HobbyKing.com New Products Flash Sale
Reply
Thread Tools
Old Jan 19, 2013, 04:29 AM
Oxford Panic
AndyOne's Avatar
United Kingdom, Oxford
Joined Feb 2003
3,659 Posts
There was a big thread about regulators in parallel here a while ago and it may still be going on such was the controversy.
Engineers know only too well that if doesn't say you can in the datasheet, then if you do, there will be some horrible unforeseen consequence. Those of us who have done any kind of design work have fallen at this fence at least once.
Having said that it does seem common practice to see two or even three regulators lined up next to each other on some ESCs. Even though the idea of doing this goes against the grain it seems to be so common these days that it has become mainstream practice.
AndyOne is offline Find More Posts by AndyOne
Reply With Quote
Sign up now
to remove ads between posts
Old Jan 26, 2013, 12:30 PM
Simple, Easy, Fun
Gabe_BigApple's Avatar
United States, NY, New York
Joined Jan 2013
558 Posts
Quote:
Originally Posted by dave1993 View Post
... most here these days acknowledge parallel lipos is ok but the regulator myth lives on. ...
Dave,
Is this why they tell you lift the positive pins on 3 of 4 ESCs in a quad?

It never made any sense to me but I never took the time to look into it.

Cheers,
Gabe
Gabe_BigApple is online now Find More Posts by Gabe_BigApple
Reply With Quote
Old Jan 28, 2013, 05:48 PM
RC beginner
New York
Joined Oct 2008
6,054 Posts
oh yeah... i love that advice to lift pins which actually does little more than reduce system current capability big time. with subsequent risk of brownout. the very thing they think they are avoiding. specially common in the multi area.

a friend of mine had great disagreements on the subject here at rcg. he seemed to know what he was talking about but just to verify i engaged in my own research. sure enough if one takes time to learn about linear regulator design and semiconductor physics it turn out to be quite silly thinking you cant parallel them. this is aside from the fact that virtually all escs have been doing this from day one w/o a problem.

naturally some will chime in with stories of how it caused their house to catch fire and take the whole neighborhood with it. there are many similar universal wives tales on these forums and its entertaining to see how few will be dissuaded from their belief systems. "better than tv" and one of the main reasons i keep coming back here. lol!

ps welcome to the monkey house. this is a great place for fun and education. i get both on daily basis.
dave1993 is offline Find More Posts by dave1993
Reply With Quote
Old Jan 29, 2013, 03:04 AM
Registered User
Staffs, UK
Joined Nov 2003
10,561 Posts
Quote:
Originally Posted by dave1993 View Post
a friend of mine had great disagreements on the subject here at rcg. he seemed to know what he was talking about but just to verify i engaged in my own research. sure enough if one takes time to learn about linear regulator design and semiconductor physics it turn out to be quite silly thinking you cant parallel them.
There's nothing wrong with a bit of amateur experimenting. But it's not going to convince me that the amateur in question now understands more about linear regulator design and semiconductor physics than e.g. National Semiconductors and Texas Instruments who designed and made most of the linear regs in use and who both warned about parallelling them. Particularly the original 78XX which had no design allowance for parallel connection, unlike the more modern designs.

BTW parallelling lipos (which are not electronic circuits) has nothing to do with it and has always been o.k. Why else would many of the early lipos have configurations like 3S2P ? Guess what the "P" stands for .

Steve
slipstick is offline Find More Posts by slipstick
Reply With Quote
Old Jan 29, 2013, 03:59 AM
Registered User
United States, ID
Joined Sep 2011
501 Posts
jakestew is online now Find More Posts by jakestew
Reply With Quote
Old Jan 29, 2013, 11:29 AM
RC beginner
New York
Joined Oct 2008
6,054 Posts
Quote:
Originally Posted by dave1993 View Post
its entertaining to see how few will be dissuaded from their belief systems. "better than tv" and one of the main reasons i keep coming back here.
lol!
dave1993 is offline Find More Posts by dave1993
Reply With Quote
Old Jan 29, 2013, 01:28 PM
Registered User
Staffs, UK
Joined Nov 2003
10,561 Posts
Yep , you've decided what you think and you don't like people trying to confuse you with inconvenient facts .

It was reputed to be Plato who first said "I'm trying to think, don't confuse me with facts" so at least you're in good company .

Steve
slipstick is offline Find More Posts by slipstick
Reply With Quote
Old Jan 29, 2013, 02:13 PM
Who let the dogs out?
phil_g's Avatar
Pontefract, Yorkshire, UK
Joined Jul 2007
1,086 Posts
Quote:
Originally Posted by dave1993 View Post
...just to verify i engaged in my own research. sure enough if one takes time to learn about linear regulator design...
Erm... its 2013 Dave, its a little late to start learning about linear regulator design...

A 7805 has a specified tolerance of 4%, which is from 4.8v to 5.2v
If you happen to parallel a particularly low one (4.8v) with a particularly high one (5.2v), then the voltage will stabilize at the high level (5.2v) and the low one will not contribute at all, until such times as the output is loaded sufficiently to drop to 4.8v because the 7805 output is an emitter follower and the emitter of the low one is being held above its regulation voltage by the other regulator.
This is all well documented in the device datasheets Dave, its nothing to do with 'belief'. Paralleling mismatched regulators may be fine but the manufacturers do not recommend it as the individual regulators will not share the current demand equally - one could be approaching its disipation limit whilst the other hasnt started to contribute.
Also some LDO regulators are particularly sensitive to anything affecting the control loop (eg another regulator across its terminals) and can spuriously oscillate if a shared output capacitor is insufficient or too high ESR for paralleled regulators. Again in a paralleled pair one will dominate whilst the other is doing nothing, until the 'high' one reaches its foldback current limit and its output voltage drops to that of the paralleled regulator.
Conclusion - you can parallel them but the manufacturers dont recommend it because it doesnt achieve what some people expect it to achieve.
Quote:
Originally Posted by dave1993 View Post
ps welcome to the monkey house. this is a great place for fun and education. i get both on daily basis.
yeah what do we monkeys know, eh? just nonsense we've learned from datasheets.

Cheers
Phil
phil_g is online now Find More Posts by phil_g
Last edited by phil_g; Jan 29, 2013 at 02:43 PM.
Reply With Quote
Old Jan 29, 2013, 02:27 PM
Stuart
srnet's Avatar
UK, Cardiff
Joined Dec 2008
3,274 Posts
Quote:
Originally Posted by dave1993 View Post
i engaged in my own research. sure enough if one takes time to learn about linear regulator design and semiconductor physics it turn out to be quite silly thinking you cant parallel them.
I seem to recall that the original designers of the common regulators do not recommend this practice.

Why do they do this ?
srnet is offline Find More Posts by srnet
Reply With Quote
Old Jan 29, 2013, 02:52 PM
RC beginner
New York
Joined Oct 2008
6,054 Posts
Quote:
Originally Posted by phil_g View Post
yeah what do we monkeys know, eh? just nonsense we've learned from datasheets.
not much apparently. i checked a few datasheets from most popular mfgs and a quick pdf search reveals no reference at all regarding parallel anything let alone a warning. a couple for your perusal:

http://www.fairchildsemi.com/ds/LM/LM7805.pdf
http://www.ti.com/lit/ds/symlink/lm340-n.pdf

im sure you may be able to dig up some oddball app note or rcg rave warning against it. but dont bother. as mentioned i will NOT be confused by those "facts". interestingly steve is implying that being confused by facts is a good thing. each to his own but i try to avoid confusion of any type.

again i will point to the millions (no exaggeration) in use this way as we speak with no hint of misbehaviour. again, dont bother dredging up tragic tales of woe as i will not be impressed. i could care less if you tell me your auntie burned down the family farm with a pair of lm340s.

the thing i like about this place is there will always be somebody here who will argue to chase away the loneliness. and best of all... no ir remote required.

dave1993 is offline Find More Posts by dave1993
Reply With Quote
Old Jan 29, 2013, 03:01 PM
Who let the dogs out?
phil_g's Avatar
Pontefract, Yorkshire, UK
Joined Jul 2007
1,086 Posts
Quote:
Originally Posted by Texas Instruments
Due to slight differences in the internal Reference, one regulator will dominate all the current when 2 or more are paralleled. Once the dominant regulator has reached its foldback current limit, or the voltage drop across its pass element (I x Rdson) causes its output voltage to drop, another parallel device will start providing power. Many regulator's have a shutdown current limit and not a foldback. If a shutdown current limit is reached that regulator will turn off and not reduce the output voltage (foldback) to limit the current. In this case direct paralleling will not work. One method to force parallel operation is to put a resistance in series with each output, to compensate for the reference offsets Rdson variables. For instance a 2Amp regulator with a reference tolerance of 2.5+-.1V will need a resistor = 0.2V/2A or 0.1ohm in series to help it share. If one regulator has a reference of 2.6V, at 2A the drop of the .1 ohm resistor will be .2V, which the other regulator will see as 2.4V, and since this is the lower limit of the reference it should have started regulating by the time the output dropped to 2.4V. It is usually preferable to choose a linear regulator suited for the output current needed.
But what do TI know, they're all idiots eh Dave.
phil_g is online now Find More Posts by phil_g
Reply With Quote
Old Jan 29, 2013, 03:25 PM
Stuart
srnet's Avatar
UK, Cardiff
Joined Dec 2008
3,274 Posts
Quote:
Originally Posted by dave1993 View Post
I checked a few datasheets from most popular mfgs and a quick pdf search reveals no reference at all regarding parallel anything let alone a warning
What a completely daft argument.

Most rational persons would assume that if a datasheet did not explicitly define and allow a practice its not an approved mode of operation.

What are you expecting data sheets to be, a couple of paragraphs of real data and then several hundred pages of daft things not to do ?
srnet is offline Find More Posts by srnet
Reply With Quote
Old Jan 29, 2013, 03:26 PM
RC beginner
New York
Joined Oct 2008
6,054 Posts
Quote:
Originally Posted by phil_g View Post
But what do TI know, they're all idiots eh Dave.
not all but i suspect a few. looks like you were careful to avoid linking that quote because even casual glance shows it is not referencing the 7805 device. quote: "For instance a 2Amp regulator".

and even if it was referring to 7805 its not at all surprising to see the line:

quote: "It is usually preferable to choose a linear regulator suited for the output current needed."

considering profit margins for high amp regulators are 10x-20x that of the generic 1a device. im trying to imagine these guys actually explaining how to cut into their bottom line.... NOPE, cant do it.
dave1993 is offline Find More Posts by dave1993
Reply With Quote
Old Jan 29, 2013, 03:29 PM
RC beginner
New York
Joined Oct 2008
6,054 Posts
Quote:
Originally Posted by srnet View Post
What a completely daft argument.

Most rational persons would assume that if a datasheet did not explicitly define and allow a practice its not an approved mode of operation.

What are you expecting data sheets to be, a couple of paragraphs of real data and then several hundred pages of daft things not to do ?
lol! i never expected to see any reference at all to parallel use because most professionals know the truth. it was steve and posse that brought up the whole subject of datasheets. i was simply showing that his statement about them was untrue. lets just say did not help establishing expert opinion. personally im stuck here in the real world.

ps use for rc electronics is also not mentioned. so by your "rational" reasoning we can "assume" (theres that word again) this is absolutely verboten too.
dave1993 is offline Find More Posts by dave1993
Last edited by dave1993; Jan 29, 2013 at 03:37 PM. Reason: if it aint mentioned you just cant do it... yeah right!
Reply With Quote
Old Jan 29, 2013, 03:39 PM
RC beginner
New York
Joined Oct 2008
6,054 Posts
its 10x better when they gang up. lol!
dave1993 is offline Find More Posts by dave1993
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Build Log DIY Headtracker (Easy build, No drift, OpenSource) Dennis Frie FPV Equipment 2395 Oct 20, 2014 11:12 PM
Discussion my very diy long range bixler (16.7km on 5.8GHZ) quix-fz FPV Talk 40 Aug 20, 2012 02:56 AM
Discussion any easy diy to measure thrust chuckinnc Power Systems 5 Jun 06, 2012 02:13 PM
Help! Diy bec jvpercia DIY Electronics 2 Jun 06, 2012 03:22 AM
Cool Very Easy DIY 4G6 FB -to- FBL head conversion with 3D and crash tests! Timewerx Micro Helis 5 Aug 11, 2010 02:57 AM