# Relay confusion

Discussion in 'Electronic Basics' started by Roger Bourne, Jun 29, 2006.

1. ### Roger BourneGuest

Hello all,

Today I was introduced to the wide world of magnetic relays (for SPDT
switching applications)... .Push a current through the coil of the
relay--> creates magnetism --> makes the mechanical METAL switch flip
position, and TA-DAH you have an SPDT with practically nil "switch on"
resistance capable of passing high currents (in the As).
The problem that I have with relays is to do with their datasheet
contents. Now I am more or less familiar with inductors; their
inductance varies with frequency alongside their series resistance but
their inductance+serires resistance is more or less constant at low
frequencies.
How is it then that the [coil inductance + the coil resistance]
caracteristics vary with the rated voltage of the coil (I assume that
the rated voltage is the voltage drop accross the inductor used to flip
the switch) in order to maintain a steady power consumption?
What I am missing ?
Does the "L+ ESR" of the coil really increase/decrease with its voltage
drop ? Seems odd.

I know that I must be not understanding a fundamental principle, since
various datasheets all have the same varying coil inductance w.r.t.
their rated voltage.

This "pseudo-problem was discovered" when scouring randomly-chosen
relay datasheets in order to find how relays are typically driven. Are
there ICs out there whose task is to drive relays. I.e. they source a
steady current of lets say 50 mA to one of the relay coil's terminal
while the other coil terminal is tethered to a power supply node ? A
typical digital control signal will simply not do the trick. Is there a
typical approach for driving relays?

All manners of feedback will be appreciated.

Kind Regards
-Roger

2. ### David L. JonesGuest

You typically drive a relay from a digital signal with an NPN
transistor.
You'd typically only use an IC driver if you wanted to drive a bunch of
relays. The ULN 2003 is a common device for this.

You can get 5V "high sensitivity" relays which can be driven directly
from a digital logic output, but they are usually your smaller signal
type relays, not the big beefy ones.

Dave

3. ### Phil AllisonGuest

** OK, cop this.

Relays operate using #direct current, so the inductance of the coil is
irrelevant to the basic operation.

Relays are made in a variety of voltages, simply to suit the needs of
various users.

Relays with higher voltages have MORE turns of finer gauge wire on the coil
than those with low voltages.

MORE turns of wire means higher resistance and inductance values.

Relays vary in the amount of power needed to make the contacts engage.

Small generally relays use less power than big ones.

Get it ?

( # Excepting those specially designed to use 50/60 Hz AC voltage )

...... Phil

4. ### John LarkinGuest

Hi, Rog,

A given relay family (all alike except for DC coil voltage) will tend
to have identical coil power requirements. So a 24-volt relay of type
X will use half the current of a 12-volt type X, so its coil
resistance will be 4x as much.

The 24-volt guy will have twice the number of turns (of smaller wire)
so will have 4x the inductance. With 4x inductance and 4x resistance,
the L/R time constant stays the same, so pullin/dropout timing will be
mostly unchanged.

Of course, as the armature moves in, the magnetic path gets shorter,
so the inductance may increase considerably after it's energized!

Normally one would tie one end of the coil to, say, +24 volts, and
switch the other end to ground (or not) with a transistor or a fet or
something. So coil current is determined by the supply voltage and the
coil resistance, not by some value programmed into the driver. A diode
is usually placed across the relay coil to keep the inductive flyback
from trashing the transistor when it turns off.

Small relays can be driven directly from logic gates, often without
even needing a diode.

AC relays are a bit more interesting.

John

5. ### John GGuest

Sorry to say this and no offence intended.
Electronics 101 starts at your nearest technical institute next year.
After that you may be ready understand these things.
News groups are good for single questions but they are not all that
great for a basic understanding of the subject.

6. ### Phil AllisonGuest

"John G"

** That is a VERY cruel remark.

The OP announced his inexperience completely honestly.

** The OP has almost certainly done what jokingly passes for "'Electronics
101" these days.

Only it was all done on a PC running a dodgy simulation program.

Only the instructors barely knew more than the students.

Only no REAL items of electrical or electronic hardware were available for
student perusal - let alone the sort of free experimentation ESSENTIAL
for the development of a *feel* of their " raison d'etre " .

** Says who?

Ever read the name of this NG ????

....... Phil

7. ### Alan BGuest

I respectfully disagree. The poster had a question germane to the
newsgroup, presented it completely and matter-of-factly, and got a very
good answer from John Larkin[1]; an answer that was probably of good use to
a great many readers of the group, including myself. It was one of those
little things that I've not spent a lot of time worrying about, and now I
know more! I've been enlightened!

[1] No offense to Phil, but your answer was just a bit snippy.

8. ### Phil AllisonGuest

"Alan Bullshit "

( snip sickening arse licking)

** LOL .....

A typically narky throw away remark from one of Larkin's fellow autistic
cock suckers - who could not be "enlightened " even if a 4 foot long
fluoro tube were shoved up him.

........ Phil

9. ### David L. JonesGuest

I'll jump to the defense of the OP too, it was a perfectly valid
question asked nicely. The OP clearly has done some form of electronics
101 (not that it matters), but didn't know much about relay driving
which is fair enough.

You can get a great understanding of a subject on newsgroups! Some
people give detailed and through replies on a subject, often much
better than you'll get in any text book or electronics class. Basic
relay operation is not hard to explain.
Other times you can often peiece together comments from different
replys and understanding of a subject can fall into place quickly from
that.

Dave

10. ### John LarkinGuest

Hmmm... that makes the dropout time slower than pullin, given a
low-impedance drive, or a diode clamp. Never thought of it that way
before. It would be interesting to measure the deenergized/energized
coil inductances of a few DC relays.

The delta-L is an important design factor in AC relays, and allows
them to be more reliable than DC relays in some situations.

John

11. ### JamieGuest

not exactly sure what your looking for but how ever.
maybe its the pull and holding currents your looking at?
relays require more current to get the contacts toward the
magnetic core than it does to hold it there.
as the contact arm gets closer to the core the magnetic
field becomes stronger! thus once it's fully on it may only
require half the current to hold it on than what it did to
get it started.
in fuel injected engines you have what is called a
Peak and Hold injector which in it self, is a magnetic actuator..
more current is applied to get it to the full open position and
then less current is needed to hold it there ! in logic this
could be stated as hysteresis;

12. ### jasenGuest

relays are crude mechanical devices, get the voiltage approximately right and
they'll work after a few microseconds.

as you have noticed they are inductive and so won't respond instantly to a
voltage, and using a constant current drive can get a slightly faster
response.

similarly sized relays need a similar ammount of magnetism to operate the
contacts, to get that from a lower voltage needs less resistance:
thiker wire and fewer turns carrying more current.

total watts will be in the same ballpark.

Bye.
Jasen

13. ### Phil AllisonGuest

** Strewth - I gotta see one of them.

........ Phil

14. ### Roger BourneGuest

Hello all,

Thanks to all (well, almost all ) for the throrough explanations. I
also am glad to see that I am not the only one who thinks that I did
not overstep the bounds of this ng.

The reason why I need a relay, is because I want to multiplex the power
supply of a DDR RAM ---they can sink up to 2A of current. Otherwise, I
would went with a simple cmos switch.

-Roger

15. ### John LarkinGuest

There are bounds?

er, milliseconds.
Expect nasty bounce and stuff. Electronic switching might be better.

John

16. ### Alan BGuest

Ah. A true sociopath. How incredibly boring.

17. ### jasenGuest

Oops! I meant a few milliseconds,

Bye.
Jasen

18. ### jasenGuest

seems an odd use for a relay. they are kind of slow in comparsion to ram
chips. and I meant milliseconds no microseconds.

Bye.
Jasen

19. ### Roger BourneGuest

#

I want to multiplex the RAM's power, not one of it's data signals.
Since a DDR RAM can draw roughly up to 2A of juice, I do not know of
another alternative than a relay. A response time of in the order of ms
is the price I have to pay .

-Roger

20. ### David L. JonesGuest

Ok, I'm curious, *why* do you have to switch power to DDR RAM?
Is this some sort of custom design?

BTW, be careful doing this, you may not be able to simply throw a relay
in series with the RAM, signal integrity issues could really come into
play here.

Dave