News:



  • March 28, 2024, 05:08:50 AM

Login with username, password and session length

Author Topic: Instrumented motor comparison test for .60-size models  (Read 1085 times)

Offline Mike Alimov

  • Trade Count: (0)
  • Commander
  • ****
  • Posts: 379
Instrumented motor comparison test for .60-size models
« on: November 27, 2018, 03:46:13 PM »
I have followed a few recent discussions on the optimal motor selection (which left quite a few questions in my mind), and faced with a need to design an electric system for next year, I decided to perform an instrumented comparison test (i.e.: where multiple parameters are measured, recorded and analyzed. No guessing or speculation, or assumptions). 
  NOTE 1: it is NOT my purpose to start a heated discussion on the validity of my test methodology, my conclusions from it, or how my results correlate with your own choices of a power system. I am sharing the results of these tests here for the benefit of those that are at the same point in the learning curve as I am, and to save others time and energy in seeking the same answers. 
NOTE 2: My focus was on motor selection. I will address ESC and battery selection very briefly, as it was not my focus.
NOTE 3: This is not intended as a complete tutorial on control line electric systems.  There are several threads out there, and other resources.  If you are not familiar with certain terns, such as Kv or mAh, seek info elsewhere.
NOTE 4: No animals were hurt in the course of these experiments.

The purpose of the test was to select the most optimal motor for a conventional .60-size model (~680 in^2 wing, ~62 oz weight, ~ 68 ft lines, ~5.1" lap time).  There are several considerations that go into my definition of "optimal", roughly in the order of importance:
1) Energy efficiency (i.e. minimal losses when converting electricity into mechanical energy).  Because this determines how big a battery my plane needs to haul, and batteries are heavy!
2) Motor's own weight and weight distribution.  Because why haul unnecessary weight?  Also, larger diameter motors will create more gyro effect, which is undesirable, so I would prefer "slimmer" motors all else being equal (i.e smaller outer diameter).
3) Construction and materials quality (esp. bearings).  Because I want the motors to run reliably and require minimal maintenance or rebuild intervals.
4) Price  - because it matters! (To most of us).

The questions I was particularly curious about:
1) Are bigger, heavier motors that much more efficient that carrying an extra 1-2 oz in the motor saves me more than that in the battery weight?
2) Are more expensive motors more efficient than the cheaper motors of the same weight to make appreciable difference?
3) Side test: is a 6S setup more energy efficient than a 5S setup, due to lower average current?

Methodology:
- First, I determined the approximate propeller size that I want to be using.  Something like APC 12x6 EP.
- The propeller pitch and RPM determine airplane speed.  I'm guessing RPM in the 9000-9500 range.
- The target RPM determined that I'm looking for motors in the 650-780 kV range.  This narrowed the choice of motors.  I had a few, and borrowed a few.
- Based on flight data (Castle Edge Lite 50 data log) on a Top Flite Score (680 in^2, 64 oz), I figured approximate average power required ~ 460 W.  Note this is averaged over the entire flight, and during actual gflight will vary up and down by as much as +/-80%, but both the motor and ESC can handle short spikes.  They can't handle sustained overload, however.
- Based on the 460W power level, I determined that an APC 10x6E prop at 10400 RPM on a test stand represents an equivalent load to a APC 12x6E in flight.  So I used it as a motor load in all my tests. 
- All other variables were controlled for: same atmospheric conditions (room temp 25C, humidity, altitude, etc), same ESC (Castle Edge Lite 75), same timer (FM-9),  same battery (Zippy Compact 2700 5S), same RPM setting. *(Note: the Plettenberg Orbit did not want to run in a Governed mode, so I had to use Set RPM mode.  As a result, the actual RPM were about 100 less than on other motors).
- Each motor was mounted on the test stand, fresh battery used every time; data was recoded for a 3-min run and downloaded from the ESC through Castle link.  I chose 3 min to be easy on the battery and save test time.  Later I extrapolate the battery consumption to the full 5'30" flight time.
- I also checked motor case temperature after each run using laser thermometer. It gave me an idea of how much energy got converted to heat (waste) instead of mechanical energy (RPM), and how thermally stressed the motor was.

The results are attached.  Seven motors were tested, ranging in weight from 6 oz to almost 8.5 oz, including some popular choices.  (Look at the 7 motors with actual test data. Other motors's stats were entered for comparison purposes).

Conclusions:
- The larger, heavier motors were indeed more efficient... but... not enough to matter.  Why? Because, even though they required 100-200 mAh less of a battery (5S), I can't take advantage of that, because I can't find 2500 mAh 5S batteries on the market!  I can find 2700-2800 mAh 5S packs, but that covers all motors in the 6-8.5 oz range!  And since I can't cut off a piece of battery to make it 100 mAh smaller and 1 oz lighter, I can't take advantage of this difference.  As a result, I vote for the 6 oz motors.
- The more expensive, "elite" brand motors like Plettenberg, Scoprion and AXi were slightly more efficient than, say, the 6-oz Turnigy, which got hotter and would need every bit of 2700 mAh to complete a pattern on a normal day.  This does mean, though, that if I need a spare 200 mAh in the pack for a faster flight on a windy day, a better motor will allow me that, while a Turnigy does not leave that margin, requiring to step up to a bigger heavier battery to leave some spare "fuel in the tank".  Vote for the more expensive motors.
- I also ran a test of 5S vs. 6S setup using the Cobra 2826-12 motor (everything else same - prop, ESC, etc).  Strangely, the 5S setup consumed 21.46 Watt*hr for the 3 min run, while the 6S setup consumed 23.3 Watt*hr.  This is contrary to what I expected.  The only explanation I  have is that due to higher voltage and fixed given Kv rating of the motor, the ESC had to throttle down so much that the overall setup became less efficient.  Since this was only one test and not the main focus of the experiment, I would not give this too much weight, other than to conclude that there was not enough convincing evidence for me to switch to 6S at this point.  There are very few 6S 2200 mAh batteries on the market to choose from, and they are not any lighter than the 5S 2700 mAh batteries. No tangible gain to be had here.
- This concludes this series of experiments.  I've narrowed my choice down to either Cobra 2826-12 or AXi 2826/13 with a 5S 2700 mAh battery.
- Other areas to study would be: propeller efficiency (what are the best props on the market for electrics) - my preliminary take is that APC is great for starters, but not good enough for Experts; ESC efficiency (is 75A ESC an overkill for the 23A average current?) - I think so; battery choice (is ThunderPower worth the price, or is Turnigy / Zippy a great deal for the price - my internal resistance measurements show that Turnigy and Zippy have lower R than more expensive brands). 

Offline Tim Wescott

  • 2016 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 12804
Re: Instrumented motor comparison test for .60-size models
« Reply #1 on: November 27, 2018, 06:09:46 PM »
If I understand your spreadsheet correctly, you're pulling 2700mAh out of a 2700mAh pack.  Do that, and you're not going to have very much battery life.  Sucking the dregs out of a LiPo cell significantly reduces the number of charge/discharge cycles it'll last through.  The rule of thumb is to use 70 to 80% of the available charge, no more.  So you want to use something like 3600mAh 5S packs.

And yes, if the ESCs of the world work like the motor drivers that I've designed, they'll be slightly less efficient with 6S -- but you'll have lots of headroom on your voltage, for when you need extra power.  I suspect this means more if you're running a Burger or a Fioretti timer, because those vary the speed where the Hubin and Renicle timers don't.
AMA 64232

The problem with electric is that once you get the smoke generator and sound system installed, the plane is too heavy.

Offline Mike Alimov

  • Trade Count: (0)
  • Commander
  • ****
  • Posts: 379
Re: Instrumented motor comparison test for .60-size models
« Reply #2 on: November 27, 2018, 06:44:24 PM »
Tim, you did not read my spreadsheet correctly.  During the instrumented tests, I was actually using around 1100 mAh for a 3 minute run (column J).  Then I projected this usage into 5'30" flight time using the 80% capacity formula, arriving at the total needed battery capacity in column L.  Looks like all motors I tested will be able to use the standard 2700 mAh 5S packs.
I totally agree with your comment about having the voltage headroom.  Which explains why those running active timers are migrating to 6S packs, while the passive timers are humming on 5S just fine.  According to a this research paper
https://apps.dtic.mil/dtic/tr/fulltext/u2/a577582.pdf
 it is more efficient to use brushless motors with the ESC at full throttle, and adjust supply voltage to achieve desired output, instead of having the ESC and throttle it down (which is what happens with ~700 Kv motors running 6S in stunt planes - the ESC really chokes it down until it's time to climb uphill).

Offline Tim Wescott

  • 2016 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 12804
Re: Instrumented motor comparison test for .60-size models
« Reply #3 on: November 27, 2018, 07:10:34 PM »
Tim, you did not read my spreadsheet correctly.

I can believe that!  Thanks for de-confusing me.
AMA 64232

The problem with electric is that once you get the smoke generator and sound system installed, the plane is too heavy.

Offline Crist Rigotti

  • 23 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 3855
  • Electric - The future of Old Time Stunt
Re: Instrumented motor comparison test for .60-size models
« Reply #4 on: November 27, 2018, 07:58:48 PM »
Mike,
Thanks for taking the the time and effort to conduct these trials and posting your results.

Here in Texas, several of us have migrated from the 28 series to the 35 series Cobra motors because the 35 series motors run a lot cooler.  When the circle tarmac is 140 degrees F. we need all the help we can to keep our equipment running as "cool" as possible.
Crist
AMA 482497
Waxahachie, TX
Electric - The Future of Old Time Stunt

Offline Mike Alimov

  • Trade Count: (0)
  • Commander
  • ****
  • Posts: 379
Re: Instrumented motor comparison test for .60-size models
« Reply #5 on: November 28, 2018, 09:01:24 AM »
Well Christ, they say everything is bigger in Texas...  I guess even stunt motors!   LL~
Wanna move to the North Coast?  We don't fly on 140 degree tarmac - we don't even have those days; but even if we do, we sit at home and sip iced tea.

But seriously, yes, I've heard others voice the heat issue.  What I'm curious whether anybody has tried to quantify it.  So what if the motors get hot (assuming they don't melt / smoke / demagnetize)? You wouldn't put your hand over that piped PA right after landing, would ya?

Now, I'm too lazy to set everything up and re-run my tests at, say, 100F ambient temp (not sure how I'd do that - blow a hair dryer air into the spinning prop? It's snowing outside at the moment...)  But I ran some numbers.  For example, Cobra 2826-12 has Rm (resistance per phase) of 68 mOhm at room temp (20C) --[although my own measurements show 60 mOhm, but we'll stick with the official data sheet].  At 23A average current, the copper losses (I^2*R) come out to be around 36 W.  With the resistive temperature coefficient of copper being ~ 0.393 per degree C, at 80C we have Rm=84 mOhm, and copper losses come out at ~44.5 Watt, which is now about 10% of average in-flight power consumed. Same calc for the porky Cobra 3520/14 gives me ~24 W at room temp and ~31 W at 80C.  Over the typical ~6 min flight time, this translates into the smaller Cobra using up about 72 mAh (5S) of battery capacity more than the chunky one; let's even be generous and say another 100 mAh.  Is that what this is about, or is there more to that?  Other than the increased battery consumption, does the motor heating manifest itself in any other negative way?  Have you tried to download and analyze the ESC data?
This is exactly the kind of discussion I wanted to have, so please share your observations.  I'm not trying to prove that I'm right or anybody's wrong; in the end, if there's enough convincing data, I might go and buy me a couple of the chunky motors.  But first I need to understand why I'm doing that.

Offline Tim Wescott

  • 2016 supporter
  • Trade Count: (0)
  • Admiral
  • *
  • Posts: 12804
Re: Instrumented motor comparison test for .60-size models
« Reply #6 on: November 28, 2018, 09:42:02 AM »
Bigger motors not only have lower winding resistance, they also have more surface area.  So they tend to run cooler.  They also have more cogging torque, iron losses, and bearing friction, so if you get too enthusiastic with going to a big motor you hit diminishing returns -- it's likely that you'll be shooting yourself in the foot with weight long before that hits you.
AMA 64232

The problem with electric is that once you get the smoke generator and sound system installed, the plane is too heavy.


Advertise Here
Tags:
 


Advertise Here