Guys,
I did a bench run on all 3 versions tonight.  I have 3 CC-45 ESC's and loaded the same settings into each software version.  I loaded the motor as close as a flight as possible.  After each run I measured the battery, motor, and ESC temps.  Then I measured them again to check for temp rise.  I also used my Eagle Tree datalogger for each run which I've included.

The results:

V3.13
Battery temp - 120/117
Motor temp - 122/127
ESC temp - 97
Mah used - 2422
Amps ave - 23.32
Min Volts - 13.83

V3.16
Battery temp - 126/122
Motor temp - 125/129
ESC temp - 102
Mah used - 2534
Amps ave - 24.84
Min volts - 13.80

V3.18
Battery temp - 124/121
Motor temp - 125/129
ESC temp - 110
Mah used - 2501
Amps ave - 24.67
Min volts - 13.76

Wow, not many comments.

We all agree that v3.16 was terrible and that Castle is heading in the right direction with v3.18. 

V3.18 uses 3.2% more mah than v3.13.  Not much at all. The increase in ESC temp might be attributed to me and the way I checked it.  In other words don't hold too much to the ESC temps. 

The motor temp for v3.18 was 2.4% higher than v3.13 and the battery temp was 3.33% higher. 

Might be my imagination but v3.18 seemed to run pretty smooth.  Take that for what it's worth.  IMHO v3.13 is still the one to keep, but CC is improving and getting closer with each revision. 

Right now I have 3 45's, 2 35's and 1 25.  I'll keep the one 45 at v3.13 as the "standard".  I will update all the others from v3.16 to v3.18 soon.  No sense in keeping v3.16.

Dennis,
I might be able to do some comparison flying this weekend between v3.13 and v3.18.  I'll report back what I find out.  I guess as they say in Michigan "Your mileage may vary."      Maybe the differences is the motors we are using?

OK, let's follow that train of thought - I admit I am skeptical but as of now any/all ideas are in play, and this is a s good a place as any other to start.  In the RC Groups discussions, have they surfaced any common threads regarding motor brand or construction?


Just for comparison purposes my apps include:

* Turnigy 36-42 - 1000 kv with Phoenix 35.  Run on 3.13, 3.14 (beta testers version) and 3.16.  Also run for a short time on ICE-50 strictly for testng using 3.14.  No differences noted.

* Rimfire 36-35 - 1200 kV with Phoenix 35 & 4S.  Has been run on 3.13 & 3.16 with no differences observed.

* Arrowind 2820-920kv with Phx 35, 2832-960kv with Phx 45, & 3526-685kV on 6S with Phx 60.

* I use the same basic parameters set-up, ALL run with governor gain set to "10" (midway between the Low and Medium gains) and head speed rate set to Medium.  All the Arrowind apps have only used Version 3.16, and all run very well. 

* Have also run the Brodak motor  whcih is 1500 kV on 3S and version 3.13.  Again this motor is off the beaten path (10 poles instead of the usual 14) so it has a place in the common software discussion.  It is now set-up to fly on 3.16 but has not been flwon this way yet.

* As a side note, I even ran the gear drive with the 4-pole inrunner on version 3.13, and while its career was short (about 10 flights) the common CC Software  it seemed to handle it just fine.  I only mention it because THAT motor combo was as far removed from an outrunner as you can get.

One consistant element in my apps is that the kV is a little tall, and throttle setting is thus low.   Perhaps WHERE we operate in the throttle range is a factor?    However the 3526 (in the Strega) has run well from 8000 RPM with a 13x6.5 to 11,000 RPM on a 13x4.  That is a pretty broad throttle range.


Crist: are your comparisons all with Scorpions?  Same motor?


I still think we're missing something in the set-up parameters, however Crist's excellent test series had not "bubbled up" anything there yet...

 

Dennis,
Are you seeing any issues with your setups? I can't tell exactly from your last comments.

And for Crist,
I am curious about your governor gain.I know it is hard to compare 2.xx settings vs the 3.xx settings since I think they may have changed the scale, but I am wondering if putting the gain right up to the max (assuming that is what you did) might conceivably cause the ESC to work somewhat harder in keeping the rpm "perfectly" constant. I have no evidence that it does, but I know one claim to fame from CC was that they made the feedback much stronger in the 3.xx version that what existed in the 2.xx.

Just throwing out possibilities!

I agree, no Fox for me!

Well, maybe a few more tries  . Why does this feel so good?

I am almost totally convinced now that the constant rpm mode in electric is superior to any glow engine--at least in providing thrust. The only think I would like to get measured would be a piped run to see how constant those runs really are. I tried "pirating" a You-Tube video of David Fitzgerald flying the pattern with a piped setup back in late 90's, and I was actually able to analyze the sound ok, but the Doppler Effect made it hopeless to get anything quantitative out of the sound.

Crist,
My comment really meant to use enough gain so that you can't measure an rpm drop. Measure (up to now at least) meant with the Eagletree.

So instead of driving it to the point just before it begins to oscillate, try putting just past the point where you can't detect any rpm drop.

Now I am not sure how close or far away from each other those two points are.

I was just curious.

So one more question---what version did you (Crist) use during last year's contests? I am guessing you were nominally satisfied with that performance.