HobbyKing.com New Products Flash Sale
Reply
Thread Tools
Old Oct 15, 2009, 01:45 PM
An itch?. Scratch build.
eflightray's Avatar
South Wales U.K.
Joined Mar 2003
13,378 Posts
Discussion
Two BEC's in parallel ?

For electric fliers building twin motor models, there has been a general recommendation that if using two ESC's, to disable one of the BEC features when connecting both to the receiver, (usually through a 'Y' lead). Disabling a BEC is usually by removing the center pin, (+ive) from the connector or cutting the wire).

Basically the thought was, you shouldn't use two voltage regulators in parallel.

I have recently been told this is a myth, and that you can connect the two BEC's to the Rx.

Can anyone confirm that having two BEC's connected doesn't cause a problem ?
eflightray is online now Find More Posts by eflightray
Reply With Quote
Sign up now
to remove ads between posts
Old Oct 15, 2009, 02:20 PM
Registered User
USA, CA, Pismo Beach
Joined Oct 2007
1,379 Posts
There is not a simple answer to this question, depending on the BEC circuit - linear or switching. If the BEC is linear, it is probably OK to operate two in parallel if you recognise that they will not be sharing the load well. The reason is that it is highly unlikely that the actual regulated voltage will be exactly the same on each BEC. Therefore the load will be serviced by the output with the highest voltage. If the load becomes large enough to cause the voltage to sag on the dominant BEC, only then will the other begin to provide current. The rate at which this happens depends on just how good a voltage regulator it is.

With switchers, it may or may not work depending on circuit stability issues and could result in high frequency oscillations which would not be desirable feeding your RX. You might check with the manufacturer.

Gw
Greywing is offline Find More Posts by Greywing
Reply With Quote
Old Oct 15, 2009, 03:32 PM
An itch?. Scratch build.
eflightray's Avatar
South Wales U.K.
Joined Mar 2003
13,378 Posts
Thank Gw

So it would seem that the standard, (cheaper), ESC's that uses a linear regulator for BEC are OK to be paralleled.

Does anyone have any information on just what 'may' go wrong, (I know that's a bit open), as I wonder what may have started the 'scare' about not connecting two BEC's in the first place.
eflightray is online now Find More Posts by eflightray
Reply With Quote
Old Oct 15, 2009, 03:44 PM
An itch?. Scratch build.
eflightray's Avatar
South Wales U.K.
Joined Mar 2003
13,378 Posts
I have just tried a quick google on, 'why disconnect one bec on a twin'.

The only comment I could find was, 'to stop the BEC's 'fighting' each other'.

Fighting ?

Now it's not too hard to visualize two ESC's having a punch-up in the fuselage, (if you have a weird sense of humor like me), but is that the only answer ?
eflightray is online now Find More Posts by eflightray
Reply With Quote
Old Oct 15, 2009, 03:50 PM
Registered User
USA, CA, Pismo Beach
Joined Oct 2007
1,379 Posts
To expand slightly, I suspect the the problem lies in assuming that with two BEC's in parallel, you can provide twice the load current which is not the case. The BEC with the highest output voltage will be providing all the current up to the point it becomes overloaded and its output voltage sags at which point the other will begin providing enough current that the voltage can come up again. So one BEC will be operating at full or near full capability whilst the other is virtually idling along.

To best take advantage of two BEC's try splitting the load with some of the servos being driven by one BEC and the balance by the other. A bit more pain with wiring but relatively straight forward. This scheme of course would apply to both linear and switching BEC's.

GW
Greywing is offline Find More Posts by Greywing
Reply With Quote
Old Oct 15, 2009, 04:24 PM
Registered User
Canada, NB, Fredericton
Joined May 2002
932 Posts
As Greywing mentions, most plausible explanations surround the imbalanced loading. I suppose it could be a serious problem on poorly matched BECs, where one might be handling the full load and heat up to the point of shutting down before the other one kicks in, then it takes the full load and gets to the shutdown point. This could cause cascading failure, or load cycling (which could cause weirdity if this results in voltage fluctations, etc)

But, there are documents on doing load sharing with linear vregs from the manufacturers...most don't recommend it, but suggest it can be done with current limiting resistors (0.1-1.5 ohm?), but this does waste some power in the resistors.

I figure the that ESC PCB & lead from the ESC to the RX adds a few hundred milliohms, so there's already some degree of current limiting resistance to help the BECs balance out. I wouldn't trust two parallel BECs to give much extra headroom over one, maybe 30-40%?

In the end, I went ahead and did it on a twin EDF with no problems. ESCs are the same brand/model, so the output should be roughly matched. A single BEC would have been sufficient...I just did it cause I could.

D.
darkith is online now Find More Posts by darkith
Reply With Quote
Old Oct 15, 2009, 05:54 PM
Registered User
North East England UK
Joined Jan 2003
1,432 Posts
Quote:
Originally Posted by darkith
I suppose it could be a serious problem on poorly matched BECs, where one might be handling the full load and heat up to the point of shutting down before the other one kicks in, then it takes the full load and gets to the shutdown point.
D.
Very unlikely as the output from each bec is going to be within a 100mv of each....

Or put another way...if the voltage difference between each bec is greater than 100mv then no issue...even that is conservative
Chippie is offline Find More Posts by Chippie
Reply With Quote
Old Oct 15, 2009, 10:05 PM
Registered User
43851E's Avatar
College Station, TX
Joined Jul 2009
882 Posts
If one placed a battery source between each B+ & negative, perhaps 4.8V (very small pack 4 1.2v NiNh AAA or a 2s LiPo) before they go to the RX and parallel to the RX, I believe the BEC outputs would then be slightly loaded charging the 4.8v cell and then the VA curve would be matched and the output of the BECs would be keeping the "reservoir" as it were (battery, very small) charged.

I know, I know, it's a "Bi Partisan" kind of thing, but it would work great.
I deleted the yellow wire for drawing simplification. It will need to be connected together & routed to the center pin. Drawn with MS Visio in 10 minutes.

Am I crazy?



Doug
43851E is offline Find More Posts by 43851E
Reply With Quote
Old Oct 15, 2009, 11:31 PM
The Fixer
Martin Y's Avatar
Canada, ON, Milton
Joined Jan 2009
3,767 Posts
If you look at the castle creations ESC's, they use 2 linear regulators in parallel. So doubleing or trippling will be OK.
Martin Y is online now Find More Posts by Martin Y
RCG Plus Member
Latest blog entry: Flyfly Fox Glider
Reply With Quote
Old Oct 16, 2009, 02:48 AM
Registered User
Joined Apr 2007
73 Posts
Quote:
Originally Posted by Greywing
Therefore the load will be serviced by the output with the highest voltage.
If you connect two regulators in parralel, they will be at the exact same voltage.
slippyr4 is offline Find More Posts by slippyr4
Reply With Quote
Old Oct 16, 2009, 03:23 AM
Registered User
Staffs, UK
Joined Nov 2003
10,550 Posts
I have for many years recommended NOT connecting two linear regulators directly in parallel, mainly because that's what the manufacturers of the devices recommended (in particular both National Semiconductor and Texas Instruments had it as a real no-no for 78xx devices). Of course it's easy to solve in a BEC with correct circuit design and some BEC manufacturers (like Castle Creations) have long done this.

And how valid the worry still is for the more modern LDO regs I don't really know but I can't seen any evidence of the semiconductor manufacturers still publishing dire warnings about it .

Steve
slipstick is online now Find More Posts by slipstick
Reply With Quote
Old Oct 16, 2009, 03:32 AM
Registered User
Staffs, UK
Joined Nov 2003
10,550 Posts
Quote:
Originally Posted by slippyr4
If you connect two regulators in parralel, they will be at the exact same voltage.
Yes that's right, one will be at its natural voltage and the other will have been forced to operate at a voltage which is unnatural for it. The resultant strain on the regulation circuit is what causes the imbalance in loading (over-simplifying horribly ).

A couple of low value resistors between the two outputs allowing each device to operate at its natural voltage even though the combined circuits obviously can have only one output voltage solves some problems...but not all ESC manufacturers build their BECs to take account of that .

Steve
slipstick is online now Find More Posts by slipstick
Reply With Quote
Old Oct 16, 2009, 04:53 AM
Registered User
Joined Aug 2009
60 Posts
Quote:
Originally Posted by slipstick
Yes that's right, one will be at its natural voltage and the other will have been forced to operate at a voltage which is unnatural for it.
Both would be at a voltage between the two values

Let me explain with an example.

When used as voltage source, for maintaining a regulation of 0.5V (5%) at 1A load, a 5V regulator must have an output impedence less than 250mOhm.

Lets say there are two 5V regulators, one with o/p voltage 4.9V and other with 5.1V. Lets say their o/p impedence is 250 mOhm each.

If you connect these two devices in parallel, in steady state, there would be 400 mA (= 0.2 Volt, 500mOhm) current flowing between the two devices (yes, one regulator feeding other: fighting?), and the drop on each regulator would be 0.1 V thus output would be 5.0V

If the regulator is rated for higher currents, the impedence would be even lesser, and such a current may easily exceed the ratings of regulator because there is nothing except the regulator's internal impedence to limit the current.

This basic analysis models the regulators as voltage source with finite internal resistance. This is not 100% accurate, but it is close enough to recommend not connecting two voltage sources in parallel. Personally, I'd prefer wiring them supplying to separate loads.
supergps is offline Find More Posts by supergps
Reply With Quote
Old Oct 16, 2009, 05:24 AM
Registered User
Joined Apr 2007
73 Posts
supergps made the correct point there.

joining simple, fixed output linear regulators without due consideration to circuit design isn't wise.

however, adjustable regulators can be joined in a carefully designed circuit; set the adjustment to be a feedback function of the output of the regulators and you can tune them to be identical.
slippyr4 is offline Find More Posts by slippyr4
Reply With Quote
Old Oct 16, 2009, 11:13 AM
Registered User
Staffs, UK
Joined Nov 2003
10,550 Posts
And I've never seen an ESC which uses adjustable regs for the built-in BEC so we're still saying that it may not be a good idea to connect two linear BECs in parallel. But it depends on the BEC circuit design because there may be more to the BEC circuit than JUST a simple linear V.Reg. .

Steve
slipstick is online now Find More Posts by slipstick
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Parallel two bec's from speed controller callen Castle Creations 8 Sep 12, 2008 02:36 PM
Discussion two bec's in parrallel BOBBY1017 Power Systems 3 Feb 16, 2008 08:22 AM
Big eHelis: Running BEC's in parallel? pwkpete Electric Heli Talk 14 Jun 24, 2005 10:46 PM
Charging Kok's in Parallel genenjr420 Batteries and Chargers 22 Aug 06, 2003 06:40 AM
Hooking two PC PS's in parallel to double A?? foss1997 Power Systems 9 Apr 04, 2002 01:10 AM