Comparing gov performance

[Steve Evans]

I've been doing an awful lot of work trying to tune the CGY750 gov, first using an Align RCE-BL60G ESC and more recently a YGE-120LV. I then decided to try the YGE's internal gov and much prefer it having played around with gains and motor PWM etc.

It's been hard to get a good visual comparison of how good the performance is, so I've been playing around with charting in Excel, and I think this may be a good basis for such comparison. Below are three sample flights with approximately equivalent manoeuvres being performed.

With the Align ESC using CGY Quick response and 20% gain, with the ESC PWM signal limited to 1836us to avoid instability, I got the following.



With the YGE with HeliFast response and with the CGY750 in Quick mode with 20% gain (the best combination I ever found) I got the following.



With the YGE internal gov using internal gains of P=1.6, I=0.09 I got the following.



It's clear that the CGY gov is less effective, but what really shows this is plotting the distribution curve, on the horizontal axis is RPM, and on the vertical axis is the number of times that RPM was seen. For this I took the first 34000 samples (the CGY/YGE flight was shorted due to rain!) of each flight one the desired head speed had been achieved and plotted the distribution between 2450 and 2850 RPM.

For the CGY750 gov and Align ESC.



For the CGY750 gov and YGE ESC.



And for the YGE internal gov.



Notice how much cleaner and narrower the YGE curve is, demonstrating that the RPM was close to the set speed for a larger proportion of the time.

Thought folks might find it a useful comparison. The top trace for the Align ESC looks better than the YGE gov using the CGY750's gov, but the heli felt less powerful, rather proving the point that useful as such charts may be, what really matters is how well the help flies! It should also be noted that the above doesn't show the head speed response to a heavy load such as a pitch pump for which the CGY/YGE combination was pretty grim with lots of over/undershoot (see HERE).

Steve

[Long Nguyen]

Very nice information!

What is the sampling rate? And what is the logging tool?

Does the GY Gov have pre-compensation?

After using Vbar Gov + YGE, YGE internal Gov, Kontronik internal Gov, I have found that both internal govs hold head speed better (smaller variance) however, the Vbar Gov + YGE gives the best feeling, and the most power; due to its pre-compensation function.

[Steve Evans]

Hi Long,

The logger, see HERE, samples at the S.Bus frame rate of 7ms (130Hz).

CGY750 has feed-forward, see HERE for examples of its operation. Without doubt the head-speed dips less under load using the CGY750 gov with some pitch and cyclic to throttle mixing, but unfortunately the instability at anything but very low gains means poor performance over all. I'm sure that the hardware is capable, so Futaba need to pull their finger out and sort out the software. Having to make up a pre-scaler lead, see HERE, to adapt the YGE phase sensor output to the CGY750 is a pain too.

Steve

[Mystic3D]

Interesting data. Is the CYG Gov best suited to Nitro?


Sent from my iPhone using Tapatalk

[Steve Evans]

The GY701 and CGY750 govs are highly regarded for nitro, but clearly need some work for electric.

Steve