27-06-2023, 12:04 PM
This flight log is from the Slick, a 90cm span 3d plane, with a matek R24 P6 rx and a Radiomaster T16 tx with an elrs internal module. All the flight data appears in one plot because I adjust each data item. Measurements are taken every 0.5 sec. The main things of interest are the blue RSS (receiver signal strength), running between -30 and -45 (which means -60 and -90, a normal range), the red receiver quality (RQly) running at 100% except for a few dips below 90%, and the pink receiver signal to noise (RSNR).
At 50 sec into the flight, multiple events occur:
- a blender manoeuvre enters its vertical dive and spin
- the throttle (THR) at 50 sec went from full throttle (10) to zero (-15) which caused the plane to flip from vertical climb to vertical dive
- the RSNR fell to 0 for less than half a second
- the transmitter power (TPWR) briefly increased from 10 mw to 25 mw
-the RQly fell to 84% and the RSS dipped slightly below -90.
The TPWR increased because the elrs was set to "dynamic power" whereby if the rx perceives a signal problem, it increases the tx signal power from 10 mw to 25 mw as happened here, or 50 mw or 100 mw if 25 does not do the job. The decision to boost tx power is based on the RQly (the same as valid frame rate) which measures the percentage of packets of data that are valid, and also the RSNR.
https://www.expresslrs.org/software/dyna...mit-power/
It appears the main cause of the TPWR jump was the transient drop of RSNR to zero, at a critical point in the blender. The RQly and RSS also fell, but not to an abnormal level.
Why did the RSNR fall? I have several planes using R24 P6 rxs, and I only get significant RSNR falls when doing the blender manoeuvre, and only in around 20% of flights. One theory is my use of a sudden cut from full- to zero throttle to flip the plane (from vertical climb to vertical dive) causes a transient interference. A fall in rx voltage is ruled out as a cause.
The tx power jump demonstrates the elrs doing its job to maintain a safe signal, and no control loss was experienced. Matek rxs require the user to solder the pwm pins to the rx circuit board, and in future I am likely to move to Radiomaster elrs rxs. I was a bit disappointed with the RM ER5A rx, and I hope the soon-to-arrive RM ER6 series of pwm rxs will be an improvement.
At 50 sec into the flight, multiple events occur:
- a blender manoeuvre enters its vertical dive and spin
- the throttle (THR) at 50 sec went from full throttle (10) to zero (-15) which caused the plane to flip from vertical climb to vertical dive
- the RSNR fell to 0 for less than half a second
- the transmitter power (TPWR) briefly increased from 10 mw to 25 mw
-the RQly fell to 84% and the RSS dipped slightly below -90.
The TPWR increased because the elrs was set to "dynamic power" whereby if the rx perceives a signal problem, it increases the tx signal power from 10 mw to 25 mw as happened here, or 50 mw or 100 mw if 25 does not do the job. The decision to boost tx power is based on the RQly (the same as valid frame rate) which measures the percentage of packets of data that are valid, and also the RSNR.
https://www.expresslrs.org/software/dyna...mit-power/
It appears the main cause of the TPWR jump was the transient drop of RSNR to zero, at a critical point in the blender. The RQly and RSS also fell, but not to an abnormal level.
Why did the RSNR fall? I have several planes using R24 P6 rxs, and I only get significant RSNR falls when doing the blender manoeuvre, and only in around 20% of flights. One theory is my use of a sudden cut from full- to zero throttle to flip the plane (from vertical climb to vertical dive) causes a transient interference. A fall in rx voltage is ruled out as a cause.
The tx power jump demonstrates the elrs doing its job to maintain a safe signal, and no control loss was experienced. Matek rxs require the user to solder the pwm pins to the rx circuit board, and in future I am likely to move to Radiomaster elrs rxs. I was a bit disappointed with the RM ER5A rx, and I hope the soon-to-arrive RM ER6 series of pwm rxs will be an improvement.