HobbyKing.com New Products Flash Sale
Reply
Thread Tools
Old Apr 06, 2012, 08:52 AM
Registered User
bri6672's Avatar
San Diego
Joined Jul 2009
2,236 Posts
Quote:
Originally Posted by erh7771 View Post
I've had 2 castle esc pop FETs with less than 1' cables....one with spark resister.

and when I say FETs its the Fets...not a regulator or some resister or the MCU got spiked...the FET cracked almost in half

They replaced the ESCs promptly

Thank goodness they were all on the ground, one guy who's been flying with them for years popped them on hook up of 6s pack 5 months ago.

It's not voltage run time variances, to fix that it's cheap and light and easy...
Yep we got it, We are talking about Caps not FETs, thats a whole other subject...
bri6672 is offline Find More Posts by bri6672
Reply With Quote
Sign up now
to remove ads between posts
Old Apr 06, 2012, 09:47 AM
Registered User
United States, TX, Grand Prairie
Joined Nov 2005
5,677 Posts
Quote:
Originally Posted by bri6672 View Post
Yep we got it, We are talking about Caps not FETs, thats a whole other subject...
There are plenty of ESCs that don't need big stiffing caps to help their components seeing fairly constant input power. If stiff input power were a real issue with transistors we'd rarely see high end audio components run off house current. Similar ESC components have been powering car audio systems for years and you can tell there are few, outside of competition systems, that have stable power.

I don't use caps to protect the components so much because even on SUPPOs over 1.5ft bat wire is not an issue.

I put bigger caps with some good gauge wire for longer flight times and to keep the packs cool. I can have longer flights because burst current is coming from the caps and not the packs and dipping below cut off doesn't happen until the packs REAL run time volts dip

Keeping the burst current from reaching the packs even for half a second will keep the heat down almost 2 fold because caps discharge almost 10 times faster than a LIPO. Unless packs are run past their C rating at a given mah I rarely have a heat issue with my big power systems.

You know, come to think of it we REALLY don't need a 10 times the discharge speed and rate of LIPO packs for longer flight times and cooler packs. What is really needed from caps is the leads need to be bigger so you can run less of them or get some with lower resistance as mentioned up thread.

Hmmmmm, I got an idea....dang it....this whole job thing is keeping me from being really addicted!

sigh
erh7771 is online now Find More Posts by erh7771
Reply With Quote
Old Apr 06, 2012, 04:55 PM
Watt me worry?
Madmax1965's Avatar
Florida
Joined Mar 2006
3,799 Posts
bri6672:
Exactly! The caps included should be sufficient for most planes but when we start making our wires super long that's the problem.

May not be true as in my F-86 the wires were super short and I still had a fire.....just a good thing the plane was on the ground so I could extinguish the fire. CC said the problem was from ripple current caused as I was "blipping" the throttle while testing the power setup before the maiden flight. According to CC tech it is not a good idea to "blip" the throttle on the high powered planes(the F-86 was @ 12s).

Speaking of caps......I read that we may be using caps as a power supply in the future instead of batteries. They are somewhat large now but could be smaller later on.

http://www.greentechmedia.com/articl...-be-batteries/
Madmax1965 is offline Find More Posts by Madmax1965
Last edited by Madmax1965; Apr 06, 2012 at 05:02 PM.
Reply With Quote
Old Apr 07, 2012, 12:56 AM
Registered User
Peacemakr40's Avatar
Joined Aug 2011
88 Posts
Yea, Think I'll jump in here as there's all sorts of confusion about this topic.
just a quick background about me for Credentials:
Currently I'm an IT manager who still performs field work.
I hold a BSEE from Buffalo State University. My desciplines were communications and power distribution.
I've worked for major corps as a design engineer for UPS/BPS systems, Smitchmode power supply design, blah blah blah.

The Purpose of filtration capacitors is to reduce drops in applied voltages. When sourcing an ESC with a battery voltage, you will, by the switching nature of the ESC, cause rippling of the supply voltage. The caps will act as high speed current sources to offset the rapid draining from the ESC. Just like Lipos and other batteries, Capacitors have a charge/discharge rating based on it's capacitance and it's ESR (equivalent series resistance). For High stability filtration, I see everyone mentioning paralleling mid level uf caps (3-700 ufs) at the 50 volt range and saying it's God's gift to ESC protection. That unfortunately is wrong. There are several factors to consider. I will touch on each of them below

Length of wires from battery to ESC:
As we all know, wire has resistance. the longer the wire, the higher the resistance. This also applies to diameter. The smaller the diameter the higher the resistance as well. Ideally, you want a huge copper block from each battery terminal connecting directly to the + and - of the ESC. This will never happen in reality where grams can dictate how well a model flies.
Another aspect of longer wiring is the inductance created (voltage induced in a wire when an electrical current is applied). Though not as critical as the wire's resistance, it can still play a role with higher power applications.

Capacitance in Parallel to the Source:
The level of filtration you have is determined by how much ripple is acceptable ripple. The Ripple you get will typically be in the 8-16 KHz (Kilohertz) range as this is the standard freqs that ESCs run at.
While having a huge honkin 10 FARAD capacitor bank rated at 50 volts (100Volts for HV systems) might seem to be the bees knees (I'm running out of euphamisms) It cannot react quickly enough to source the additional current that the ESC is demanding. Because of this, the implimentation of a couple of smaller uf caps will solve that issue.

Oscillation of an LCR network:
Because of the larger sized total capacitance to be used vs the wiring resistance and the wiring inductance, a harmonic ringing can occur. This ringing can feed back both into the ESC and to the battery reducing their overall efficiency and could damage the ESC. Again, small "snub" caps can be used.
A snub cap is a small value NON-ELECTROLYTIC cap used to snub out the oscillations. I prefer to use .1 uf caps as they're cheap and easy to get.

Considerations for adding cap banks:
1 - what level of filtration do you want to achieve?
2 - What is your source voltage and C rating?
3 - How many Amps are you pulling under full load?
4 - How long are the wires from the battery to the ESC?
5 - what is your weight limit ( some filter caps can weigh quite a bit)?
6 - How much room do you have to work with near your ESC?

Given the typicals for filtration. I would go with no less than 1000 uf per battery Cell for high current, 500 uf per cell for low power. The voltage rating should be 3 times that of your total battery voltage. Place the capacitors as close as possible to the ESC and use Low ESR electrolytic caps with 105 degree C rating.
Personally I like Nichicon (United Chemicon) caps over Rubicons but I shop around for the best deals. Panasonics and Sanyo's seem to work pretty well too.
2 snub caps per capacitor bank and at least 1 47uf at the rated voltage.
Using more capacitors in Parallel for the same total uf value will give a lower ESR.
Also, use the heaviest wire you can get away with (due to weight restrictions). 2 wires of the same gauge may not have the same resistance. Smaller strands in a wire will give you a lower resistance than the same gauge wire with larger strands.

I hope this sheds some light on the need and use of Cap packs.
Peacemakr40 is offline Find More Posts by Peacemakr40
Reply With Quote
Old Apr 07, 2012, 08:08 AM
Registered User
Doug Bateman's Avatar
Swarthmore, PA 19081
Joined Dec 2004
2,838 Posts
Good stuff Peace......I will try to absorb and apply this.
Doug Bateman is offline Find More Posts by Doug Bateman
Reply With Quote
Old Apr 07, 2012, 09:00 AM
Registered User
Joined Jun 2002
860 Posts
It's nice to see an EE response to this issue ~ Thank You!
Ron Pogue is offline Find More Posts by Ron Pogue
Reply With Quote
Old Apr 07, 2012, 09:53 AM
awrightbrother's Avatar
Florida
Joined Nov 2007
2,456 Posts
Hello Peacemakr40
As one EE to another, I have a question.

Instead of adding caps and therefore weight, and more components that can fail, why not locate the ESCs at the battery end and extend the motor wires?

Due to the pulsed nature of this side of the ESC, there is the risk of RF interference that may effect the RX. Can this be mitigated by locating the RX some distance from the motor wires, and/or twisting the wires, and/or using shielded wires, and/or shielding the Rx?

Has anyone tried some or all of the above?
awrightbrother is online now Find More Posts by awrightbrother
Reply With Quote
Old Apr 07, 2012, 10:13 AM
Pursuit of Happiness
Ron101's Avatar
Brentwood, California
Joined Jul 2007
7,189 Posts
Quote:
Instead of adding caps and therefore weight, and more components that can fail, why not locate the ESCs at the battery end and extend the motor wires?
In some models there is no room or cooling for the ESC.... so the ESC is located in the tailcone of the fan for good cooling, this leave long leads to the packs
Ron101 is offline Find More Posts by Ron101
Reply With Quote
Old Apr 07, 2012, 10:21 AM
Capt. Z
falcon5's Avatar
Tonopah, Nevada
Joined Dec 2004
2,540 Posts
OK peacemakr40 I have the mother of all installs going here. Curious as to what an EE would suggest about it,

Considerations for adding cap banks:
1 - what level of filtration do you want to achieve? = A good level
2 - What is your source voltage and C rating? = 10S per fan . 30C batts
3 - How many Amps are you pulling under full load? = 120amps per fan
4 - How long are the wires from the battery to the ESC? = 45"
5 - what is your weight limit ( some filter caps can weigh quite a bit)? = No real limit
6 - How much room do you have to work with near your ESC? = allot

And the vedict is????
falcon5 is offline Find More Posts by falcon5
Reply With Quote
Old Apr 07, 2012, 10:33 AM
Gorilla Glue Expert
PikeStaff's Avatar
United States, PA, Doylestown
Joined Sep 2010
2,667 Posts
Quote:
Originally Posted by falcon5 View Post
OK peacemakr40 I have the mother of all installs going here.

And the vedict is????
Holy cow.
PikeStaff is online now Find More Posts by PikeStaff
Reply With Quote
Old Apr 07, 2012, 11:08 AM
Pursuit of Happiness
Ron101's Avatar
Brentwood, California
Joined Jul 2007
7,189 Posts
Quote:
OK peacemakr40 I have the mother of all installs going here. Curious as to what an EE would suggest about it,

Considerations for adding cap banks:
1 - what level of filtration do you want to achieve? = A good level
2 - What is your source voltage and C rating? = 10S per fan . 30C batts
3 - How many Amps are you pulling under full load? = 120amps per fan
4 - How long are the wires from the battery to the ESC? = 45"
5 - what is your weight limit ( some filter caps can weigh quite a bit)? = No real limit
6 - How much room do you have to work with near your ESC? = allot

And the vedict is????

I love Andy's build!! in fact so much that I hope to head to vegas for the ace in the hole event to see this beast!

Andy if you don't mind my input here is what I would do. Castle recommends 1 of there cap packs for every 8 inches, that's four panasonic 50 volt 330 uf caps per 8 inches. One cap every 2 inches ( 45 / 2 = 22.5) so lets call it 24 caps to be safe...
You need 45" (I need 30 in my F-18 so I know the feeling ) I wouldn't want to run 24 caps ... so I'm going to step up to the bigger Panasonic 680 uf 50 volt....cut your number in half and add 12 of the larger caps. Maybe that thats over kill and your fine with 10?

link to the smaller 330 caps that castle uses in there cap packs:
http://search.digikey.com/scripts/Dk...9-ND+&x=8&y=10

Here is the larger 680 uf panasonic fm series caps:

http://search.digikey.com/us/en/prod...2402-ND/613763

I just ordered a bunch of both... I'll add them in line myself.....much cheaper than the cap packs
Ron101 is offline Find More Posts by Ron101
Reply With Quote
Old Apr 07, 2012, 11:17 AM
Registered User
Wilsonville, OR
Joined Jun 2010
555 Posts
Quote:
Originally Posted by Ron101 View Post
I love Andy's build!! in fact so much that I hope to head to vegas for the ace in the hole event to see this beast!

Andy if you don't mind my input here is what I would do. Castle recommends 1 of there cap packs for every 8 inches, that's four panasonic 50 volt 330 uf caps per 8 inches. One cap every 2 inches ( 45 / 2 = 22.5) so lets call it 24 caps to be safe...
You need 45" (I need 30 in my F-18 so I know the feeling ) I wouldn't want to run 24 caps ... so I'm going to step up to the bigger Panasonic 680 uf 50 volt....cut your number in half and add 12 of the larger caps. Maybe that thats over kill and your fine with 10?

link to the smaller 330 caps that castle uses in there cap packs:
http://search.digikey.com/scripts/Dk...9-ND+&x=8&y=10

Here is the larger 680 uf panasonic fm series caps:

http://search.digikey.com/us/en/prod...2402-ND/613763

I just ordered a bunch of both... I'll add them in line myself.....much cheaper than the cap packs
Ron, you should do a "Rons House" episode (like the anti-spark) when you install. With all the responses in such a short time I think it'd be great!
Wisdumb is offline Find More Posts by Wisdumb
Reply With Quote
Old Apr 07, 2012, 11:24 AM
Pursuit of Happiness
Ron101's Avatar
Brentwood, California
Joined Jul 2007
7,189 Posts
Quote:
Ron, you should do a "Rons House" episode (like the anti-spark) when you install. With all the responses in such a short time I think it'd be great!
Great idea!! in fact I'm going to add them to my Hawk which I plan to fly very soon... that will be next week so I'll do something then
Ron101 is offline Find More Posts by Ron101
Reply With Quote
Old Apr 07, 2012, 11:29 AM
Registered User
Nav_Aids's Avatar
Canada, MB, Winnipeg
Joined Nov 2011
331 Posts
A different view

Ok this is going to be long winded but, bare with me it may be worth while. First we are dealing with very large current flows and things can get a bit weird.

Quote:
Originally Posted by Madmax1965 View Post
.......
Concerning longer wiring...........I have posted this thread before(many times)and there is good info in it concerneing wiring and caps.
http://www.rcgroups.com/forums/showt...3#post10926525
This link that Madmax1965 provided to this old post is a very good read and should be read by all concerned. It comes very close to what is happening.


Quote:
Originally Posted by bri6672 View Post
I have spent alot of time with castle on the phone discussing this and here are my findings and this has worked great for me:

You first need to establish a baseline of ripple current using some fresh batteries. Start by running up a simulated short flight on the ground.(here is why) anything under .05 is fine but as batteries get older the average ripple current will increase, so if you are at let say .045 to start you will be fine for awhile but as the batteries age you will go over .05 and that can lead to problems in the ESC. .......
bri6672, I'm sorry but. What unit of measurement are you using? If we are talking about amps or volts then those numbers don't mean much. The values are so small that they become insignificant, if they are Ka or Kv then that's a different story. I just don't understand what youíre trying to say.


Quote:
Originally Posted by erh7771 View Post
K, ....sigh

Why doesn't caste simply add bigger caps to the ESC to smooth out current flow?

Caps are relative inexpensive and light...and easy to attach....

Cause it's not the caps, it's the FETs are out of tolerance from factory...SUPPO esc's went through a bout of bad FETs on their esc's and they were gimped because of it.

It only takes one to fail or heat up to the point were its properties change and make others fail...

Yaws take?

TIA
Well the FETs are the end result of a problem and you are correct in saying that they should be better but, that costs money and the ESC market is very competitive.


Quote:
Originally Posted by Peacemakr40 View Post
........ The Purpose of filtration capacitors is to reduce drops in applied voltages. When sourcing an ESC with a battery voltage, you will, by the switching nature of the ESC, cause rippling of the supply voltage. The caps will act as high speed current sources to offset the rapid draining from the ESC. Just like Lipos and other batteries, Capacitors have a charge/discharge rating based on it's capacitance and it's ESR (equivalent series resistance). For High stability filtration, I see everyone mentioning paralleling mid level uf caps (3-700 ufs) at the 50 volt range and saying it's God's gift to ESC protection. That unfortunately is wrong. There are several factors to consider. I will touch on each of them below ..............
I too have built switching PSs and yes, noise and oscillations are a pain and require lots of de-coupling and filtration. I also have a small company and we specialize in R&D for other companies. Your thinking that this is a filtration problem and in a limited way it is but itís not the problem, I'll explain below.


I hope most people know how an ignition coil works and/or why we get a spark when we unplug an electrical cord. Well the same thing is happening here except no spark, I hope. I could get very technical and start quoting formulas but, I wonít. Anyway, when a current flows through a wire a magnetic field is created. The larger the current flow the larger the magnetic field. When that field collapses, the magnetic lines of flux will cut across the wire and induce current flow in the same direction that the current was flowing in. These lines of flux also create a voltage to appear on the wire (aka. voltage spike.) These voltage spikes can be large enough to blow capacitors or cause punch through on the FETs and open them or worse short them out. If the FETs open then the ESC is dead, if the FETs short then the current flow will be so high they will either pop (crack) or over heat and burn along with the plane. I know about these effects because I have spent hundreds if not thousands watching FETs, IGBTs, and other stuff die or explode.

So how do the caps help? Well Peacemakr40 was on the right track when he mentioned LRC. Now here is something to remember:

ELIe The ICEman, where:
E = voltage
L = inductance
I = current
This means that voltage leads current by 90 degrees in a purely inductive circuit. And
I = current
C = capacitance
E = voltage
This means that current leads voltage by 90 degrees in a purely capacitive circuit.

I agree that the caps should be as close to the ESC as possible on both sides. On the battery side the caps provide 2 things, surge current to the ESC quicker than the battery can and noise filtering. On the motor side the magnetic flux in the wire will try to collapse through the wire to the motor because of ELIe and the caps have ICE, the 2 are trying to cancel out each other. So when you crank up the power the caps slow down the power going to the motor until the caps are charged up. This is actually a good thing because current surges are bad. When you throttle or power down the caps slow down the collapse of the magnetic flux because of the voltage they hold. -90 + 90 = 180 degrees out of phase, and try to cancel each other out. Eventually the load (motor) will discharge the caps along with the magnetic flux. Because the magnetic flux canít collapse very fast the voltage spike isnít near as high as it was without the caps. I have also seen where if the timing is right you can create a huge voltage spike if you reapply power at just the right time when the magnetic field is collapsing. I use this effect on an ignition Iím developing.

I think Peacemakr40 would agree with me here that one can calculate the size and voltage rating of the caps needed using LRC formulas. Do use caps with the lowest ESR rating you can find or afford. For caps on the battery side use the largest you can fit or carry (weight) with a voltage rating of at least 150% of the stated charger voltage. For a caps on the motor side this is a little more difficult because the caps have to follow the wire size. The best way is to find the peak voltage when the throttle is cut from full. You want to use the smallest size cap that will keep the voltage spike the lowest or as close to battery voltage as possible. This all brings up another question; are large or larger cables better? Well based on what I have said here, I would have to say no. The larger the wire size the larger the inductive kick. I would say use a wire size that has 125% larger current capability as the peak current used. Catch 22, you gotta love it.

I hope all this stuff helps and makes sense, if it doesnít oh well I tried and besides itís just my 2 cents worth. MOSFET manufacturers have released some new products in the last 6-12 months that should really help the electric R/Cers. Like FETS that can handle 200 amps at 1500 volts, now I'd like to see someone blow one of these up.

Ray
Nav_Aids is offline Find More Posts by Nav_Aids
Reply With Quote
Old Apr 07, 2012, 11:43 AM
Pursuit of Happiness
Ron101's Avatar
Brentwood, California
Joined Jul 2007
7,189 Posts
for $40 bucks I bought enough caps to make up 20 castle cap packs
That's $500 dollars worth of castle cap packs
Ron101 is offline Find More Posts by Ron101
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion HK ESC + Turnigy Motor Timing Issue lipoflyer Power Systems 6 Mar 20, 2012 10:19 AM
Discussion Miracle of the Avus, or the one time it's good to have your seatbelt fail.. CashRC Life, The Universe, and Politics 1 Feb 23, 2012 01:27 PM
Discussion WWF ad just in time for 9-11 E-MO Life, The Universe, and Politics 68 Sep 03, 2009 06:28 PM
Question Nikon coolpix 3700 ad on ebay...is this a good deal/good beginner camera? Picoflyer Aerial Photography 15 Jul 26, 2004 01:13 PM