Searching \ for '[EE] Diode ORing supplies' in subject line. ()
Help us get a faster server
FAQ page: massmind.org/techref/index.htm?key=diode+oring+supplies
Search entire site for: 'Diode ORing supplies'.

Exact match. Not showing close matches.
'[EE] Diode ORing supplies'
2009\04\09@123445 by

Typically if you have two power sources, you diode OR them, so they will not backfeed to each other.  Application is battery backed systems, etc.

If you have one supply, say the battery sitting on a trickle charger, and aother wall supply, both feeding a power plane,and if one supply is at a slightly higher potential, will it draw more current from the lower potential supply in order to normalize the voltage on the plane?

alan smith wrote:
> Typically if you have two power sources, you diode OR them, so they will
> not backfeed to each other.  Application is battery backed systems, etc.
>
> If you have one supply, say the battery sitting on a trickle charger, and
> aother wall supply, both feeding a power plane,and if one supply is at a
> slightly higher potential,

So you basically have THREE power supplies, all feeding the same circuit?
Trickle charger, battery, and the wall supply?

And the "ground" levels of the trickle charger and the wall supply, are
slightly different? So like, the trickle charger's Gnd is at 0V, and the
wall supply's Gnd is at +0.5V?

> ..will it draw more current from the lower potential supply in order to
> normalize the voltage on the plane?

Either I don't understand the question, or the question is wrong. The power
supplies will be fighting each other, so they will both be supplying more
current than is required by the circuit.

Vitaliy

The grounds are all connected together. So that is of no concern.
The highest voltage in front of a diode will provide the major current for
the application.
The other diodes will be back biased if the highest voltage is a few tenths
more than the rest.
You can soften each up by current limiting it with a resistor in series with
each diode.

Voltage critical application will probably need voltage regulation down
stream?

{Original Message removed}

Two supplies...actually.  That as an example.  Two supplies, one off the wall, one from a UPS.  Both identical switching supplies, but due to a cable length, one supply may be at a lower potential than the other by around a volt or two.  The grounds are all tied together on the board.  They shouldnt fight each other due to the diode..I think.

In the past where I tied multiple supplies to a single power plane/bus bar system, the outputs of each were diode or'ed but they were all at the same output voltage.

--- On Thu, 4/9/09, Vitaliy <spammaksimov.org> wrote:

{Quote hidden}

> --
alan smith wrote:

> In the past where I tied multiple supplies to a single power
> plane/bus bar system, the outputs of each were diode or'ed but they
> were all at the same output voltage.

The diode of the supplies with lower voltage will be reverse biased.

Gerhard
Michael Hagen wrote:
> The grounds are all connected together. So that is of no concern.

Are you sure? It seems to follow that if one "ground" is a couple of volts
above the other "ground", there will be current flow from one to the other.

Vitaliy

True...if the grounds are tied back, or referenced to earth ground, right? but switching supplies are rather virtual grounds I believe, so if you reference both switching supplies together, then they become the same reference point...rather a star grounding scheme.

So if one supply..same reference...is actually a volt to maybe two volts different, then what...I am thinking that they still will want to equalize the potential between them?

--- On Thu, 4/9/09, Vitaliy <spammaksimov.org> wrote:

{Quote hidden}

> --
alan smith wrote:
> True...if the grounds are tied back, or referenced to earth ground, right?
> but switching supplies are rather virtual grounds I believe, so if you
> reference both switching supplies together, then they become the same
> reference point...rather a star grounding scheme.
>
> So if one supply..same reference...is actually a volt to maybe two volts
> different, then what...I am thinking that they still will want to equalize
> the potential between them?

I am not an expert on switching supplies, but there's got to be *some*
current flowing b/w the grounds, to equalize them, no? Maybe it is small
enough to be negligible, but wouldn't it still be there?

Vitaliy

Thats the issue...I think when they are close, physically close, so the point of load or even distribution is a very short connection (wire or trace) there is a mininal about of differential, such that any ground currents is nearly nothing.  Its when there is a large differential and the supplies (and return paths to ground) are long, then you start having problems.  As a result of this discussion, even me typing out responses, has convinced me that I have to solve this a different way.

--- On Thu, 4/9/09, Vitaliy <spammaksimov.org> wrote:

{Quote hidden}

> --
Something is getting confused in this discussion. Can you perhaps draw
a quick schematic of what you have in mind?

Sean

On Thu, Apr 9, 2009 at 4:55 PM, alan smith <micro_eng2yahoo.com> wrote:
{Quote hidden}

>> -

yeah..give me a day and I will.  So many issues, so little time in the day (anyone else work 18 hour days as well? haha)

--- On Thu, 4/9/09, Sean Breheny <shb7cornell.edu> wrote:

{Quote hidden}

> Its when there is a large differential and the supplies (and return paths
> to ground) are long, then you start having problems

A project I spent an awful lot of time getting right has been completely
fudged up at the remote installation. The symptoms they reported to me
just didn't make sense compared with the faultless performance here

So I got *their* whole thing back yesterday. Ay caramba. Not a common
ground in sight. I measure 0.8V and 150mA between section grounds. No
wonder it trips whenever a light or a motor is turned on or something is
(dis)connected or anyone looks at it funny
alan smith wrote:
> So many issues, so little time in the day (anyone else work 18 hour days
> as well? haha)

No, not lately -- but I've done it as recently as a few weeks ago. :-)

Take it easy. As the Russian proverb has it, "work is no wolf, it won't run
off into the forest".

;-)

Vitaliy

----- Original Message -----
From: "Michael Hagen" <picdogdslextreme.com>
To: "Microcontroller discussion list - Public." <piclistmit.edu>
Sent: Thursday, April 09, 2009 10:03 AM
Subject: Re: [EE] Diode ORing supplies

{Quote hidden}

>> --
I am trying to resend this, it never made it the first times?

----- Original Message -----
From: "Michael Hagen" <picdogdslextreme.com>
To: "Microcontroller discussion list - Public." <piclistmit.edu>
Sent: Thursday, April 09, 2009 6:37 PM
Subject: Fw: [EE] Diode ORing supplies

{Quote hidden}

Michael Hagen wrote:
>I am trying to resend this, it never made it the first times?

It made it to the list every single time.

>
> {Original Message removed}

More... (looser matching)
- Last day of these posts
- In 2009 , 2010 only
- Today
- New search...