Does the governor mode allow you to change any gains?  Or did loading in the new code wipe out some old settings?  It sounds like the control loop is tuned too tight (or there's a garden variety error in there some place).

Crist,

Thanks but I'm talking Firmware here. I'm running 3.26 Software for my Castle Link program. The Firmware is downloaded into the ESC by the Software.

By the way I tried a bunch of settings changes and they did not help. That makes me suspect that the problem is in the Firmware -- this was a Phoenix 10 Amp ESC -- dialing back to the previous 3.21 made it go away.

John

Crist
I think he means 3.24.  I tried using Firmware 3.24 but my RPM dropped 900. Switched back to 3.20 and RPM returnrd to normal.
Firmware 3.24, 3.22 and 3.21 load with Software 3.26.  Firmware 3.21 is Phoenix ICE 100 specific.

Norm

Crist and Norm,

All I know is that 3.23 Beta popped up at the top of the list when I responded to the message that there was new firmware available and did I want to change it etc (don't the Castle  Link and ESC handy to recheck right now - all packed up for a trip to Northern MI). Maybe there was scroll possibility that I missed which would have allowed me to see/load higher versions.

The printout of the parameters for the current settings shows "Firmware version 3.23 Beta" at the bottom of the page (or did before I went back to 3.21).

Norm, How do you know that 3.21 is specific to the ICE 100 ESC?? I guess that fact slipped past me but it seems to work for the Phoenix 10 ESC.

John

Norm,

You are correct the version 2.23 Beta is not listed on Castle's site it just skips from 3.2 to 3.4. There may be a message in that fact. LOL Why it showed up on my computer when I opened Castle Link escapes me but there it was and it messed up my ESC. l also see on their site that 3.21 is specific for the ICE 100. When the pop up menu appears on opening Castle Link as a firmware option it is not flagged as ICE 100 specific.

Right now 3.21 is working for me and I will mess with other firmware next week when I am home and have some time.

Thanks,

John

They have a graphical explanation of changes in ver. 3.23 -- It seems that you need to recalibrate your endpoints when changing to 3.23 (I'm not sure that I ever calibrated my endpoints for any previous version, but apparently my ESC is happy with whatever I've done).  Anyway, this may explain the surging and Crist's RC plane.




http://www.castlecreations.com/downloads/cl_beta_info_071910.html?subId=696

They have a graphical explanation of changes in ver. 3.23 -- It seems that you need to recalibrate your endpoints when changing to 3.23 (I'm not sure that I ever calibrated my endpoints for any previous version, but apparently my ESC is happy with whatever I've done).  Anyway, this may explain the surging and Crist's RC plane.




http://www.castlecreations.com/downloads/cl_beta_info_071910.html?subId=696

I did recalibrate my end points. 

Crist
How do you set your endpoints?  I don't have a clue.
Norm

Here's a link on how to do it.  Remember this is for my R/C airplane and not my C/L airplane.

http://www.rcgroups.com/forums/showpost.php?p=15036064&postcount=1

I contacted CC and they are aware of the problem plus others.  A new version will be posted within a week......that's what they said.

Castle Creations software release notes on their Version 3.23 and above note that the governed range becomes from 20% to 90% of full throttle rather than the 10% to 90% of version 3.20 and before. (10% to 20% is now labeled as the "auto-rotate range"--not something we need!) One consequence is that (as various flyers have noted) the FM-9 Phoenix High RPM calibration doesn't give RPMs very close to the tach values any more. Another consequence is that the curve of RPM vs. throttle gets steeper in the range of RPMs we usually use, resulting larger RPM increments between adjacent settings (corresponding to 1/2 of 1% throttle changes).

(All R/C throttle channels of all R/C receivers may not give the full 0% to 100% of full throttle [i.e., 1.0 ms to 2.0 ms pulse width] that our timers easily give, at least partly because they allow a "thumbnail" adjustment/fine tuning. So apparently some helicopter pilots couldn't get the low end of their throttle channel output to 1.0 ms.)

I just made some test stand runs of an Eflite 32 motor with the ICE 50, using both versions 3.20 and 3.24. Sure enough, the motor starts to turn (at a little over 7000 RPM) at 10% throttle with 3.20 but only at 20% throttle with version 3.24.

The 3.20 version of the ICE 50 High RPM calibration with the Eflite is pretty close (usually within 60 or 90 RPM, or pretty much within error limits of measurement) to the Phoenix High RPM now in the FM-9 Programmer, so that is currently usable.

I'm planning to replace the Hacker 1/2A calibration in the Programmer with the 3.24 ICE/Eflite 32 calibration this weekend, and also check the v3.23ff firmware calibration for the ICE/Axi in the near future--which I'm hoping is similar. (The firmware in the FM-9 Programmer is pretty easy to update, if necessary, because all but the first ten or so have the microcontroller in a socket.) The latest firmware also includes 1/2 of 1% spacing when used in either the pure throttle mode or in the compensated throttle mode, so that can always be used to establish one own's RPM calibration.
.