Searching \ for '[EE]: Combining voltages from different sources' in subject line. ()
Help us get a faster server
FAQ page: www.piclist.com/techref/power.htm?key=voltage
Search entire site for: 'Combining voltages from different sources'.

Exact match. Not showing close matches.
'[EE]: Combining voltages from different sources'
2002\07\23@174424 by

Let's say that I have two 12V 1.3A batteries. If I connect the positives and
the negatives together, I get the same voltage but longer run time (I
think).  If I connect battery 1 negative to battery 2 positive and the
voltage diff between bat1 pos and bat2 neg is 24V.

My question is this:  What happens if I connect battery 1 to battery 2 (pos
to pos, neg to neg) but with about 100meters apart and when there are
circuits between them?  Like this:

bat1                             bat2
+ ---*----------*--------  +
-  ---|-*---------|--*----  -
|  |              |   |
|  |              |   |
circuit 1      circuit 2

(Lousy drawing!)

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

> Let's say that I have two 12V 1.3A batteries. If I connect the
> positives and the negatives together, I get the same voltage but
> longer run time (I think).  If I connect battery 1 negative to
> battery 2 positive and the voltage diff between bat1 pos and bat2
> neg is 24V.
>
> My question is this:  What happens if I connect battery 1 to
> battery 2 (pos to pos, neg to neg) but with about 100meters apart
> and when there are circuits between them?  Like this:

Well, that all depends on how much resistance there is in your wire.
Then you'll have to look at thevenin equivalents for the two
circuits.  If you can measure the resistances, then I'm sure that we
can work some data out for you.

- --Brendan

-----BEGIN PGP SIGNATURE-----
Version: PGPfreeware 6.5.8 for non-commercial use <http://www.pgp.com>

iQA/AwUBPT3QbAVk8xtQuK+BEQL9gQCgsryfL/M2muqMzYW+Ll/20BS60awAoKEW
48pVMKD4+q62031n+5Xx0uID
=OsZQ
-----END PGP SIGNATURE-----

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.

Omer,
What you intend to do, usualy is not used for some reasons:

-- 100m at 12V is too much except for small curent sinking/sourcing
-- any battery/accu may have local chargers

So assuming you have an alarm system or something else, if one battery is
dying the other one will can't support the load because of the long
distance. Use the Ohm law on a full circuit ( including battery internal
resistance) to compute the result aplying also the superposition theorem
( consider battery1 disconected and battery2 functioning, then battery1
functioning and battery2 disconected, the final result is a sum of those
two condition ). For line, use the resitor theorem: R = ro x l/s, where ro is
the resistivity ( material dependent ), l is the wire lenght ( x2 ), s is
the total crossower section of the wire ( add all independent sections for
a multiwire cable ). Computing the voltage result on any wire lenght,
measured from battery1 or battery2 on a known load is also possible.

Vasile

On Wed, 24 Jul 2002, [windows-1254] Ömer Yalhý wrote:

{Quote hidden}

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listservmitvma.mit.edu with SET PICList DIGEST in the body

Well, what I have not told you is that the circuits have one component the PIC 16f628 2 resistors and a 100nF cap.  The voltage from bat1 and bat2 comes regulated to 5V with 7805.  My concern is that doing what I have asked, would it damage the whole system in a way, short circuit maybe or something electrically wrong that I might not know.  Also, if there is a voltage drop after 100m from bat1 to last circuit to 3V (I know there won't be that much of a drop but let's say), again doing what I have asked, would it cancel the voltage drop and supply 5V to the whole system.

To make things short, this is what I am trying to do: circuit 1 is powered from bat1 and there is long comm line, and at the end, there is circuit 2 (same circuit).  Now, if the comm line is too long so that circuit 2 does not receive info correctly I need to install another circuit in between (like a repeater), this circuit will be powered from bat2 and will transmit any info that came from circuit1 to circuit2.  In this model, bat1 and bat2 pos as well as bat1 neg and bat2 neg line will be connected together (12V from bat, regulated to 5V with 7508 then powers the circuits).  Would this be wrong, electrically speaking?

Wow, hope I did better explaining this time.

Thank you for taking the time and answering my questions.  I really appricate it and it's always nice to know there are very nice people out there trying to help and with no expectations.  Do not take this granted, I don't.

Ã–mer YALHI
oyalhiteksan.com.tr
http://www.teksan.com.tr
Tel : +90 212 613 22 00
Fax: +90 212 544 70 35

{Original Message removed}
In the known circumstances:

a. if the circuit is sending and receiving datas on the same power lines
( +12V and Gnd ) I'll use current supply for the secondary unit, so the
second battery is not necessary.
b. If the a. is not applicable then I'll suply both microcontroller boards
with local batteries/accumulators and use a standard RS485 line which is
fine for 100m. The batteries doesn't need to be connected together.
But the "remote" unit acumulators must be charged somehow.
c. if you have another situation ( which ? ) a. and b. aren't good for you

best, Vasile

On Wed, 24 Jul 2002, [utf-8] Ã–mer YalhÄ± wrote:

{Quote hidden}

> {Original Message removed}
First of all I am actually more of a electronics programmer than a electrical guru so I have no idea what a current supply is :-)   For now I am stuck with rs-232 and with my tests, lengths up to 305m are ok (cat 5 cable).  However, just to make sure I would like to use repeaters every 100m.  So it would be like:

bat1 12V --->  7805 5V ----> circuit1 ---------100m---------  repeater (with another bat2/7805 combo) ------------100m----------- repeater (repeater (with another bat3/7805 combo) ---------100m-------- circuit2

In this setting, bat1, bat2 and bat3 will all be connected together, (so that the whole systems power line is connected together - common power line as well as each bat powering the circuit it is connected to) would this create some kind of short circuit?

P.s. Whole cable consists of 4 wires.  Power, ground, tx and rx.  And all powers are connected together, all grounds are connected together, and tx's and rx's.  Each battery is connected to a 220V power line and constantly charging (lead battery).

Ã–mer YALHI
oyalhiteksan.com.tr
http://www.teksan.com.tr
Tel : +90 212 613 22 00
Fax: +90 212 544 70 35

{Original Message removed}
please forgot everything you've told me in your last mail,
then, after your mind is empty and open, take a look at:

http://www.national.com/an/AN/AN-1057.pdf
http://www.arcelect.com/485info.htm.
http://www.geocities.com/SiliconValley/Peaks/4125/myfinal.htm

then choose a RS485 pair ( emiter + receiver ) SN75156 ?
convert RS232 to 485 and done without repeaters.

Sasha has some great experience with RS485 :o)

succes,
Vasile
http://www.geocities.com/vsurducan

On Wed, 24 Jul 2002, [utf-8] Ã–mer YalhÄ± wrote:

{Quote hidden}

> {Original Message removed}
On Wed, 24 Jul 2002 18:06, you wrote:
> please forgot everything you've told me in your last mail,
> then, after your mind is empty and open, take a look at:
>
> http://www.national.com/an/AN/AN-1057.pdf
> http://www.arcelect.com/485info.htm.
> http://www.geocities.com/SiliconValley/Peaks/4125/myfinal.htm
>
> then choose a RS485 pair ( emiter + receiver ) SN75156 ?
> convert RS232 to 485 and done without repeaters.
>
> Sasha has some great experience with RS485 :o)
>
> succes,
> Vasile
> http://www.geocities.com/vsurducan

Hi all,

Also those of us dealing in industrial automation (including DMX 512 ;o) have
great experiance in long line comms using RS 485 and RS 422.

I have a design using polled comms with upto 200 units online running at
9600bps with a loop length over 10Km (using shielded twisted pair) At a
Higher  baud rate loop length must be reduced, I have tested the system to
19K2 but I think from memory this was the limit of the driver and reciever
chip slew. It breaks a few design rules, but is used in another product which
is in service all over the world, in various configurations (star wired,
loop, with fewer units and shorter line lengths). This system has been in use
for over 20 years now ( which explaines the funny parts selection ;o) I may
update the design to use newer drivers :o)

Regards,

Kat.

--
--------------------------------------------------------------------------------------------
K.A.Q. Electronics
Software and Electronic Engineering
Perth, Western Australia
Phone +61 (0) 419 923 731
--------------------------------------------------------------------------------------------

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listservmitvma.mit.edu with SET PICList DIGEST in the body

Katinka,

It will be very nice from you if we can see your nice project.
And I'm talking maybe about a web tech page, you know, the schematic of
one module, the theory of long line data transfers, why a twisted pairs
is better than a coaxial one, maybe a  few words about the PIC routines
( RS485 ), implementing your own protocol ( interogation, avoiding bus
colision ) and stuff like this... in a more detalied manner like the
secondary web address below.

best regards,
Vasile

"even if you are good it doesn't worth a penny if the other peoples don't
know that"  unknown author

On Wed, 24 Jul 2002, Katinka Mills wrote:

{Quote hidden}

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listservmitvma.mit.edu with SET PICList DIGEST in the body

Kat.  Are you offering help or service?
How can we check out the design?

Ömer YALHI
oyalhiteksan.com.tr
http://www.teksan.com.tr
Tel : +90 212 613 22 00
Fax: +90 212 544 70 35

{Original Message removed}
On Wed, 24 Jul 2002 20:30, you wrote:
> Katinka,
>
> It will be very nice from you if we can see your nice project.
> And I'm talking maybe about a web tech page, you know, the schematic of
> one module, the theory of long line data transfers, why a twisted pairs
> is better than a coaxial one, maybe a  few words about the PIC routines
> ( RS485 ), implementing your own protocol ( interogation, avoiding bus
> colision ) and stuff like this... in a more detalied manner like the
> secondary web address below.
>
>
> best regards,
> Vasile
>
> "even if you are good it doesn't worth a penny if the other peoples don't
> know that"  unknown author

Lol, well for starters the current code is in an uDP1802B (mmm shame I can not
buy them any more, or the fact my z80 s100 bus computer died, was an
excerlent cpm machine)

I would need to update it lol.

Collison detection was easy, MASTER computer (one of my loverly Dastra boxes
from the USA (cost ~us\$10K in 1980) sent a polling command out, it would ask
unit one are you here ? then it would wait before going to unit 2 etc up till
the maximum unit # programmed in the master. when a unit has a request for
information, it waits untill it is polled (polling was continuous and could
poll 200 units in under 20 seconds) then during the polling was a window
where the slave could send data too (an event happened) or request data from
the host. the time was sent on the minute every minute as a way of keeping
all the software RTC's in sync (did not have to worry about how close the
2.48Mhz clock was ;o)

I will when I get the time update this, but bewarned it will be in avr ASM as
I do not have the time or development gear to do pic stuff atm (esp not when
an ice200 for the mid range atmel AVR chips is now us\$99.

Regards,

Kat.

--
--------------------------------------------------------------------------------------------
K.A.Q. Electronics
Software and Electronic Engineering
Perth, Western Australia
Phone +61 (0) 419 923 731
--------------------------------------------------------------------------------------------

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listservmitvma.mit.edu with SET PICList DIGEST in the body

On Wed, 24 Jul 2002 21:11, you wrote:
> Kat.  Are you offering help or service?
> How can we check out the design?
>
>
> Ömer YALHI
> oyalhiteksan.com.tr
> http://www.teksan.com.tr
> Tel : +90 212 613 22 00
> Fax: +90 212 544 70 35

Hi Omer,

I can do the hardware design no problems, but as I do not use pics that often, If I was to offer software aswell it would need to be for an AVR.

If you would like to talk further for me to subcontract, please feel free to contact me off list. If you would like to pick my brains about RS422 / RS 485 long line comms, please feel free to do it on list as it is then archived for others :o)

Regards,

Kat.

-- --------------------------------------------------------------------------------------------
K.A.Q. Electronics
Software and Electronic Engineering
Perth, Western Australia
Phone +61 (0) 419 923 731
--------------------------------------------------------------------------------------------

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listservmitvma.mit.edu with SET PICList DIGEST in the body

I did the same with the time checks when using
slaves with software clocks.

My challenge was that at any time the master could
go off line, or one or more slaves could loose
power.  When the slaves lost power, their software
clocks would not be running and they would have a
earlier time when they were powered back on.

The solution was to add a "Figure Of Merit" (FOM)
to each time message sent.  The master, having a
RTC, was the standard with a FOM of 0. (love it
when I make up my own three letter words)  If a
slave received a FOM better (lower) than it's own,
then it would set it's software clock to the time
and set it's FOM to one more than the received
FOM.

When the master went off line, one of the slaves
with the lowest FOM would start transmitting time
checks for the rest of the slaves.  When one of
the slaves were powered down and back up, it would
use the time from the transmitting slave and set
it's FOM to one more than the FOM of the time

So as long as the master and all of the slaves
were not powered down at the same time, a "good"
time clock was kept.

Bill

{Original Message removed}
how many slaves have you Bill ?
As I had understood, one of your slaves take the master's place when
this one was shutted off, but just for time checks and no more else.
Is that a better methode than polling the slaves ( small numbers of slaves
), the slave which does not respond is passed on an "error list" Succesive
pollings on that slave without succes it means an off/broken slave.

regards,
Vasile
http://www.geocities.com/vsurducan

On Thu, 25 Jul 2002, Bill & Pookie wrote:

{Quote hidden}

>   {Original Message removed}
Could have up to 30 slaves, and the master did all
the polling.  And all the slaves would transmit
was the time when the master was not on line
transmitting it.

If any slave didn't receive a time that was as
good or better than his in a certain time, then he
would start transmitting his time as often as the
master would.  The slave with the best time would
be transmitting and all others would listen.

The slaves ram was battery backed, so they could
log data for a few days without loosing any if the
master was not on line polling for the data.

And during that time, the individual slaves might
be powered down and that would stop their software
clocks from keeping time while no power.  So when
powered up again, that slave's clock would start
from where it left off timing when powered down.
Then get good time from the master or another
slave.

****

As long as a slave was receiving time packets that
were as good or better than his, all was well with
that slave.

When a slave didn't receive a time packet within a
certain interval that was as good or better than
his, he would start sending his time at regular
intervals.  And stop only when he received a time
that was equal or better than his.

The interval a slave would wait before starting to
send his time was 3 times the interval the master
sent time plus a random value. (random value was
to keep slaves from having line contention.)

Bill

{Original Message removed}
The problem with connecting power supplies in paralel but with long
cables between them is that if, say, circuit #1 (close to battery #1)
drains more current (power) than circuit #2 (close to battery #2) then
part of this current difference will come from battery #2, thus
generating a current supply along the wires, and this current will
generate a voltage difference between the wire taps. In fact, circuit #1
will not see the full voltage from battery #2, as this voltage drop will
apear in both wires.

A better method is to leave the positives opened, connecting only the
grounds. And of course you would need a local battery charger, as

But I agree to the best method: only one power supply and a differential
comm link as RS485 or 422.

Francisco

C mer YalhD1 wrote:

{Quote hidden}

>{Original Message removed}
Ok, now I got it.

Thank you Francisco.

Ömer YALHI
oyalhiteksan.com.tr
http://www.teksan.com.tr
Tel : +90 212 613 22 00
Fax: +90 212 544 70 35

--
http://www.piclist.com hint: To leave the PICList
piclist-unsubscribe-requestmitvma.mit.edu

More... (looser matching)
- Last day of these posts
- In 2002 , 2003 only
- Today
- New search...