SMALL - espritmodel.com SMALL - Telemetry SMALL - Radio
Reply
Thread Tools
Old Feb 26, 2013, 04:50 PM
Registered User
Australia, VIC, Melbourne
Joined Nov 2006
10,717 Posts
When you clock the fan, even just a handful of degrees away from the truly perfect position can mean you will never get it running near as best it can... it is a big change even at that small error remaining.
So you need to go right down to just a few degrees in steps, to really pin point the final area. Get it dead on the right spot and it is a big difference.

So you want to find that dead perfect (absolute minimal vibration you can get) spot first. Then move on to 'weight checking' so you can also fix the weight error it will still surely have too.
You can either clock a test weight (same process as fan clocking, but with a weight), or use a 4 point balance method for that.

Start with a small weight... bluetac or tape... of about 0.1g or 0.2g. Start anywhere - mark a line - and clock around, as per fan method, checking for the lowest vibration position of that weight. Then you will know WHERE it needs weight.
Next is test various weights to pinpoint how much weight exactly is required there. (4 point balance method's math can tell you that weight.... to a pretty good degree at least).

This final 'weight clocking' will move a fan to a far higher level of great balance than just clocking the rotor. Clocking ONLY the rotor can only ever work out 'great' IF the rotor's weight imbalance miraculously flukes being exactly what all other imbalance totals are - thus offsets it perfectly! That will pretty well never occur!
You will always do far better with correct weight also.
PeterVRC is offline Find More Posts by PeterVRC
RCG Plus Member
Latest blog entry: SBEC's connected in parallel
Reply With Quote
Sign up now
to remove ads between posts
Old Feb 26, 2013, 09:08 PM
C.E.R.N. field tester
supercollider's Avatar
United States, NE
Joined Sep 2012
949 Posts
Very true , I don't think an outrunner's imbalance will ever perfectly cancel out a fan rotor's imbalance. Have to use some weight. I have a few outrunners and none of them are even close to being balanced without help. I think inrunners are the way to go anymore.

Cheers
supercollider is online now Find More Posts by supercollider
Reply With Quote
Old Feb 26, 2013, 09:48 PM
Registered User
Australia, VIC, Melbourne
Joined Nov 2006
10,717 Posts
Yeah, I am disliking outrunners... but there are some great cheap combos you can make from those!
You can balance the bell too, to help improve them. Even using the phone app thing still for that too.
Same principal.... just clocking a bit of tape weight around them, until optimally balanced. Then some epoxy dab/smear pout inside the bell.
PeterVRC is offline Find More Posts by PeterVRC
RCG Plus Member
Latest blog entry: SBEC's connected in parallel
Reply With Quote
Old Feb 27, 2013, 03:06 PM
Registered User
PeterF's Avatar
United Kingdom, England, North York
Joined Nov 2004
337 Posts
OK, first fan nearly done. Ran around in 45 increments to home in on the lowest point which was a 270 rotation from the datum mark. Checked either side of this and confirmed that 270C was at the minimum. Measured AC voltage reduced from 0.77V at the datum from the fan vendor to 0.54V at the best position. I then added strips of tape between the 7 blades (51.4 increment) one at a time to get the best position for weighting. This reduced the measured voltage to 0.37. I then built up the weight until I got to 0.18V. This is about 1/4 of the vibration level at the datum position from the fan vendor. I ran out of time but I may be able to add a little more weight and reduce this even further. Polar plots attached below.

I then put it back on the scales to see if I had any more thrust or better efficiency. The thrust and power consumption were the same as before, so the amount of energy being wasted by the vibration as delievred was small. I believe the sound of the fan is better, there is less fan noise and more rushing of air noise, but that is subjective.
PeterF is offline Find More Posts by PeterF
Reply With Quote
Old Feb 27, 2013, 04:28 PM
Registered User
Australia, VIC, Melbourne
Joined Nov 2006
10,717 Posts
Interesting info to see, thank PeterF

I now have the Dynex Vortex balancer unit (It comes as just electronics, no cradle).
One thing that makes it stand out is the PC Scope software can do frequency analysis (spectrum), and you will find a wide variation of harmonic and spurious freqs coming from various fan combos. eg even two of the same parts total will have different spectrums.

The content of the spectrum is energy..... so the total resultant you see in a Accelerometer is ALL those freqs. At least up until the freq level the ACC, or circuit, read up to. Not your choice in a Phone, it just 'sees' what it sees.
This means the balancing you do is not purely on the imbalance itself. But also the energies of those various 'who knows what created them' freqs also. Thus you are using numbers that are not quite truly just imbalance numbers.

If you set filtering to remove them, then you get different results.
This showed me (it seems) why the Phone app can have such 'wavery' levels - overly crazy, versus how the pure RPM vibration is virtually a perfect sine wave result really.
An imbalanced mass, rotating a complete circle = perfect sine wave.
So if those extra freqs are not removed, then the results used are a bit askew.

This is where the final bit of manual fine tuning helps to correct that error margin.
But those extra freqs are also signs of other sources of vibration issue... like bearings, coupled imbalance (eg motor bell/armature out of balance), slight shaft error, etc.
So to have a precise balance you need those other freqs (vibration sources) reduced/removed too.
PeterVRC is offline Find More Posts by PeterVRC
RCG Plus Member
Latest blog entry: SBEC's connected in parallel
Reply With Quote
Old Feb 27, 2013, 04:57 PM
Registered User
PeterF's Avatar
United Kingdom, England, North York
Joined Nov 2004
337 Posts
Yes it is quite complex and on a simple level, with other noise in the system you will never balnce it perfectly using this type of method, nor with the smartphone, but at least it gets someway there and with a 55mm fan it is probably good enough. The Dynex system looks very nice.

Anyway, I had a final tweak with the balance weight again and got the reading down to 0.08V, which is about 10% of the value I started at so I am pleased with the method even if it is not the best. Now I have 3 more fans to do.
PeterF is offline Find More Posts by PeterF
Reply With Quote
Old Feb 27, 2013, 05:02 PM
Registered User
Rutland, England
Joined Feb 2005
217 Posts
Quote:
Originally Posted by PeterF View Post
OK, first fan nearly done. Ran around in 45 increments to home in on the lowest point which was a 270 rotation from the datum mark. Checked either side of this and confirmed that 270C was at the minimum. Measured AC voltage reduced from 0.77V at the datum from the fan vendor to 0.54V at the best position. I then added strips of tape between the 7 blades (51.4 increment) one at a time to get the best position for weighting. This reduced the measured voltage to 0.37. I then built up the weight until I got to 0.18V. This is about 1/4 of the vibration level at the datum position from the fan vendor. I ran out of time but I may be able to add a little more weight and reduce this even further. Polar plots attached below.

I then put it back on the scales to see if I had any more thrust or better efficiency. The thrust and power consumption were the same as before, so the amount of energy being wasted by the vibration as delievred was small. I believe the sound of the fan is better, there is less fan noise and more rushing of air noise, but that is subjective.
Well done Peter. However, i'm not quite following you in the red bit??? I thought that the next stage after finding the best relative position for the fan (rel. to motor etc) was to take the fan off, put it on a balancer, then find the position where it is lightest (ie the point on the 'up' side where it comes to rest). Then isn't that the spot where you add the weight?

Mind you... it's a bit late, bit tired; i may have lost the plot

Rgs, Andy
Andy B is offline Find More Posts by Andy B
Reply With Quote
Old Feb 27, 2013, 05:09 PM
Registered User
PeterF's Avatar
United Kingdom, England, North York
Joined Nov 2004
337 Posts
Andy,

Yes you are correct, I should have mentioned this. I put it on my magnetic balancer but such a small rotor was having difficulty coming up with a uniform light side. It did identify a light side which was at about 180. However, when I put tape on at this point and span it up the vibrations got a little better and then worse as I added more. It was apparent that where the balancer was coming out was not the real light side, possibly because the cone sat into the carbon on the front of the rotor, not the aluminium adapter, so the fan did not sit central. Therefore, I had to adopt the other method of identifying where to add the balance weight which was to make a small bit of tape up and move that arund the fan to find out where it made the most difference, which was then the location for the weight. Some of the posts in this thread metnion this exact problem as being a possibility.
PeterF is offline Find More Posts by PeterF
Reply With Quote
Old Feb 27, 2013, 06:21 PM
Registered User
Rutland, England
Joined Feb 2005
217 Posts
Quote:
Originally Posted by PeterF View Post
Andy,

Yes you are correct, I should have mentioned this. I put it on my magnetic balancer but such a small rotor was having difficulty coming up with a uniform light side. It did identify a light side which was at about 180. However, when I put tape on at this point and span it up the vibrations got a little better and then worse as I added more. It was apparent that where the balancer was coming out was not the real light side, possibly because the cone sat into the carbon on the front of the rotor, not the aluminium adapter, so the fan did not sit central. Therefore, I had to adopt the other method of identifying where to add the balance weight which was to make a small bit of tape up and move that arund the fan to find out where it made the most difference, which was then the location for the weight. Some of the posts in this thread metnion this exact problem as being a possibility.
Ah, i think i understand now. Thanks Peter.
Andy B is offline Find More Posts by Andy B
Reply With Quote
Old Feb 27, 2013, 06:34 PM
Registered User
Australia, VIC, Melbourne
Joined Nov 2006
10,717 Posts
Don't bother with static balancing at all.... or rather, use it to get a rough base and that is that.
Because of mounting issues (whether cones or other reasons) you won't find the perfect true balance anyway, and don't even WANT that! Because we don't have all fan components 'perfect', so you are certainly going to need an 'imperfect' rotor balance to give the best overall result. (offsetting other imbalances)
PeterVRC is offline Find More Posts by PeterVRC
RCG Plus Member
Latest blog entry: SBEC's connected in parallel
Reply With Quote
Old Feb 27, 2013, 07:47 PM
C.E.R.N. field tester
supercollider's Avatar
United States, NE
Joined Sep 2012
949 Posts
Yes, even the shaft used to balance the fan rotor most likely will not be "true". I have a couple and one the shaft is not straight, and the other one has the spear shaped ends not even, its always trying to center itself and with any weight on it just keeps going round and round. Works OK on larger props but not good for edf fans. I need a balancer for my balancer so to speak. Lol

Cheers
supercollider is online now Find More Posts by supercollider
Reply With Quote
Old Feb 28, 2013, 06:17 AM
Registered User
Rutland, England
Joined Feb 2005
217 Posts
Some money saved :-)

Quote:
Originally Posted by PeterVRC View Post
Don't bother with static balancing at all.... or rather, use it to get a rough base and that is that.
Because of mounting issues (whether cones or other reasons) you won't find the perfect true balance anyway, and don't even WANT that! Because we don't have all fan components 'perfect', so you are certainly going to need an 'imperfect' rotor balance to give the best overall result. (offsetting other imbalances)
Quote:
Originally Posted by supercollider View Post
Yes, even the shaft used to balance the fan rotor most likely will not be "true". I have a couple and one the shaft is not straight, and the other one has the spear shaped ends not even, its always trying to center itself and with any weight on it just keeps going round and round. Works OK on larger props but not good for edf fans. I need a balancer for my balancer so to speak. Lol

Cheers
Ah, that sounds really sensible advice guys - thanks; as a newbie to edf - you've just saved me bothering to upgrade my prop balancer!

I'm going to give Peter's (PeterF) accelerometer board method a try too - he's convinced me that this is a most viable bit-of-kit; ... and that will probably save me upgrading my phone too! (i'll stick with my Luddite version).

Thanks guys

ps: It sounds like PeterF's rotor may have come from the manufacturer 'too well balanced' doesn't it. Still... i guess we'd not want to ask them to make them deliberately inaccurate :-) I'm also kinda not all that surprised that Peter's results indicate that you don't actually save any significant amount of power with an accurately balanced system - however, the bearing loads must at least be reduced... and in the longer term this can only be good for motor life. AND IT WILL SOUND BETTER!
Andy B is offline Find More Posts by Andy B
Last edited by Andy B; Feb 28, 2013 at 06:28 AM.
Reply With Quote
Old Feb 28, 2013, 07:20 PM
C.E.R.N. field tester
supercollider's Avatar
United States, NE
Joined Sep 2012
949 Posts
"ps: It sounds like PeterF's rotor may have come from the manufacturer 'too well balanced' doesn't it. Still... i guess we'd not want to ask them to make them deliberately inaccurate :-) I'm also kinda not all that surprised that Peter's results indicate that you don't actually save any significant amount of power with an accurately balanced system - however, the bearing loads must at least be reduced... and in the longer term this can only be good for motor life. AND IT WILL SOUND BETTER!"

Precisely!

Cheers
supercollider is online now Find More Posts by supercollider
Reply With Quote
Old Mar 01, 2013, 05:12 AM
Registered User
Rutland, England
Joined Feb 2005
217 Posts
Ok guys, i've now ordered my MMA7361 chip set... and should have my very first edf motors and CS12 fans arriving anytime now. When I've prepared a rig and done some playing - I'll let you know how I get on. My Vampire kit arrived yesterday too; it's all very exciting!

Rgs, Andy
Andy B is offline Find More Posts by Andy B
Reply With Quote
Old Mar 06, 2013, 01:17 PM
Registered User
Rutland, England
Joined Feb 2005
217 Posts
Quote:
Originally Posted by PeterVRC View Post
You can either clock a test weight (same process as fan clocking, but with a weight), or use a 4 point balance method for that.
.
What's the 4 Point Balance method Peter?
Andy B is offline Find More Posts by Andy B
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
For Sale Don's Wicked 4800 mounted in a dynamically balanced 64mm EDO Type I EDF Unit. jhamm Aircraft - Electric - Jets (FS/W) 1 Jun 06, 2010 07:03 AM
Question How to modify foam to fit larger EDF unit? Silverstreak328 Foamy EDFs 8 Feb 09, 2010 10:31 AM
Discussion How to balance an EDF impeller ? i3dm Electric Ducted Fan Jet Talk 7 Sep 16, 2009 05:07 AM
Discussion Looking to dynamically balance fans. Sniper.308 Electric Ducted Fan Jet Talk 4 Nov 18, 2008 01:12 AM
Discussion How to make a pitch gauge with a rubberband and a cell phone ? mguw Electric Heli Talk 0 Jul 28, 2008 07:58 AM