Servo movements on bench
In any case I recommend to set Stick Deadband at least at 4.
Spirit System developer
I wanted a channel so that I can assign it to fine tune parameters such as sensor gains as example in flight.ZeXx86 wrote:Thank you very much for very interesting finding.
From the FBUS processing point of view, there is no difference between 8 - 16 channels.
For what purpose are you using this channel?
But in my case it’s reproduced even when nothing is configured on the spirit to make use of ch12 it’s only setup as a new channel on the transmitter with a slider via a free mix.
Steps to reproduce
Create a new freemix on ethos
Input source as any free slider or pot
Output channel set to 12
Move the slider so that the ch12 valve is around +5% and move the pitch smoothly. I experience a nonlinear pitch servo movement.
Hello Senaka,
could that be maybe a problem from the mixer in ETHOS ?
I don't know how your mixer look like - maybe you can show some photos of the configured mixer ?
If there is no problem it could also be a problem of ETHOS itself ? That would be not the first time...
In this case I would set this problem in your place to github maybe there are some others where this error occurs too or some developer can say something about this problem -
Best regards
Matthias
could that be maybe a problem from the mixer in ETHOS ?
I don't know how your mixer look like - maybe you can show some photos of the configured mixer ?
If there is no problem it could also be a problem of ETHOS itself ? That would be not the first time...
In this case I would set this problem in your place to github maybe there are some others where this error occurs too or some developer can say something about this problem -
Best regards
Matthias
Hi Mattes61
The mix is as recorded here in the video link
https://youtu.be/GNIQAGwxrkE?feature=shared
I have narrowed down a bit further and found out that this issue isn’t reproducible using sbus protocol. I have reached out to Frsky and created the following for ethos
https://github.com/FrSkyRC/ETHOS-Feedba ... ssues/5522
The mix is as recorded here in the video link
https://youtu.be/GNIQAGwxrkE?feature=shared
I have narrowed down a bit further and found out that this issue isn’t reproducible using sbus protocol. I have reached out to Frsky and created the following for ethos
https://github.com/FrSkyRC/ETHOS-Feedba ... ssues/5522
Hello Senaka,
I have looked in github and saw that one reported this problem, I think you are this one (dehigama) -
in the changelog Tomas also said that the FBUS protokoll/Handle is changed see here from the hardcopy, maybe here could be also a problem ?
Best Regards
Matthias
I have looked in github and saw that one reported this problem, I think you are this one (dehigama) -
in the changelog Tomas also said that the FBUS protokoll/Handle is changed see here from the hardcopy, maybe here could be also a problem ?
Matthias
Hi Mattes61,
I don't think it's an issue which got introduced on spirit version 4.0.0 as I rolled back to version 3.7.2 and the issue persisted. Im wondering if I can capture the logs of the pitch channel and the 3 swash servo movements in the logs. so that maybe I can see if it's before it gets processed (received pitch output value) or after, by inspecting the servo output values(output values controlled by Spirit), and enter the values in a spreadsheet to and check for a saw toothed curve in the graph.
I don't think it's an issue which got introduced on spirit version 4.0.0 as I rolled back to version 3.7.2 and the issue persisted. Im wondering if I can capture the logs of the pitch channel and the 3 swash servo movements in the logs. so that maybe I can see if it's before it gets processed (received pitch output value) or after, by inspecting the servo output values(output values controlled by Spirit), and enter the values in a spreadsheet to and check for a saw toothed curve in the graph.
Hello Senaka,
in your video I do not understand what are you doing with the free-mix in conjunction with the pitchchannel it makes for me no sense in this test - If you want to to fine tune a gain I understand that, so you have a possibility to change to the best value of the gain for example the gain for the rudder - I have done this also but I have limited the area from -5 to +10 and also made a mix to add this value to get the optimal rudder gain -
I'm familiar with the behavior that the servo supposedly exhibits from very fast servos, and I'd already researched what information there was on the subject. What I ultimately found was that this only occurs with very fast servos. I'd already noticed this in setups from a few years ago that were configured only using Spektrum satellites. I don't know if that helps, but I just wanted to mention it. Perhaps something else will come to light. Let's wait and see what happens.
Best regards
Matthias
in your video I do not understand what are you doing with the free-mix in conjunction with the pitchchannel it makes for me no sense in this test - If you want to to fine tune a gain I understand that, so you have a possibility to change to the best value of the gain for example the gain for the rudder - I have done this also but I have limited the area from -5 to +10 and also made a mix to add this value to get the optimal rudder gain -
I'm familiar with the behavior that the servo supposedly exhibits from very fast servos, and I'd already researched what information there was on the subject. What I ultimately found was that this only occurs with very fast servos. I'd already noticed this in setups from a few years ago that were configured only using Spektrum satellites. I don't know if that helps, but I just wanted to mention it. Perhaps something else will come to light. Let's wait and see what happens.
Best regards
Matthias
Hi Mattes61,
The freemix which has CH12 has no association with the pitch channel, this is what for me is very strange why it's causing the swash servos to move non-linear.
The other freemix I created is to move the pitch channel via a 3 way switch more smoothly at a consistent rate using the slow feature set to 0.5 seconds for both up and down travel, than me moving the pitch stick by hand , this was just to check and record the visual outputs on the spirit software diagnostics page, which now I have deleted it.
Thanks for sharing the info, but what I'm thinking is if is due to fast servos, why does the issue only occur when CH12 is at +5%. also CH12 in this test case was not even configured in the spirit software to make use of it In the channel assignment page. It a simple freemix created with 3 clicks like as follows:
freemix-->input source=any free slider or pot --> output = CH12
to me this looks something very low level affecting in the fbus frames somehow corrupting it only when ch12 is set to +5%
The freemix which has CH12 has no association with the pitch channel, this is what for me is very strange why it's causing the swash servos to move non-linear.
The other freemix I created is to move the pitch channel via a 3 way switch more smoothly at a consistent rate using the slow feature set to 0.5 seconds for both up and down travel, than me moving the pitch stick by hand , this was just to check and record the visual outputs on the spirit software diagnostics page, which now I have deleted it.
Thanks for sharing the info, but what I'm thinking is if is due to fast servos, why does the issue only occur when CH12 is at +5%. also CH12 in this test case was not even configured in the spirit software to make use of it In the channel assignment page. It a simple freemix created with 3 clicks like as follows:
freemix-->input source=any free slider or pot --> output = CH12
to me this looks something very low level affecting in the fbus frames somehow corrupting it only when ch12 is set to +5%
Hello Senaka,
Now I understand you fully it is channel 12 without any connection to another channel -
Can this be an latency problem ?
Can you verify this behavior with only your receiver too ? (Archer plus R10)
Your receiver has some outputs put there an servo to test the behavior if its replicable -
In this way we can look if its an problem of the receiver or not -
Best regards
Matthias
Now I understand you fully it is channel 12 without any connection to another channel -
Can this be an latency problem ?
Can you verify this behavior with only your receiver too ? (Archer plus R10)
Your receiver has some outputs put there an servo to test the behavior if its replicable -
In this way we can look if its an problem of the receiver or not -
Best regards
Matthias