Author Topic: RPM Discrepancy (Read 2764 times)

tom hampshire · « **on:** May 22, 2011, 12:09:45 PM »

Hi Dean - I have had quite a time with my SV-11. The system is the following:
Motor Scorpion SII-3026-890-V2
890 kv, 14 poles

ESC Castle ice lite 50

Timer Hubin FM9

Battery Neu 4s1p 3900 mah

The esc is set up exactly per Norm Whittle's cookbook, except for the 890kv motor. The 'test prop' is a piece of 1/16 ply 1.5 x 6, center drilled, but with no pitch. Saves having to restrain the airplane for test runs.

The problem is that the RPM does not follow the timer setting. Some data:
Set RPM Measured RPM Delta
9870 8940 930
10,020 9090 930
10,245 9300 945

The RPM is stable, but off from the timer setting. How do I approach this to get the RPM setting to match the tach?

Crist Rigotti · « **Reply #1 on:** May 22, 2011, 12:46:27 PM »

Tom,
Have you set the KV of the motor in the ESC menu to 890? Also be sure to set the gearing to 1:1. Could you post your ESC set up? You can print it out.

tom hampshire · « **Reply #2 on:** May 22, 2011, 01:55:27 PM »

Crist - I just double checked it. 890 kv and 1:1. I'm off to try and print the whole setup. Thanks!

Crist Rigotti · « **Reply #3 on:** May 22, 2011, 02:28:02 PM »

Great Tom. I'm sure we'll be able to determine what is going on. Do you have the ESC set up for 4S?

tom hampshire · « **Reply #4 on:** May 22, 2011, 03:38:50 PM »

Here's the setup file for the ESC...

Crist Rigotti · « **Reply #5 on:** May 22, 2011, 04:06:31 PM »

Tom,
Do you have the timer box set on "Phoenix new High RPM"?

tom hampshire · « **Reply #6 on:** May 22, 2011, 04:39:21 PM »

Crist - Phoenix high RPM. I have not updated for a year or so. Considered it now, but there is a current thread describing problems with the current version, so I passed. I will do it as soon as the postings describe that the problem is resolved.

Crist Rigotti · « **Reply #7 on:** May 22, 2011, 06:04:21 PM »

Are you using the tach outdoors? Using a tach indoors you'll get errors unless you are using a strong flashlight for the light source.

tom hampshire · « **Reply #8 on:** May 22, 2011, 06:26:59 PM »

Indoors, with a bright LED flashlight. I can get variation if either the flashlight or the tach are more than about a half inch from the prop disc. Aim the flashlight directly at the nose of the tach to overwhelm the flourescent shop lights. I trust the readings after they are stable for more than about 2-3 seconds, but not before.

tom hampshire · « **Reply #9 on:** May 23, 2011, 01:59:59 PM »

More trials... Changed the Hubin FM9 for another one, same results. Used a different tach, same results. So whatever the glitch is, it seems repeatable.

There is a sentence in Norm Whittle's cookbook about the desired head speed 1 - 3 indicating it will affect rpm from the controller. (This is shown in the settings printout at the bottom of the first paragraph.) I will try modifications of those settings, but suggestions are welcome. Anything I dream up will be a pure shot in the dark. Comment??

Tim Wescott · « **Reply #10 on:** May 23, 2011, 02:44:07 PM »

I've been reading this thread with interest, 'cause design control systems for a living. (Not ESC's, though). The only thing that comes to my mind is that any time you're testing a complex system and getting weird results, it's a good thing to try to make things as close to 'real' as possible.

So -- try it with a real prop on there, instead of something that loads the motor unnaturally lightly?

I can't see a rational reason for it to make much difference, but as a general rule any time you start using a component that runs off of software (like both the Hubin timer and the ESC) then weirdness needs to be taken in stride, because you're gonna see it!

tom hampshire · « **Reply #11 on:** May 23, 2011, 02:53:10 PM »

Tim - I'll try it, but you greatly underestimate my ability to create disaster any time electricity is involved. Thanks for the help.

Tim Wescott · « **Reply #12 on:** May 23, 2011, 02:55:41 PM »

Quote from: tom hampshire on May 23, 2011, 02:53:10 PM

Tim - I'll try it, but you greatly underestimate my ability to create disaster any time electricity is involved. Thanks for the help.

Well, don't chop any fingers off, and if you do smoke things, remember that the actual to commanded RPM ratio will be easy to calculate!

William DeMauro · « **Reply #13 on:** May 23, 2011, 04:32:21 PM »

Hi Tom,
That's the exact same setup that I am using in my SV 11 and will be using in my SV22. My first question to you is ,Why set RPM? That setup works absolutely fantastic in what I call "pure gov mode". I do have the setting printouts from when I was running version 2 of the software. I was using "set rpm" at that time. We don't have any major discrepancies so I think you are on target. My pwm was set for 12 which was the biggest difference that I see. What prop do you intend on using? I usually run up with a small prop to check my rpms. I like to put a small load on the motor. I use an 8x6 on that motor. I have found that my tach never matches anything anyway but I am within about 200-300 in most cases. so I usually fly and adjust to the laptimes I want as necessary. If you want I'll pull my settings files off my laptop for my SV 11 and email you the file, but those are for "pure gov" and not "set rpm" . Then all you have to do is load the settings from the file.
William

William DeMauro · « **Reply #14 on:** May 23, 2011, 04:54:02 PM »

I just re-read this and your problem might be right here
"Crist - Phoenix high RPM. I have not updated for a year or so. Considered it now, but there is a current thread describing problems with the current version, so I passed. I will do it as soon as the postings describe that the problem is resolved."
If running in "set rpm" your box and timer need to be in "set rpm" not "phx high" Again I have not used "set rpm" in over a year so I may be wrong

Dennis Toth · « **Reply #15 on:** May 23, 2011, 06:08:43 PM »

Tom,
Hope all is well. One thing I noticed was that the motor timing is set at a Custom value of 2. I think the Scorpions like a lot more from what I read, greater than 12. I have been using a value of 8. I have had about a one hundred rpm delta from the set rpm to the measured rpm.

Best, DennisT

Bob Hudak · « **Reply #16 on:** May 23, 2011, 07:01:49 PM »

Tom,
Which FM9 timer do you have? If you have the FM9V you can dial the rpm with the variable resistor(potentiometer). I have the FM9V with the +/- 500 rpm option. Works like a champ in "New Hi Gov mode".I set the timer to 11220rpm and fine tune with the dial till it reads what the timer is set at.
I plan on flying my electric Vector at Brodak's.
See ya' there!
Bob

Dean Pappas · « **Reply #17 on:** May 23, 2011, 07:23:17 PM »

Hi Gang, Hi Tom!
Wow, I get busy for a couple of days and look what I miss. Motor timing and even Kv do not affect the RPM setpoint of the timer / ESC / motor combo.
The pole count does. I don't know the Scorpion motors: somebody out there, are they 12-pole? That probably isn't it.
I would put on a prop that results in a 10 or 15 amp load, and recheck. It beats me though an old version of ESC code could be the issue.
Guys with Scorpions turning very close to the set numbers ... software revisions, please?
The FM9 takes into account the throttle curve built into the CC ESC.

The Kv and cell count are used by the ESC to calculate the lowest SET RPM levels that Castle wish to see.

I'm waiting to hear what the issue is, though always setting for ~ 110% of the desired figure is no disaster.

take care,
Dean

tom hampshire · « **Reply #18 on:** May 23, 2011, 07:33:54 PM »

Hi Will - I have always used governor mode, simply because I don't have a laptop and want to use the timer to make on the spot RPM changes at the field. I set the magnet pole setting at 14 because the motor spec sheet calls out no. of stator arms=12, no. of magnet poles =14. I guessed that the stator had to be the armature windings on the fixed part of the motor. So the magnet poles have to be the magnets in the outer rotating can? When the motor is cogged over, I clearly count 12 steps. Might this explain it? (14/12 x 9050 = 10,558). Most important, your experience seems to suggest that if the system looks staqble, it will give adequate flight performance by just tuning the RPM to give good results in terms of flight trimm, and ignore the discrepancy. Any thoughts?

Hi Dennis - My guess is that the timing wouldn't affect the RPM discrepancy. I do plan to move it up after I get things flying. One thing that might be of interest is that the front end of this airplane is open, and it has only a spinner nut the width of the prop hub. Accordingly, the cooling holes in the motor bell face directly into the airflow. There is also a big air exhaust in the cockpit, with a cutaway canopy to operate as a rear facing airscoop. Apparently the cooling is fully functional, as I couldn't detect any difference between ambient temperature and the bell temp after the first few test flights. I have an IR thermometer, and elected not to bother with it when I felt the motor on the first flight. It was a smoking flight, 3.9 sec. laps, yet the motor stayed way cool. We shall see if advancing the timing makes it run any warmer.

tom hampshire · « **Reply #19 on:** May 23, 2011, 07:39:45 PM »

Whoops, missed Dean and Bob -

Dean - See the previous post, stator arm has 12 count, magnet poles have 14 count, when I cog the motor over I count 12. I'm lost as to how the two could be different, but that's what the motor spec sheet says.

Bob - The timer is a Hubin FM9, without the pot, settings are all done with the programming box.

Crist Rigotti · « **Reply #20 on:** May 23, 2011, 08:15:05 PM »

The Scorpion is definitely a 14 pole motor.

tom hampshire · « **Reply #21 on:** May 23, 2011, 08:21:13 PM »

Just ran another test with the castle link software changing the ESC pole count to 12. As before, the tach RPM was 9090, and the set RPM on the timer was 10,020. So the change in pole count from 14 to 12 made no difference whatever. This is still with the no load test prop. I'll round up a helper tomorrow and try it with a small prop.

BTW, the castle link software is v. 3.29.0.

William DeMauro · « **Reply #22 on:** May 24, 2011, 02:05:37 AM »

Tom,
Scorpions are 14 pole like Crist says. If you are in Gov mode, Forget those set rpm numbers as they mean nothing. You RPM is set off the prog box. What does that say? Lower it there if necessary. If you still have a problem call me after 3 today. I'll walk you through it.
William

tom hampshire · « **Reply #23 on:** May 24, 2011, 06:10:52 AM »

Bill, my language was clumsy. The term set RPM in the data table in the first post refers to the RPM displayed in the Hubin programming box, not the esc mode. The term Measured RPM refers to the tach reading.

tom hampshire · « **Reply #24 on:** May 24, 2011, 07:40:36 AM »

Another test run, this time with a Zinger 9x4 mounted backwards. The tach RPM was 9090, and the RPM set on the Hubin timer was 10,020. This is identical with the RPM measured at that timer setting with a 'no load' test prop, and (in a separate run) also with the magnet pole count on the esc changed to 12. The speed was closely regulated, with no variation at all. And all of these runs have been stable and repeatable. So it looks like there is nothing else to try, and I just set the Hubin timer 930 RPM over the desired prop RPM. But it does seem that no other users have encountered an RPM discrepancy this large.

Will Hubin · « **Reply #25 on:** May 24, 2011, 10:38:09 AM »

Tom,
If you are running in the Governed High RPM mode of the Castle ESC and have Castle firmware of 3.23 or later, you should be using the FM-9 mode of "Phoenix new High" (and not "Phoenix High RPM") because Castle changed the calibration, and tach values will be lower than the FM-9 values if you use the former rather than the latter. In later versions of the FM-9 Programmer, I exchanged the "Hacker 1/2A" mode for the "Phoenix new High". I'm not sure if your programmer includes this version or not, but it is a simple chip exchange if not.
Will

tom hampshire · « **Reply #26 on:** May 24, 2011, 11:41:39 AM »

Hi Will - THANK YOU and to all the others for all of the effort. I'll send the programming box back to you. How much for the upgrade and S&H? Thanks again!

Dean Pappas · « **Reply #27 on:** May 24, 2011, 02:33:57 PM »

Quote from: Will Hubin on May 24, 2011, 10:38:09 AM

Tom,
If you are running in the Governed High RPM mode of the Castle ESC and have Castle firmware of 3.23 or later, you should be using the FM-9 mode of "Phoenix new High" (and not "Phoenix High RPM") because Castle changed the calibration, and tach values will be lower than the FM-9 values if you use the former rather than the latter. In later versions of the FM-9 Programmer, I exchanged the "Hacker 1/2A" mode for the "Phoenix new High". I'm not sure if your programmer includes this version or not, but it is a simple chip exchange if not.
Will

AHA!
The old throttle curve changed with the new rev of software trick! Sadly I had suspected this ...

later,
Dean P.

Will Hubin · « **Reply #28 on:** May 24, 2011, 09:31:16 PM »

Castle moved the helicopter governing range from its original 1.1 to 1.9 ms to 1.2 to 1.9 ms, to give helicopter pilots the 1.0 to 1.1 ms range for auto-rotatation ... and the CL governing changed as well, requiring higher throttle settings for a given RPM.

It is simple enough to swap in a new chip in the FM-9 box, since it is in a socket, and the same chip can be re-programmed. So the cost of updating is $10 (for S&H) if I do it or $5 (for the chip & S&H) if a user wants to try it (with explicit instructions -- so, aside from a bent pin, replacing it with the proper orientation shouldn't be a problem).

Will

tom hampshire · « **Reply #29 on:** May 25, 2011, 10:55:25 AM »

Hi Will - It's on its way to you. Thanks again.

Tim Stagg · « **Reply #30 on:** May 25, 2011, 01:08:29 PM »

Tom,

I have the same issue with my RPM's not being consistent with the actual, but I guess I am a little lazy because I just set the RPM on the box and then after flying a 1.5 minute lap, I bump up or down to get the desired lap time, and don't really worry about what the box says.

I guess I should be more methodical but........ I dont have a whole lot of hair left to scratch off

tom hampshire · « **Reply #31 on:** May 25, 2011, 01:13:39 PM »

Hi Tim - I think that would probably work reliably, my worry was the 930+ RPM difference between the timer setting and the tach RPM. Flew it this AM after borrowing a prog box from Bob Hunt and it worked flawlessly. Tom

n b I still have hair, but no guarantees as to the workability of the 'systems' below the thatch.

Dean Pappas · « **Reply #32 on:** May 25, 2011, 02:48:10 PM »

... below the thatch ...
I like that!
Dean

Tim Stagg · « **Reply #33 on:** May 26, 2011, 05:16:41 AM »

Too Funny