SR6 to throw in there. Mind you my radio was already setup and working, throws, directions, etc. had long since been set.
The first thing you need to do is "calibrate" the receiver by going through a series of "set it like this and wait" motions. I am not sure what chip they use in their receivers, but what I do know is the chip itself does not need calibration. On a side project I have been playing with some 9 DoF (Degrees of Freedom) chips, which contain an accelerometer, gyro, and magnetometer, all having 3 axis of sensing hence the 9 degrees. I don't think these receivers need the compass, so I assume they are using 6 DoF which contain accelerometers and gyros. Here is a data sheet from one and you will notice, they are very clear about output values in relation to the "dot" on the chip.
Given the chips all know what direction they are moving why does FrSky need us to calibrate? Well, I am guessing that is so they are free to mount the chip any direction they want, and the calibration process lets them determine which way is "up". This process though a bit annoying, is a minor issue and I don't really care about having to do it.
Next, you have to mount the receiver, and this is where it starts to get annoying. Rather than letting me just tell the software the direction the receiver is facing, I am forced into a very limited set of directions. Why? This is lame. My aircraft is large and I am able to mount the receiver the way they require me to, but this certainly could pose issues for smaller aircraft.
Next, I setup my radio with a switch for channel 10, the gyro mode switch, and a knob for channel 9, gyro gain, and I ran the self check. I then enable the receiver using the SXR Enable button,
and then I check the direction of correction, and during this I notice the direction of correction is wrong for both the elevator and the ailerons. No biggie right? There is a bunch of "Inverted" switches in the settings, so I reverse both surfaces, and viola, they are now doing the proper correct. Job done right? Nope. Now both surfaces apply proper correction to motion, however when I move the sticks on my transmitter, the surfaces move the wrong direction!
Why? I cannot think of any reason, the "invert" switches cannot just change the direction of correction. At this point I was confused as I thought I had the receiver mounted wrong or ? so I spent time trying to find a picture to make sure my "pins to the rear, label up" was indeed referred to as "Level". In OpenTx I believe there is a wizard that shows pictures of the various positions, not so in Ethos. I did have it mounted correctly and at this point, I gave up and just disabled it.A couple days later, I decided to man up and solve this stupid thing. I double checked everything I had done to this point and decided I had not messed anything up. So, I then decided that since I had to invert the surfaces at the receiver level, I would reverse them on the transmitter, which should make them respond correctly to stick inputs and the motion correction would remain correct. Of course when I reverse the channel output, guess what else happens. Yep, the outputs were not at exactly 1500µs so both ailerons and the elevator need to be recentered. Arggggg. Ok no biggie, this is again another annoyance of their system but not a deal breaker. So, with stabilization in the off position, I recenter my surfaces, adjust my throws so they are once again even and correct, and then I run the self-check, which should pick up the new center and end positions. Whew, finally done and ready to fly, right?
I flip my stabilizer switch from off, to stabilize. Both ailerons move up a bit, and the elevator moved down. OMG!!!! Why? I am sure there is something I didn't do quite right, but this is the lamest thing yet. There is zero need to move controls unless the stabilizer detects motion.
I QUIT!!!!
I doubt FrSky will ever "fix" these issues, so I will never again use one of their stabilized receivers. I will buy one of the $15 stand alone units used in quads.
I really like FrSky products, I have had an X9D, the X9D 2019, an X20S, and I just got myself on the list for the X20 Pro AW. Their receivers, as receivers have always been rock solid for me, I have used a number of their Neuron ESCs, but their stabilization is...I am just going to say it....horrific! Never again.
(one final note, I have set one of these up before (and used it)).
