Inductive clamp energy

[The Phantom]

There have been several threads about clamping relay coils, etc. There are
some subtleties connected with this whole topic that I haven't seen
mentioned. I've posted a paper from 1968 over in ABSE that explains it
all.

This is an interesting topic with practical applications, and I hope the
paper stimulates some discussion.

 

[Tony Williams]

There have been several threads about clamping relay coils, etc. There are
some subtleties connected with this whole topic that I haven't seen
mentioned. I've posted a paper from 1968 over in ABSE that explains it
all.

That was posted over 8 hours ago. Nothing in abse yet.

 

[Winfield]

On Thu, 21 Jun 2007 15:13:42 -0700, The Phantom wrote


That was posted over 8 hours ago. Nothing in abse yet.

I was able to grab it about 7pm EDT last night, 9 hrs ago.
But I haven't read or even more than glanced at it yet (it
looked complicated), so you're at no disadvantage, so far!

 

[The Phantom]

That was posted over 8 hours ago. Nothing in abse yet.

Well, it does take a little while to digest it all.

Briefly, what they're showing is something that, to borrow a phrase John
Popelish used nearly 3 years ago, "...we knew but didn't realize." Any of
us could figure it out, if only we realized we should.

As shown in this app note:

http://relays.tycoelectronics.com/appnotes/app_pdfs/13c3311.pdf

the suppressor (clamp) can be connected across the switch *or* across the
relay coil (inductor).

If the clamp is across the inductor, then a simple zener can't be used,
because the zener will conduct in the forward direction when the switch is
on; another diode must be placed in series with the zener. A bidirectional
suppressor (two zeners in one package, or a MOV) solves that problem, but
may be more expensive or hard to get.

If the clamp is across the switch, then a simple zener can be used, but as
shown in the paper I posted on ABSE, the dissipation in the clamp is not
just due to the energy stored in the inductance.

The thing to realize is that when the clamp is across the switch and the
switch turns off, the current in the inductor is decaying and the power
source, inductor and clamp are all carrying the same current. The
direction of current in the DC power source is such as to represent energy
flow *out* of the source. Since we believe in conservation of energy, we
see that the energy that must be absorbed by the dissipative elements
(mainly the resistance of the wire of the inductor and the clamp) will be
*not only* the energy stored in the inductor at turn-off. The energy
supplied by the power source during time while the inductor current is
decaying must be absorbed, as well as the inductor's stored energy.

The authors of the paper I posted over on ABSE show that if the clamp
voltage (when the clamp is across the switch) is only slightly higher than
the power source voltage, the time for the inductor current to decay is
increased, and the energy the clamp must absorb can be substantially more
than L*I^2/2.

The situation is different if the clamp is across the inductor, and this
factor should be taken into account when deciding where to place the clamp.

The increase in energy to be absorbed when the clamp, or snubber, is across
the switch is a phenomenon that also should be taken into account in
switching power supplies.

 

[John Popelish]

The Phantom wrote:
(snip)

As shown in this app note:

http://relays.tycoelectronics.com/appnotes/app_pdfs/13c3311.pdf

the suppressor (clamp) can be connected across the switch *or* across the
relay coil (inductor).

If the clamp is across the inductor, then a simple zener can't be used,
because the zener will conduct in the forward direction when the switch is
on; another diode must be placed in series with the zener. A bidirectional
suppressor (two zeners in one package, or a MOV) solves that problem, but
may be more expensive or hard to get.

If the clamp is across the switch, then a simple zener can be used, but as
shown in the paper I posted on ABSE, the dissipation in the clamp is not
just due to the energy stored in the inductance.

The thing to realize is that when the clamp is across the switch and the
switch turns off, the current in the inductor is decaying and the power
source, inductor and clamp are all carrying the same current. The
direction of current in the DC power source is such as to represent energy
flow *out* of the source. Since we believe in conservation of energy, we
see that the energy that must be absorbed by the dissipative elements
(mainly the resistance of the wire of the inductor and the clamp) will be
*not only* the energy stored in the inductor at turn-off. The energy
supplied by the power source during time while the inductor current is
decaying must be absorbed, as well as the inductor's stored energy.

The authors of the paper I posted over on ABSE show that if the clamp
voltage (when the clamp is across the switch) is only slightly higher than
the power source voltage, the time for the inductor current to decay is
increased, and the energy the clamp must absorb can be substantially more
than L*I^2/2.

The situation is different if the clamp is across the inductor, and this
factor should be taken into account when deciding where to place the clamp.

The increase in energy to be absorbed when the clamp, or snubber, is across
the switch is a phenomenon that also should be taken into account in
switching power supplies.

The possible up side of having the snubber across the switch
(that might pay for the higher power dissipation during
inductive discharge, in some cases) is that the supply
current is continuous during the process, so the supply
voltage is not bumped up as much as when the snubber is
across the coil and the supply current switches of as fast
as the switch current drops.

Generally, I would rather use bypass capacitance to smooth
the bump than dump extra energy into the clamp.

 

[John Popelish]

The Phantom wrote:
(snip)

Briefly, what they're showing is something that, to borrow a phrase John
Popelish used nearly 3 years ago, "...we knew but didn't realize." Any of
us could figure it out, if only we realized we should.

(snip)

When I read this, I assumed you were remembering something
someone else had said, but had my name attached to the
memory. I was wrong. I found it with Google and enjoyed
reading the thread all over again. It is a shame my brain
is full and I have to lose a memory to make room for a new one.

 

[Jon]

There have been several threads about clamping relay coils, etc. There are
some subtleties connected with this whole topic that I haven't seen
mentioned. I've posted a paper from 1968 over in ABSE that explains it
all.

This is an interesting topic with practical applications, and I hope the
paper stimulates some discussion.

I don't get out much...What is ABSE?

 

[Jonathan Kirwan]

I don't get out much...What is ABSE?

If your newsgroup browser supports it, you can use this link to get
directly to the indicated article:

Otherwise, look for alt.binaries.schematics.electronic and search for
the subject looking like: Energy in clamps--from SED - Inductive
Energy Calculations...

Jon

 

[Rich Grise]

If your newsgroup browser supports it, you can use this link to get
directly to the indicated article:

Otherwise, look for alt.binaries.schematics.electronic and search for
the subject looking like: Energy in clamps--from SED - Inductive
Energy Calculations...

He's a googlie - they don't get binaries.

What Jon needs to do is contact his ISP, and get the IP/URL of his
newsserver, and find a real newsreader (butthook distress sucks), and then
he can post/read anywhere he wants to.

Cheers!
Rich