EDF Balancer - Page 75 - RC Groups
View Poll Results: What applications would you use balancing for?
helicopters 2 11.76%
Turbines 4 23.53%
Electric Ducted Fans 16 94.12%
Drones 2 11.76%
Aircraft Props 7 41.18%
Multiple Choice Poll. Voters: 17. You may not vote on this poll

Thread Tools
Apr 03, 2017, 12:49 AM
Good vibrations.
dynexhobby's Avatar

Multi-blade analysis


Hey All

Apologies for the late replay I didn't know there was activity on the forum

The basic formula for imbalance is F = m . (w^2) . r

where

F = Force blade imparts on assembly
M= mass of blade
w = rational velocity
r = radius of blade CG from axis of rotation

Now if you add up all "F" values (as vectors) and they equal zero then your system is statically balanced (in theory). In the formula "w" is constant for all blades hence you can knock that out of the formula.

I created a Excel sheet that illustrates the "F" term and plotted it in a polar format. I assumed that the CG was constant for each blade, which I know is not really true because it varies from blade to blade. The Excel sheet also calculates the residual imbalance based on blades weights and CG locations.

You can change the blade locations until your residual imbalance gets very small.... and yes many permutations and combinations.

But my question is, why would you want to go to this effort? Assuming the blades are strong enough and are not flexing a great deal, the dynamic balance method should take care of all this. In fact, there are some dynamic effects that purely weighing blades cannot not deal with so the ultimate method is to dynamically balance.

Anyway, have fun and feel free to double check the numbers as this is only a quick 5 min calc.
Sign up now
to remove ads between posts
Apr 03, 2017, 01:10 PM
Multifaceted Heli Nut
heli-headcase's Avatar
Quote:
Originally Posted by PeterVRC
You might need to try 'rotating' the table of values to all 12 positions. To find the truly worst case.
Shifting around blades within the '6 group' would alter things.
Also if you had a 'polar' (?) chart of all 12 blade directions, I suspect it would not be a circle.... it would have 'lobes' and some further out than others.
You could have equality on a line through 1&7, but poor balance match out through 4&10. Grouping only 'halves' does not account for this possibility.

Now I am getting a headache thinking if the 'Cosine maths' (totals of all 11 other blades referenced to one) I was using has any issue like that also.
Seeing it adds up all 12 'vectors' (via Cosine of their angles) I think it is valid.
Okay, I'm back! Long weekend away from email...

You're very correct assuming the polar chart would have lobes in the circle, that's expected. The values of the lobes aren't important as long as one lobe or group of lobes has equal area 'under it' to those same lobes on the opposing side of the circle. And to perform my test of 180 groups of six each properly (for one arrangement of blades) I'd have to do the comparison maths six times. It's a good check because it could catch a gross error in time. But let's move on from here.
Apr 03, 2017, 01:14 PM
Multifaceted Heli Nut
heli-headcase's Avatar
Quote:
Originally Posted by KarlHeinz57
The code tests all possible permutations. Of course, each permutation wasn't clocked as this would not yield any different result. Hence, the first blade/mass is fixed at the zero rotational angle position. This reduces the number of permutations by a factor of 12.

Each permutation's test result is computed by determining the X and Y components of each blade's mass (or vector if you like), taking into account it's angle relative to the first blade (at zero degrees), summing the X and Y components of all the blades, and then computing the net "distance" from center via Pythagorean's theorem.

You could think of each blade's mass as a length vector with all vector's meeting at (0,0) and the vector's direction defined by the blade's position. In essence, you'd have a "star" with 12 arms, each positioned at 30 degrees relative to each other, where each arm's length corresponds to the blade's mass.

The permutation with the lowest "distance" is deemed the best answer.

This method makes me want to purchase a few ChangeSun 70mm EDFs and give this a try. While I certainly appreciate having my DynexHobby balancer, the thought of being able to balance a multi-blade impeller simply by rearranging the order of the blades is very appealing.

I also have a four-blade P-51 propeller with separate blades that would benefit from this methodology.
Karl, that's very 'sweet'! Is it possible to share the code? Thanks!
Apr 03, 2017, 01:28 PM
Multifaceted Heli Nut
heli-headcase's Avatar
Quote:
Originally Posted by dynexhobby
Hey All

Apologies for the late replay I didn't know there was activity on the forum

The basic formula for imbalance is F = m . (w^2) . r

where

F = Force blade imparts on assembly
M= mass of blade
w = rational velocity
r = radius of blade CG from axis of rotation

Now if you add up all "F" values (as vectors) and they equal zero then your system is statically balanced (in theory). In the formula "w" is constant for all blades hence you can knock that out of the formula.

I created a Excel sheet that illustrates the "F" term and plotted it in a polar format. I assumed that the CG was constant for each blade, which I know is not really true because it varies from blade to blade. The Excel sheet also calculates the residual imbalance based on blades weights and CG locations.

You can change the blade locations until your residual imbalance gets very small.... and yes many permutations and combinations.

But my question is, why would you want to go to this effort? Assuming the blades are strong enough and are not flexing a great deal, the dynamic balance method should take care of all this. In fact, there are some dynamic effects that purely weighing blades cannot not deal with so the ultimate method is to dynamically balance.

Anyway, have fun and feel free to double check the numbers as this is only a quick 5 min calc.
Ah ha Mark! I knew you'd come through

Dynamic balancing is ultimately the way to achieve the final, fine tuning but especially in the case of my (friend's) multi-piece impellers it will pay off with less mass correction in the dynamic stage if balanced statically as best practical. The entire 120mm chunk of spinning plastic and aluminum, save for the little nose cone, will be glued together as a permanent assembly. "Subtractive" balancing is the most practical way to achieve dynamic balance. There's only so much metal in the hub available for removal and carving off the least possible is my goal. Yes, manipulation of the blade's weight values will take some time but there's minimal guesswork involved and all blade material stays intact.

As I did state that I'll be using epoxy to permanently bond all the blades and hub parts together, I'll have to be mindful of the possible of voids in the epoxy-blade-hub interface. Small bubbles will undoubtedly form but as long as they're small I think everything will be okay.

I can't wait to try out your spreadsheet setup. I hope I can fumble through the process correctly Thanks a bunch!!!
Apr 03, 2017, 03:31 PM
Registered User
KarlHeinz57's Avatar

BestBalance Java Class


Quote:
Originally Posted by dynexhobby
Now if you add up all "F" values (as vectors) and they equal zero then your system is statically balanced (in theory). In the formula "w" is constant for all blades hence you can knock that out of the formula.
As "r" is also constant (assuming the CG of each blade is the same), it can be omitted for the purpose of determining the optimal blade arrangement.

I agree that dynamically balancing will take care of any imbalance. That being said, I think it's not a bad idea to get the blades arranged as close to balanced as possible during assembly.

Plugging the masses into your spreadsheet in the order that the program identified as optimal, the residual X and Y components were of the order of 0.005 - 0.009. If nothing else, getting the net CG that close to center would minimize the amount of weight that would need to be added if not eliminate it altogether.

I've uploaded the Java class file for those wishing to give it a try. To run the program, unzip the class file to a directory, open a command window, cd to that directory, and enter the command "java BestBalance mass1 mass2 mass3 etc...".

Include all of the blade masses, separated by spaces, on the command line. The program will accept any number of blades and will indicate the number of permutations to be tested and progress.

The 12-blade example here took about 3 seconds to complete while a 13-blade example took about 30 seconds and 14-blades... still running... Of course, these run times will depend on the power of your computer.

You should be able to copy and paste your results into the spreadsheet and visually verify the results.
Last edited by KarlHeinz57; Apr 03, 2017 at 03:49 PM.
Apr 03, 2017, 05:17 PM
Multifaceted Heli Nut
heli-headcase's Avatar
Hi Karl,

I'm itching to see your and Mark's processes work. Too bad I won't have access to a PC until much later tonight. Will just have to be patient I guess.

Thanks again guys
Apr 07, 2017, 09:49 AM
AMA 79169
rdstarwalt's Avatar
Hey Dynex... have we determined (or have a specification for) the maximum measurable frequency for the Impulse? I see the maximum input signal level and the sampling rate.

I suspect the answer is related to Nyquist's theorem as it relates to sampling rate.

In other words will the Impulse measure LF, HF, VHF, etc?

-=Doug
Apr 08, 2017, 03:58 AM
Good vibrations.
dynexhobby's Avatar
Umm.. thats an unusual question. Its mainly designed for vibrations so if using the acelereometer the maxium frequency is around 1.5kHz. However if you measure from another input source it can measure up to 20kHz. So to answer your question "no" it won't be able to measure that.
Apr 08, 2017, 08:16 AM
AMA 79169
rdstarwalt's Avatar
I didn't think it would, but I had to ask. I am working on another project that could use an RF range Spectrum Analyzer. I have a couple 'boat anchor' weight SAs at present. Packing up an Impulse size unit would be sweet.

I think the chip/pcb costs for a hobby grade RF SA would push the total cost beyond your current price point.

-=Doug
Apr 08, 2017, 11:01 AM
Multifaceted Heli Nut
heli-headcase's Avatar
Quote:
Originally Posted by rdstarwalt
I didn't think it would, but I had to ask. I am working on another project that could use an RF range Spectrum Analyzer. I have a couple 'boat anchor' weight SAs at present. Packing up an Impulse size unit would be sweet.

I think the chip/pcb costs for a hobby grade RF SA would push the total cost beyond your current price point.

-=Doug
Doug, have you heard of the RF Explorer hand-held SA?
https://www.rcgroups.com/forums/show...z-supported%21
May 06, 2017, 05:38 PM
Good vibrations.
dynexhobby's Avatar
This is amazing!! - https://www.facebook.com/f35model
May 06, 2017, 05:53 PM
Good vibrations.
dynexhobby's Avatar
A bit of nostalgia https://www.rcgroups.com/forums/show...im-Spectre-Jet
May 30, 2017, 11:22 PM
Good vibrations.
dynexhobby's Avatar
New software update available. Includes new solution for 2 blade propellers.
Jun 15, 2017, 04:47 PM
Registered User

video's on youtube


Hello

I looked at the videos of dynexhobby for EDF and i saw only 2 video's for the Vortex 4 Point ans Strobe method.
Are there no video's for the Impulse balancer for EDF's ?
Jul 11, 2017, 07:59 AM
Registered User

Impulse Dynamic Balancer out of stock


I would like to ask when Impulse will be back in stock. My APM Quad is getting still EKF errors because of high vibrations and have no other way to balance the propellers.


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Question EDF Impeller Balance old_dude Electric Ducted Fan Jet Talk 8 Nov 16, 2012 05:24 AM
For Sale Optical EDF Balancer for sale yakhakker Aircraft - Electric - Jets (FS/W) 2 Jul 29, 2012 01:16 PM
Sold Pair of balanced Wicked 4800 motor and matching ESC for twin EDF setup The Don Aircraft - Electric - Jets (FS/W) 4 Jun 25, 2012 08:34 PM
Discussion EDF fan balancing Mike873 Electric Ducted Fan Jet Talk 7 Apr 24, 2012 08:29 PM
Sold Exceed-RC stock 70mm EDF/Motor (balanced!!).......30 shipped RC_Bob Aircraft - Electric - Jets (FS/W) 2 Mar 26, 2012 12:46 PM