The video stoped at a sudden when I still would like to watch though. Thank you for developing such great tools and contents. It’s really helpful to be analytical when you wanna improve flight performance of a quad when just “feel” can’t provide such valuable info.
Yes it’s definitely more meaningful when we think in degs at specific freqs instead f ms. 2-3ms doesn’t carry much meaning to people but when you can see this causing 30-40deg phase shifts then the issue is clear. Got thinking about this more through conversation with Chris Thompson.
Thanks for the informative videos (and the great software). In terms of the order in which the components of latency are shown, wouldn’t the latency attributed to “rates” (i.e. thumb movement) come first, since your thumbs move the sticks (and rates are computed) before the RX latency comes into play? Similarly, wouldn’t the filter latency occur before the setpoint delay, since the quad is only determining how to move relative to the filtered gyro data? I suppose the ordering doesn’t really matter if the point is to enumerate the various sources of latency and discuss their relative contributions, but I still wonder about the sequencing.
Yes that’s right, the order here is not exactly meaningful. Though rx latency is certainly before since set point (rates) are computed from that signal after it arrives in the FC right
@@PIDtoolbox Sure I watched it, but I must admit I skipped the first parts about rc link and stuff. Not much new for me… I have a really tricky 10 inch and after having a bad fc and a noisy esc I might still be running to high filtering (had massive freak outs and this is no fun with 10inch and 2,5kg). And this is my main thing: how much filtering is really necessary. For me I guess I use too much. There was one video by uavtech where he tuned a 3inch and it looked so noisy at first but it was flying fine. But yes how much filtering is really necessary. Sadly I was not able to fly for a long time, so I still have to try out what I learned from all you guys.
@@jas-FPV depends on if you have the bosch gyro (radix), which has less internal filtering. Also a 10inch is mainly for cruising so more filtering wont kill you too much, but opt for more dterm filtering over gyro LPFs. also maximize RPM and DNs so you dont need as much gyro LPF filtering. the gyro LPFs are the real killer
@@PIDtoolbox Yeah. I got misled by all the hardware issues I had with it. Iflight dual, both gyros bad, replacement fc also. Iflight 60amp esc noisy as hell…. This leads to wrong decisions regarding filtering 🤷🏼♂️ Yes I will restart the tuning with gyro at 2.0 and go from there.
@@PIDtoolbox Thanks! I enjoyed the discussion. Definitely could have kept watching. I'm no programmer but it seems like code could be implemented that in propwash the FC could calculate the phase delay on the fly and advance the P freq, sorta like a feedforward for P thus reducing the phase delay.
The video stoped at a sudden when I still would like to watch though. Thank you for developing such great tools and contents. It’s really helpful to be analytical when you wanna improve flight performance of a quad when just “feel” can’t provide such valuable info.
Great stuff Brian. LOVE THE NEW VERSION!!
thanks Mark!
Nice connection explanation of filter delay and phase delay man!!
Yes it’s definitely more meaningful when we think in degs at specific freqs instead f ms. 2-3ms doesn’t carry much meaning to people but when you can see this causing 30-40deg phase shifts then the issue is clear. Got thinking about this more through conversation with Chris Thompson.
It‘s nice to have filter latency as low as possible but I find a proper PID tune along with FF makes like 80% of a good tune.
7:50. man, this is beautiful.
all feel is just horizon delay as a function of time.
exactly!
Thanks for the informative videos (and the great software).
In terms of the order in which the components of latency are shown, wouldn’t the latency attributed to “rates” (i.e. thumb movement) come first, since your thumbs move the sticks (and rates are computed) before the RX latency comes into play? Similarly, wouldn’t the filter latency occur before the setpoint delay, since the quad is only determining how to move relative to the filtered gyro data?
I suppose the ordering doesn’t really matter if the point is to enumerate the various sources of latency and discuss their relative contributions, but I still wonder about the sequencing.
Yes that’s right, the order here is not exactly meaningful. Though rx latency is certainly before since set point (rates) are computed from that signal after it arrives in the FC right
@@PIDtoolbox but a rc curve (thus a rate) could be set. maybe that was his idea
Thanks.
Great !
Hey, it just got interesting and video stops 🙈
yea, unfortunately nothing more interesting happened lol. but Im surprised you watched till the end. I was afraid it was getting sleepy lol
@@PIDtoolbox
Sure I watched it, but I must admit I skipped the first parts about rc link and stuff. Not much new for me…
I have a really tricky 10 inch and after having a bad fc and a noisy esc I might still be running to high filtering (had massive freak outs and this is no fun with 10inch and 2,5kg). And this is my main thing: how much filtering is really necessary. For me I guess I use too much. There was one video by uavtech where he tuned a 3inch and it looked so noisy at first but it was flying fine.
But yes how much filtering is really necessary. Sadly I was not able to fly for a long time, so I still have to try out what I learned from all you guys.
@@jas-FPV depends on if you have the bosch gyro (radix), which has less internal filtering. Also a 10inch is mainly for cruising so more filtering wont kill you too much, but opt for more dterm filtering over gyro LPFs. also maximize RPM and DNs so you dont need as much gyro LPF filtering. the gyro LPFs are the real killer
@@PIDtoolbox
Yeah. I got misled by all the hardware issues I had with it. Iflight dual, both gyros bad, replacement fc also. Iflight 60amp esc noisy as hell…. This leads to wrong decisions regarding filtering 🤷🏼♂️
Yes I will restart the tuning with gyro at 2.0 and go from there.
Frustratingly I can't get the new version to run :( Splash screen into nothing.
Sorry to hear that. PM me and I’ll do my best to help.
@@PIDtoolbox Thanks! I enjoyed the discussion. Definitely could have kept watching. I'm no programmer but it seems like code could be implemented that in propwash the FC could calculate the phase delay on the fly and advance the P freq, sorta like a feedforward for P thus reducing the phase delay.