DC relay latching voltage

[Pimpom]

A quick question:
Is it reasonable to assume that most small general-purpose DC relays
will not engage at 50% or the rated coil voltage? If not, how about 40%?
(This is *not* about the release voltage).

[Arie de Muijnck]

Is it reasonable to assume that most small general-purpose DC relays will not engage at 50% or the rated coil voltage? If not, how about 40%?
(This is *not* about the release voltage).

Not safe. If an external magnetic field is present (e.g. from nearby relays) or a mechanical shock occurs it can activate and hold.
Been there, had that.

Arie

[Pimpom]

Post by Arie de Muijnck

Post by Pimpom
Is it reasonable to assume that most small general-purpose DC relays
will not engage at 50% or the rated coil voltage? If not, how about 40%?
(This is *not* about the release voltage).

Not safe. If an external magnetic field is present (e.g. from nearby
relays) or a mechanical shock occurs it can activate and hold.
Been there, had that.
Arie

Thanks. The immediate application I have in mind is highly unlikely to
be subjected to such external factors. and it won't have disastrous
consequences if it did happen. But it's good to know for future
consideration.

The application *is* likely to cause 20-25% of the nominal coil voltage
to appear across the terminals for some 50 milliseconds though.

[piglet]

Post by Pimpom
Is it reasonable to assume that most small general-purpose DC relays
will not engage at 50% or the rated coil voltage? If not, how about 40%?
(This is *not* about the release voltage).

That roughly fits my observations but I don’t think it can be trusted ,
aging, vibration or temperatures outside room temp could make it pull in
earlier. You might need to characterise the relays you have in mind.

[Pimpom]

Post by piglet

Post by Pimpom
Is it reasonable to assume that most small general-purpose DC relays
will not engage at 50% or the rated coil voltage? If not, how about 40%?
(This is *not* about the release voltage).

That roughly fits my observations but I don’t think it can be trusted ,
aging, vibration or temperatures outside room temp could make it pull in
earlier. You might need to characterise the relays you have in mind.

Thanks for the reply. Actually, my calculations indicate that less than
25% of the nominal voltage will appear across the coil for some tens of
milliseconds when it should be disengaged. This is caused by a decaying
supply voltage, not an inductive spike.

[John Robertson]

Post by Pimpom

Post by piglet

Post by Pimpom
Is it reasonable to assume that most small general-purpose DC relays
will not engage at 50% or the rated coil voltage? If not, how about 40%?
(This is *not* about the release voltage).

That roughly fits my observations but I don’t think it can be trusted ,
aging, vibration or temperatures outside room temp could make it pull in
earlier. You might need to characterise the relays you have in mind.

Thanks for the reply. Actually, my calculations indicate that less than
25% of the nominal voltage will appear across the coil for  some tens of
milliseconds when it should be disengaged. This is caused by a decaying
supply voltage, not an inductive spike.

I'd suggest using a power supply at the maximum decay voltage you expect
then bang the relay a bit to see if it will engage. I assume you have
the normal back diode across the coil for clamping the ringing voltage.
You could use the NO contacts in the relay to keep it energized (Hold)
if it momentarily closes during testing.

John :-#)#

--
(Please post followups or tech inquiries to the USENET newsgroup)
John's Jukes Ltd.
#7 - 3979 Marine Way, Burnaby, BC, Canada V5J 5E3
[phone] (Pinballs, Jukes, Video Games)
www.flippers.com
"Old pinballers n

[Pimpom]

Post by John Robertson

Post by Pimpom

Post by piglet

Post by Pimpom
Is it reasonable to assume that most small general-purpose DC relays
will not engage at 50% or the rated coil voltage? If not, how about 40%?
(This is *not* about the release voltage).

That roughly fits my observations but I don’t think it can be trusted ,
aging, vibration or temperatures outside room temp could make it pull in
earlier. You might need to characterise the relays you have in mind.

Thanks for the reply. Actually, my calculations indicate that less
than 25% of the nominal voltage will appear across the coil for  some
tens of milliseconds when it should be disengaged. This is caused by a
decaying supply voltage, not an inductive spike.

I'd suggest using a power supply at the maximum decay voltage you expect
then bang the relay a bit to see if it will engage. I assume you have
the normal back diode across the coil for clamping the ringing voltage.
You could use the NO contacts in the relay to keep it energized (Hold)
if it momentarily closes during testing.
John :-#)#

That would give a good indication but wouldn't guarantee the same result
with another sample.

I can't give too much detail but this is broadly what the circuit has to do:
The relay and associated circuits are to be a part of a multi-section
project. The relay switches another section on at power-up and then
quickly disengage that section soon after power is switched off.

Normal power supply decay will not release the relay quickly enough. So
I've added a section that turns the 12V relay off when the decaying PS
drops below 7.5V. Works fine in simulation.

However, due to interaction with other sections, the gate voltage of the
transistor driving the relay rises again briefly to about 2.5V before
the shutdown process is complete. But by this time, the 12V supply has
dropped to less than 3V.

I don't think any normal 12V relay will re-engage at 3V, but I wanted to
be sure. When I cited 50% and 40% of the rated coil voltage, I was being
conservative and night have misled readers. Sorry about that.

[Liz Tuddenham]

Pimpom <***@invalid.invalid> wrote:
[...]

Post by Pimpom
Normal power supply decay will not release the relay quickly enough. So
I've added a section that turns the 12V relay off when the decaying PS
drops below 7.5V. Works fine in simulation.
However, due to interaction with other sections, the gate voltage of the
transistor driving the relay rises again briefly to about 2.5V before
the shutdown process is complete. But by this time, the 12V supply has
dropped to less than 3V.

Add a few diode drops in the supply to the Base of the transistor and a
pull-down resistor from Base to Earth. You could even use a 7.5v Zener
so the transistor switched directly off the PS and no other active
components were needed.

[Pimpom]

Post by Liz Tuddenham
[...]

Post by Pimpom
Normal power supply decay will not release the relay quickly enough. So
I've added a section that turns the 12V relay off when the decaying PS
drops below 7.5V. Works fine in simulation.
However, due to interaction with other sections, the gate voltage of the
transistor driving the relay rises again briefly to about 2.5V before
the shutdown process is complete. But by this time, the 12V supply has
dropped to less than 3V.

Add a few diode drops in the supply to the Base of the transistor and a
pull-down resistor from Base to Earth. You could even use a 7.5v Zener
so the transistor switched directly off the PS and no other active
components were needed.

Now that sounds like a good idea. Thanks. I'll check it out to see if it
suits my purpose. For one thing, switching the transistor will not be a
snap action, but that shouldn't matter as long as the relay contacts
snap open. The relay is likely to be a 12V 400Ω model so that the slow
turn off wouldn't unduly stress the transistor.

No offense, but this is one of those "Why didn't I think of that"
moments.But then no one else seems to have done so either. :-)

[Liz Tuddenham]

...The relay is likely to be a 12V 400Ω model so that the slow
turn off wouldn't unduly stress the transistor.

At half-voltage and half-current, the most stressful condition, the
dissipation of the transistor would be 90 mW.

[john larkin]

Post by Liz Tuddenham
[...]

Post by Pimpom
Normal power supply decay will not release the relay quickly enough. So
I've added a section that turns the 12V relay off when the decaying PS
drops below 7.5V. Works fine in simulation.
However, due to interaction with other sections, the gate voltage of the
transistor driving the relay rises again briefly to about 2.5V before
the shutdown process is complete. But by this time, the 12V supply has
dropped to less than 3V.

Add a few diode drops in the supply to the Base of the transistor and a
pull-down resistor from Base to Earth. You could even use a 7.5v Zener
so the transistor switched directly off the PS and no other active
components were needed.

A 2-resistor voltage divider can reduce mosfet gate swing.

Or an R-C in his case of a transient gate drive spike.

Or both.

[piglet]

Post by john larkin

Post by Liz Tuddenham
[...]

Post by Pimpom
Normal power supply decay will not release the relay quickly enough. So
I've added a section that turns the 12V relay off when the decaying PS
drops below 7.5V. Works fine in simulation.
However, due to interaction with other sections, the gate voltage of the
transistor driving the relay rises again briefly to about 2.5V before
the shutdown process is complete. But by this time, the 12V supply has
dropped to less than 3V.

Add a few diode drops in the supply to the Base of the transistor and a
pull-down resistor from Base to Earth. You could even use a 7.5v Zener
so the transistor switched directly off the PS and no other active
components were needed.

A 2-resistor voltage divider can reduce mosfet gate swing.
Or an R-C in his case of a transient gate drive spike.
Or both.

If the C is gate to drain (or base to collector) then the miller effect
slows rate of flux collapse enough to remove need for any additional
components across the coil.

[Don Y]

I assume you have the normal
back diode across the coil for clamping the ringing voltage.

If you care about the life of the relay (and how well it "opens"),
a better approach is to add a zener rated at the coil voltage in
series with that diode. The diode, by itself, delays opening of
the contacts and leads to more opportunities for arcing. The
zener speeds up this transition.

[John Robertson]

Post by Don Y

I assume you have the normal back diode across the coil for clamping
the ringing voltage.

If you care about the life of the relay (and how well it "opens"),
a better approach is to add a zener rated at the coil voltage in
series with that diode.  The diode, by itself, delays opening of
the contacts and leads to more opportunities for arcing.  The
zener speeds up this transition.

Hi Don,

Don't you mean that one could use a zener diode in place of the back-EMF
diode - or parallel with it?

We've been using 1N400X (and 3A in some cases)diodes since the 70s for
protecting pinball driver transistors - not too worried about the decay
time so haven't really looked deeper into it before.

Thanks,

John :-#)#

--
(Please post followups or tech inquiries to the USENET newsgroup)
John's Jukes Ltd.
#7 - 3979 Marine Way, Burnaby, BC, Canada V5J 5E3
[phone] (Pinballs, Jukes, Video Games)
www.flippers.com
"Old pinbal

[Don Y]

Post by Don Y

I assume you have the normal back diode across the coil for clamping the
ringing voltage.

If you care about the life of the relay (and how well it "opens"),
a better approach is to add a zener rated at the coil voltage in
series with that diode.  The diode, by itself, delays opening of
the contacts and leads to more opportunities for arcing.  The
zener speeds up this transition.

Don't you mean that one could use a zener diode in place of the back-EMF diode
- or parallel with it?

In series. It speeds up the decay of the magnetic field (and, thus, the
opening of the armature) by allowing a higher potential to exist across
the coil while it is "opening". The ideal condition is with NO catch diode
(but that tends to fry solid state switches! :> ) You're making a
snubber, of sorts.

We've been using 1N400X (and 3A in some cases)diodes since the 70s for
protecting pinball driver transistors - not too worried about the decay time so
haven't really looked deeper into it before.

It really only matters if you are really concerned over the actual opening
time (delay) of the relay and/or how long you want the relay (contacts) to
last. (many of my designs have to have service lives of decades or more.)

<https://www.te.com/commerce/DocumentDelivery/DDEController?Action=srchrtrv&DocNm=13C3264_AppNote&DocType=CS&DocLang=EN>
will give you a quick overview.
I have better notes from (relay) application engineers but I'm busy baking,
tonight, so can only take quick pokes at my mail, etc.

[John Robertson]

Post by John Robertson

Post by Don Y

I assume you have the normal back diode across the coil for clamping
the ringing voltage.

If you care about the life of the relay (and how well it "opens"),
a better approach is to add a zener rated at the coil voltage in
series with that diode.  The diode, by itself, delays opening of
the contacts and leads to more opportunities for arcing.  The
zener speeds up this transition.

Don't you mean that one could use a zener diode in place of the back-
EMF diode - or parallel with it?

In series.  It speeds up the decay of the magnetic field (and, thus, the
opening of the armature) by allowing a higher potential to exist across
the coil while it is "opening".  The ideal condition is with NO catch diode
(but that tends to fry solid state switches!  :> )  You're making a
snubber, of sorts.

Post by John Robertson
We've been using 1N400X (and 3A in some cases)diodes since the 70s for
protecting pinball driver transistors - not too worried about the
decay time so haven't really looked deeper into it before.

It really only matters if you are really concerned over the actual opening
time (delay) of the relay and/or how long you want the relay (contacts) to
last.  (many of my designs have to have service lives of decades or more.)
<https://www.te.com/commerce/DocumentDelivery/DDEController?
Action=srchrtrv&DocNm=13C3264_AppNote&DocType=CS&DocLang=EN> will give
you a quick overview.
I have better notes from (relay) application engineers but I'm busy baking,
tonight, so can only take quick pokes at my mail, etc.

Good reading! It appears counterintuitive, but if I warp (!) my head
around it I should be able to sort it out.

Thanks, that is very helpful.

John :-#)#

--
(Please post followups or tech inquiries to the USENET newsgroup)
John's Jukes Ltd.
#7 - 3979 Marine Way, Burnaby, BC, Canada V5J 5E3
[phone] (Pinballs, Jukes, Video Games)
www.flippers.com
"Old pinballer

[Don Y]

Good reading! It appears counterintuitive, but if I warp (!) my head around it
I should be able to sort it out.

Think about the extremes:

- with NO diode or snuber, there is nothing to keep the magnetic
field intact when the switch opens. So, the spring force is the
sole actor in play, pulling the contacts open
- with a diode, some current continues to "recirculate" in the coil
thus keeping the magnetic field active -- so it is trying to hold
the relay closed while the spring is trying to open it. *When*
the recirculating current falls, the field collapses and the relay
can open.

The zener lets the circuit look more like the "no diode" case.

[Don Y]

Good reading! It appears counterintuitive, but if I warp (!) my head around
it I should be able to sort it out.

- with NO diode  or snuber, there is nothing to keep the magnetic
  field intact when the switch opens.  So, the spring force is the
  sole actor in play, pulling the contacts open
- with a diode, some current continues to "recirculate" in the coil
  thus keeping the magnetic field active -- so it is trying to hold
  the relay closed while the spring is trying to open it.  *When*
  the recirculating current falls, the field collapses and the relay
  can open.
The zener lets the circuit look more like the "no diode" case.

i.e., imagine that case as having an infinite voltage zener (open
circuit) in series with the diode -- NO diode!

[john larkin]

On Sat, 24 Aug 2024 23:22:01 -0700, Don Y

Post by Don Y

Good reading! It appears counterintuitive, but if I warp (!) my head around
it I should be able to sort it out.

- with NO diode  or snuber, there is nothing to keep the magnetic
  field intact when the switch opens.  So, the spring force is the
  sole actor in play, pulling the contacts open
- with a diode, some current continues to "recirculate" in the coil
  thus keeping the magnetic field active -- so it is trying to hold
  the relay closed while the spring is trying to open it.  *When*
  the recirculating current falls, the field collapses and the relay
  can open.
The zener lets the circuit look more like the "no diode" case.

i.e., imagine that case as having an infinite voltage zener (open
circuit) in series with the diode -- NO diode!

The losses in a real relay will limit the peak voltage, which will
actually be a damped ringing thing.

L changes substantially as the parts move around, decreasing as the
magnetic loop goes from closed/seated to open/air gap. Envision some
weird waveforms.

That's the reason that AC-coil relays are prefered over DC relays in
industrial control systems.

[piglet]

Post by John Robertson

Post by John Robertson

Post by Don Y

I assume you have the normal back diode across the coil for clamping
the ringing voltage.

If you care about the life of the relay (and how well it "opens"),
a better approach is to add a zener rated at the coil voltage in
series with that diode.  The diode, by itself, delays opening of
the contacts and leads to more opportunities for arcing.  The
zener speeds up this transition.

Don't you mean that one could use a zener diode in place of the back-
EMF diode - or parallel with it?

In series.  It speeds up the decay of the magnetic field (and, thus, the
opening of the armature) by allowing a higher potential to exist across
the coil while it is "opening".  The ideal condition is with NO catch diode
(but that tends to fry solid state switches!  :> )  You're making a
snubber, of sorts.

Post by John Robertson
We've been using 1N400X (and 3A in some cases)diodes since the 70s for
protecting pinball driver transistors - not too worried about the
decay time so haven't really looked deeper into it before.

It really only matters if you are really concerned over the actual opening
time (delay) of the relay and/or how long you want the relay (contacts) to
last.  (many of my designs have to have service lives of decades or more.)
<https://www.te.com/commerce/DocumentDelivery/DDEController?
Action=srchrtrv&DocNm=13C3264_AppNote&DocType=CS&DocLang=EN> will give
you a quick overview.
I have better notes from (relay) application engineers but I'm busy baking,
tonight, so can only take quick pokes at my mail, etc.

Good reading! It appears counterintuitive, but if I warp (!) my head
around it I should be able to sort it out.
Thanks, that is very helpful.
John :-#)#

An even simpler coil catcher is a resistor. Many times the extra power
consumption is not an issue and resistors are even cheaper and more
reliable than zener + diode.

[Don Y]

Post by piglet
An even simpler coil catcher is a resistor. Many times the extra power
consumption is not an issue and resistors are even cheaper and more
reliable than zener + diode.

A lot depends on the relay's (coil) duty cycle.

To limit the peak transient (to approximately the supply voltage) when
the coil opens, R needs to be on the order of the coil resistance.
If the this halves the load resistance that must be driven when the
coil is active.

Of course, if the duty cycle is low and 2X the driving current is
within the capabilities of the device you had already planned on
using for the switch, then this has minimal impact.

OTOH, if the coil current is higher, already taxing the capabilities
of the switch or the duty cycle has the coil energized more often
than not, then this can be a significant factor.

And, as R approaches the coil's resistance, the opening time/delay of
the relay *increases*. I.e., you want R to approach infinity to get the
best opening transition but this also gives the highest switching
transient voltage -- often many times the supply voltage!

[Don Y]

Post by Don Y

Post by piglet
An even simpler coil catcher is a resistor. Many times the extra power
consumption is not an issue and resistors are even cheaper and more
reliable than zener + diode.

A lot depends on the relay's (coil) duty cycle.
To limit the peak transient (to approximately the supply voltage) when
the coil opens, R needs to be on the order of the coil resistance.
If the this halves the load resistance that must be driven when the
coil is active.
Of course, if the duty cycle is low and 2X the driving current is
within the capabilities of the device you had already planned on
using for the switch, then this has minimal impact.
OTOH, if the coil current is higher, already taxing the capabilities
of the switch or the duty cycle has the coil energized more often
than not, then this can be a significant factor.
And, as R approaches the coil's resistance, the opening time/delay of
the relay *increases*.  I.e., you want R to approach infinity to get the
best opening transition but this also gives the highest switching
transient voltage -- often many times the supply voltage!

<https://www.te.com/commerce/DocumentDelivery/DDEController?Action=srchrtrv&DocNm=13C3311_AppNote&DocType=CS&DocLang=EN>

[John Robertson]

Post by Don Y

Post by Don Y

Post by piglet
An even simpler coil catcher is a resistor. Many times the extra power
consumption is not an issue and resistors are even cheaper and more
reliable than zener + diode.

A lot depends on the relay's (coil) duty cycle.
To limit the peak transient (to approximately the supply voltage) when
the coil opens, R needs to be on the order of the coil resistance.
If the this halves the load resistance that must be driven when the
coil is active.
Of course, if the duty cycle is low and 2X the driving current is
within the capabilities of the device you had already planned on
using for the switch, then this has minimal impact.
OTOH, if the coil current is higher, already taxing the capabilities
of the switch or the duty cycle has the coil energized more often
than not, then this can be a significant factor.
And, as R approaches the coil's resistance, the opening time/delay of
the relay *increases*.  I.e., you want R to approach infinity to get the
best opening transition but this also gives the highest switching
transient voltage -- often many times the supply voltage!

<https://www.te.com/commerce/DocumentDelivery/DDEController?
Action=srchrtrv&DocNm=13C3311_AppNote&DocType=CS&DocLang=EN>

Is there an index to these useful Application Notes?
I only found this when I searched for "application notes":

https://www.te.com/en/products/relays-and-contactors/relays/intersection/application-notes.html

And did you happen to write some of them by chance?

John :-#)#

--
(Please post followups or tech inquiries to the USENET newsgroup)
John's Jukes Ltd.
#7 - 3979 Marine Way, Burnaby, BC, Canada V5J 5E3
[phone] (Pinballs, Jukes, Video Games)
www.flippers.com
"Old pinb

[Don Y]

Post by John Robertson

Post by Don Y
<https://www.te.com/commerce/DocumentDelivery/DDEController?
Action=srchrtrv&DocNm=13C3311_AppNote&DocType=CS&DocLang=EN>

Is there an index to these useful Application Notes?

No idea. I'm just leaning on lessons learned decades ago and searching for
documents that would likely confirm those issues.

Post by John Robertson
https://www.te.com/en/products/relays-and-contactors/relays/intersection/application-notes.html
And did you happen to write some of them by chance?

No. I don't publish anything "for general consumption"; only specific
issues for specific audiences. Not enough time in the day to do the
things I *want* to do while considering "optional" efforts.

Don't kill yourself worrying about this sort of thing for a pin table.
You're likely not as concerned with durability. The board's already
(likely) designed. And, it's already got an established (tolerable?)
level of EMI.

Rather, it should be seen as a counter to the "old saw" that you *just* use
a recirculating diode without considering the consequences. Just like
considering how to *drive* the coil based on how it will be used.

[Remember the "pull in" coils and "end of stroke" switches on flipper
solenoids?]

[John Robertson]

Post by John Robertson

Post by Don Y
<https://www.te.com/commerce/DocumentDelivery/DDEController?
Action=srchrtrv&DocNm=13C3311_AppNote&DocType=CS&DocLang=EN>

Is there an index to these useful Application Notes?

No idea.  I'm just leaning on lessons learned decades ago and searching for
documents that would likely confirm those issues.

Post by John Robertson
https://www.te.com/en/products/relays-and-contactors/relays/
intersection/application-notes.html
And did you happen to write some of them by chance?

No.  I don't publish anything "for general consumption"; only specific
issues for specific audiences.  Not enough time in the day to do the
things I *want* to do while considering "optional" efforts.
Don't kill yourself worrying about this sort of thing for a pin table.
You're likely not as concerned with durability.  The board's already
(likely) designed.  And, it's already got an established (tolerable?)
level of EMI.

Actually the 2nd link you provided that speaks of Coil Suppression with
DC relays pointed out that the zener and regular diode in series had a
drop -out time of almost the same as an unprotected coil, but the EMI
was limited to the zener's rating along with the diode voltage drop.
What is interesting is if you have a regular diode across the coil it
takes around 5 times as long to decay. That may actually matter in
pinball games - one could get snappier coil action with the zener/diode
combination (or zener across the driver transistor - skipping the
regular diode as mentioned in the note) and folks may be able to notice
that. Hmm, 1.9ms vs 9.8ms - can humans detect that when playing
considering that for the most part 100ms is considered 'instantaneous'?
I may have to set up a game and see...

"Many engineers use a rectifier diode alone to provide the transient
suppression for relay coils. While this is cost effective and fully
eliminates the transient voltage, its impact on relay performance can be
devastating. Problems of unexplained, random "tack welding" frequently
occur in these systems."

Rather, it should be seen as a counter to the "old saw" that you *just* use
a recirculating diode without considering the consequences.  Just like
considering how to *drive* the coil based on how it will be used.
[Remember the "pull in" coils and "end of stroke" (EOS) switches on flipper
solenoids?]

Oh, yes, we constantly deal with them, and pitted contacts on the EOS
switches in our shop. Would zener diodes across the contacts help reduce
pitting?

John :-#)#

--
(Please post followups or tech inquiries to the USENET newsgroup)
John's Jukes Ltd.
#7 - 3979 Marine Way, Burnaby, BC, Canada V5J 5E3
[phone] (Pinballs, Jukes, Video Games)
www.flippers.com
"Old pinballers never die, they just fli

[Don Y]

Post by Don Y
Don't kill yourself worrying about this sort of thing for a pin table.
You're likely not as concerned with durability.  The board's already
(likely) designed.  And, it's already got an established (tolerable?)
level of EMI.

Actually the 2nd link you provided that speaks of Coil Suppression with DC
relays pointed out that the zener and regular diode in series had a drop -out
time of almost the same as an unprotected coil,

Yes. What that implies is the contacts don't "linger" in close
proximity/partial contact. The opening action is more "snappy",
as intended by the choice of armature spring.

but the EMI was limited to the
zener's rating along with the diode voltage drop.

Yes. You also have to remember that you are going to see a larger voltage
spike than the simple diode (that would have clamped it to the supply rail).
If the semiconductor switch can't handle this...

What is interesting is if you
have a regular diode across the coil it takes around 5 times as long to decay.

That's the point; the relay is operating in a mode that wasn't intended
(i.e., shit or get off the pot!)

That may actually matter in pinball games - one could get snappier coil action
with the zener/diode combination (or zener across the driver transistor -
skipping the regular diode as mentioned in the note) and folks may be able to

It depends on what you are talking about. In an old electromechanical game
where the relays implemented the logic, the effects would largely be unnoticed.
Driving a pop bumper or flipper solenoid could see some improvement in the
"liveliness" of the action.

But, in the case of kicking targets, remember that the setting of the
sense contacts probably has a bigger impact on the feel of the game;
if the target reacts too soon or too late, the kinematics change as
the ball is in a different place with respect to the design of the
kicker.

notice that. Hmm, 1.9ms vs 9.8ms - can humans detect that when playing
considering that for the most part 100ms is considered 'instantaneous'? I may
have to set up a game and see...

An easier way to do it would be to have a little piece of code that allows
you to vary the delay between contact sense and solenoid actuation. Remember,
most kicking targets do their work on the "activation stroke"; thereafter, they
are just "recovering".

"Many engineers use a rectifier diode alone to provide the transient
suppression for relay coils. While this is cost effective and fully eliminates
the transient voltage, its impact on relay performance can be devastating.
Problems of unexplained, random "tack welding" frequently occur in these systems."

On an old machine, such a failure would be pretty obvious. I don't know if
newer (electronic) machines take any action to determine if their hammer
drivers are failed or coils faulty.

[I spent a shitload of time designing the interface to the irrigation
valves (solenoids) in my irrigation controller to avoid these sorts
of failures -- as well as detecting "forced" failures (e.g., user wired
the valves -- ir god knows what else! -- to the controller incorrectly).
But, the cost of a repair, there, far exceeds that of hauling a pin
table into the back room!]

Post by Don Y
Rather, it should be seen as a counter to the "old saw" that you *just* use
a recirculating diode without considering the consequences.  Just like
considering how to *drive* the coil based on how it will be used.
[Remember the "pull in" coils and "end of stroke" (EOS) switches on flipper
solenoids?]

Oh, yes, we constantly deal with them, and pitted contacts on the EOS switches
in our shop.

But, you have the advantage of being able to pull the machine off the floor,
and burnish and regap the contacts. And, you get some idea of how likely
the need base on how many plays it sees.

If the device in question is on a mountain in tibet... (yes! <frown>)

Would zener diodes across the contacts help reduce pitting?

An RC snubber might work better.

But, *maintaining* a pin table is half the fun! (unless, of course, you
are in a business to make money from them!)

[John Robertson]

Post by John Robertson

Post by Don Y
Don't kill yourself worrying about this sort of thing for a pin table.
You're likely not as concerned with durability.  The board's already
(likely) designed.  And, it's already got an established (tolerable?)
level of EMI.

Actually the 2nd link you provided that speaks of Coil Suppression
with DC relays pointed out that the zener and regular diode in series
had a drop -out time of almost the same as an unprotected coil,

Yes.  What that implies is the contacts don't "linger" in close
proximity/partial contact.  The opening action is more "snappy",
as intended by the choice of armature spring.

Post by John Robertson
but the EMI was limited to the zener's rating along with the diode
voltage drop.

Yes.  You also have to remember that you are going to see a larger voltage
spike than the simple diode (that would have clamped it to the supply rail).
If the semiconductor switch can't handle this...

Post by John Robertson
What is interesting is if you have a regular diode across the coil it
takes around 5 times as long to decay.

That's the point; the relay is operating in a mode that wasn't intended
(i.e., shit or get off the pot!)

Post by John Robertson
That may actually matter in pinball games - one could get snappier
coil action with the zener/diode combination (or zener across the
driver transistor - skipping the regular diode as mentioned in the
note) and folks may be able to

It depends on what you are talking about.  In an old electromechanical game
where the relays implemented the logic, the effects would largely be unnoticed.
Driving a pop bumper or flipper solenoid could see some improvement in the
"liveliness" of the action.
But, in the case of kicking targets, remember that the setting of the
sense contacts probably has a bigger impact on the feel of the game;
if the target reacts too soon or too late, the kinematics change as
the ball is in a different place with respect to the design of the
kicker.

Post by John Robertson
notice that. Hmm, 1.9ms vs 9.8ms - can humans detect that when playing
considering that for the most part 100ms is considered
'instantaneous'? I may have to set up a game and see...

An easier way to do it would be to have a little piece of code that allows
you to vary the delay between contact sense and solenoid actuation.
Remember,
most kicking targets do their work on the "activation stroke";
thereafter, they
are just "recovering".

Post by John Robertson
"Many engineers use a rectifier diode alone to provide the transient
suppression for relay coils. While this is cost effective and fully
eliminates the transient voltage, its impact on relay performance can
be devastating. Problems of unexplained, random "tack welding"
frequently occur in these systems."

On an old machine, such a failure would be pretty obvious.  I don't know if
newer (electronic) machines take any action to determine if their hammer
drivers are failed or coils faulty.
[I spent a shitload of time designing the interface to the irrigation
valves (solenoids) in my irrigation controller to avoid these sorts
of failures -- as well as detecting "forced" failures (e.g., user wired
the valves -- ir god knows what else! -- to the controller incorrectly).
But, the cost of a repair, there, far exceeds that of hauling a pin
table into the back room!]

It is still cheaper than sending a repair person out.

Post by John Robertson

Post by Don Y
Rather, it should be seen as a counter to the "old saw" that you *just* use
a recirculating diode without considering the consequences.  Just like
considering how to *drive* the coil based on how it will be used.
[Remember the "pull in" coils and "end of stroke" (EOS) switches on flipper
solenoids?]

Oh, yes, we constantly deal with them, and pitted contacts on the EOS
switches in our shop.

But, you have the advantage of being able to pull the machine off the floor,
and burnish and regap the contacts.  And, you get some idea of how likely
the need base on how many plays it sees.

Pinball games are more service trouble than most operators want to deal
with. Anything we can do to make them more reliable is part of our
business model.

And I hate repeat service calls.

If the device in question is on a mountain in tibet...  (yes!  <frown>)

Post by John Robertson
Would zener diodes across the contacts help reduce pitting?

An RC snubber might work better.
But, *maintaining* a pin table is half the fun!  (unless, of course, you
are in a business to make money from them!)

Ah, but you see that is the point of doing repairs that are better than
(improve upon) the original design. Back in the 80s I figured out why
one manufacturer's games were blowing up their driver transistors
randomly and published the answer in the trade journals of the day. It
was a ground issue, where the commons were done through Molex pins and
as the pins aged the ground connections generated resistance, which led
to transistors not fully biasing off and burning out. The fix was to
beef up the ground connections and all subsequent games we serviced
never blew the driver transistors any more and customers were happy!

Need I mention the factory never acknowledged the error and did NOT
implement my recommendations.
By the 80s they gained a reputation for unreliability that plagued them
until they finally closed shop in the mid-90s. Oddly enough the same
company was considered the Rolls-Royce of pinball up to the Solid State
machines. Their solid state games looked and played great until the
ground issues started randomly occurring (fried coils and driver
transistors) and then operators would get rid of that manufacturer's
games and buy other brand's machines.

There are two cases I deal with - commercial and home clients. Both
would rather see less of my shop then they have to and I am happy to
oblige them because I hate fixing simple things that could be made more
durable. Also we've gained a bit of a reputation of machines lasting
longer after being serviced by us and I wish to improve that score.

So, reducing contact arcing is very important to me and my customers!

We only get more business the better we fix the games!!

John :-#)#



--
(Please post followups or tech inquiries to the USENET newsgroup)
John's Jukes Ltd.
#7 - 3979 Marine Way, Burnaby, BC, Canada V5J 5E3
[phone] (Pinballs, Jukes, Video Games)
www.flippers.com
"Old pinballers neve

[Don Y]

Post by Don Y

Post by John Robertson

Post by Don Y
Rather, it should be seen as a counter to the "old saw" that you *just* use
a recirculating diode without considering the consequences.  Just like
considering how to *drive* the coil based on how it will be used.
[Remember the "pull in" coils and "end of stroke" (EOS) switches on flipper
solenoids?]

Oh, yes, we constantly deal with them, and pitted contacts on the EOS
switches in our shop.

But, you have the advantage of being able to pull the machine off the floor,
and burnish and regap the contacts.  And, you get some idea of how likely
the need base on how many plays it sees.

Pinball games are more service trouble than most operators want to deal with.
Anything we can do to make them more reliable is part of our business model.
And I hate repeat service calls.

Of course! My father-in-law (and HIS father-in-law) ran a route. So,
unlike an arcade -- where there IS a "back room" -- problems meant a
trip to <someplace>, often during THEIR working hours. (So, it's
neither convenient for you or them!)

Post by Don Y
If the device in question is on a mountain in tibet...  (yes!  <frown>)

Post by John Robertson
Would zener diodes across the contacts help reduce pitting?

An RC snubber might work better.
But, *maintaining* a pin table is half the fun!  (unless, of course, you
are in a business to make money from them!)

Ah, but you see that is the point of doing repairs that are better than
(improve upon) the original design.

Yes, John. My tongue was so firmly in my cheek that I looked like a chipmunk!
Pin tables -- even solid state -- are perpetual maintenance problems. Whether
its a broken rubber, solenoid, jammed coin mech, blown general illumination,
misadjusted targets, etc. SOMETHING *always* needs attention.

Or, is giving you clues that something WILL need attention.

I always prefered older tables to the electronic ones. So, that much *more*
maintenance. "Hmmm... the hundreds reel only advanced by 400 instead of 500.
Something needs to be regapped (and lets pray it's not a mechanism issue!)"

But, I think that was part of the appeal -- keeping such a machine in top
working condition was a point of pride.

[Until you run out of hours in a day/week/month/year to keep up with it!]

Back in the 80s I figured out why one
manufacturer's games were blowing up their driver transistors randomly and
published the answer in the trade journals of the day. It was a ground issue,
where the commons were done through Molex pins and as the pins aged the ground
connections generated resistance, which led to transistors not fully biasing
off and burning out. The fix was to beef up the ground connections and all
subsequent games we serviced never blew the driver transistors any more and
customers were happy!
Need I mention the factory never acknowledged the error and did NOT implement
my recommendations.

(sigh) Don't get me started on that! I had a piss-poor opinion of many of the
(hardware) "practitioners" in that era. I used to say that, as cavemen, they
rode around in cars with square wheels (think the Flintstone's car). And,
after a while, the corners of the wheels would wear off, leaving them ROUNDed.

At which point, they would promptly REPLACE them with nice, new SQUARE wheels!

<frown>

Of course, it's hard to know how to apportion blame. There was a lot of
emphasis on reuse -- not just designs but actual *assemblies*. If you're
only making a few thousand of an item -- and, that item is REALLY REALLY
REALLY similar to the last item you made! -- there is a strong incentive to
reuse "what worked" (even if it didn't work all that well!)

Software folks were always pleading for common-sense changes to designs
("Why do I have to serialize a bit-stream to drive the CVSD? That
sorely limits the types of sounds -- speech! -- that can be generated
and consumes 100% of the CPU! Why can't you put a frigging shit register
there, instead??? Even if *I* have to clock it, at least I won't
ALSO have to shift the data in the accumulator!") But, a lot of
pressure on cost (I wonder how much of a price increase the market
could have tolerated to improve the quality/appeal of the product and
reduce time to market?)

Larry & Eugene would have "wet themselves" for something as trivial as
a 2MHz design...

By the 80s they gained a reputation for unreliability that plagued them until
they finally closed shop in the mid-90s. Oddly enough the same company was
considered the Rolls-Royce of pinball up to the Solid State machines. Their
solid state games looked and played great until the ground issues started
randomly occurring (fried coils and driver transistors) and then operators
would get rid of that manufacturer's games and buy other brand's machines.

To be fair, you are also competing against the garage shops that want
to sell conversion kits (much cheaper than new games -- at least for video)
at the expense of the "original designs".

There are two cases I deal with - commercial and home clients. Both would
rather see less of my shop then they have to and I am happy to oblige them
because I hate fixing simple things that could be made more durable. Also we've
gained a bit of a reputation of machines lasting longer after being serviced by
us and I wish to improve that score.

Of course. I gave away all of my arcade pieces. But, always to someone who
*could* maintain them (technicians, etc.). I knew that, otherwise, the
games would quickly be unplayable and just collect dust in someone's basement.

[I've toyed with getting -- or making -- a Robotron (something along the
lines of a DINING ROOM TABLE, instead of a cocktail, so it would hide in
plain sight. A full upright would be like storing a refrigerator in the
living room! :<) But, suspect I would find too many excuses to play it
instead of doing "productive" work. And, SWMBO would quickly tire of the
hearing that deep bass "thumping" all day! (plus the next wave fanfares!)]

SWMBO is the only exception to the "could maintain it" rule -- but, then again,
she's got *me*! :-/

So, reducing contact arcing is very important to me and my customers!
We only get more business the better we fix the games!!

[john larkin]

On Fri, 23 Aug 2024 23:29:48 -0700, Don Y

Post by Don Y

Post by Don Y

I assume you have the normal back diode across the coil for clamping the
ringing voltage.

If you care about the life of the relay (and how well it "opens"),
a better approach is to add a zener rated at the coil voltage in
series with that diode.  The diode, by itself, delays opening of
the contacts and leads to more opportunities for arcing.  The
zener speeds up this transition.

Don't you mean that one could use a zener diode in place of the back-EMF diode
- or parallel with it?

In series. It speeds up the decay of the magnetic field (and, thus, the
opening of the armature) by allowing a higher potential to exist across
the coil while it is "opening". The ideal condition is with NO catch diode
(but that tends to fry solid state switches! :> ) You're making a
snubber, of sorts.

Essentially all modern discrete mosfets avalanche safely at ballpark
120% of rated abs max drain voltage. Try it.

The damage limit becomes average power dissipated, which would be low
in most cases of driving a small relay.

I recently designed a board brickwalled with about 130 power relays.
Each has a sot-23 logic-level mosfet on the bottom of the board,
snuggled between the thru-hole relay pins. No catch diodes.

Loading Image...

Post by Don Y

We've been using 1N400X (and 3A in some cases)diodes since the 70s for
protecting pinball driver transistors - not too worried about the decay time so
haven't really looked deeper into it before.

It really only matters if you are really concerned over the actual opening
time (delay) of the relay and/or how long you want the relay (contacts) to
last. (many of my designs have to have service lives of decades or more.)
<https://www.te.com/commerce/DocumentDelivery/DDEController?Action=srchrtrv&DocNm=13C3264_AppNote&DocType=CS&DocLang=EN>
will give you a quick overview.
I have better notes from (relay) application engineers but I'm busy baking,
tonight, so can only take quick pokes at my mail, etc.

Baking what? I'm getting whining that we are out of both biscuits and
bread pudding, so I won't get much electronics done today.

[john larkin]

Post by John Robertson

Post by Don Y

I assume you have the normal back diode across the coil for clamping
the ringing voltage.

If you care about the life of the relay (and how well it "opens"),
a better approach is to add a zener rated at the coil voltage in
series with that diode.  The diode, by itself, delays opening of
the contacts and leads to more opportunities for arcing.  The
zener speeds up this transition.

Hi Don,
Don't you mean that one could use a zener diode in place of the back-EMF
diode - or parallel with it?
We've been using 1N400X (and 3A in some cases)diodes since the 70s for
protecting pinball driver transistors - not too worried about the decay
time so haven't really looked deeper into it before.
Thanks,
John :-#)#

Is there a community of pinball machine people?

At my new office, the next-door neighbor's husband has pinball
machines, and I invited him over if he needs help with electronics.

The downside is that he plays drums too. I don't understand why anyone
would do that. Lotta noise, both cases.

[John Robertson]

Post by john larkin

Post by John Robertson

Post by Don Y

I assume you have the normal back diode across the coil for clamping
the ringing voltage.

If you care about the life of the relay (and how well it "opens"),
a better approach is to add a zener rated at the coil voltage in
series with that diode.  The diode, by itself, delays opening of
the contacts and leads to more opportunities for arcing.  The
zener speeds up this transition.

Hi Don,
Don't you mean that one could use a zener diode in place of the back-EMF
diode - or parallel with it?
We've been using 1N400X (and 3A in some cases)diodes since the 70s for
protecting pinball driver transistors - not too worried about the decay
time so haven't really looked deeper into it before.
Thanks,
John :-#)#

Is there a community of pinball machine people?

rec.games.pinball - but it has pretty much died. When google stopped
regulating posts a few years back it was overwhelmed by crap and was
abandoned by the remaining regulars except for a few diehards.

Nowadays it is the online forums that soak the natterers.

Post by john larkin
At my new office, the next-door neighbor's husband has pinball
machines, and I invited him over if he needs help with electronics.

If you need manuals/schematics I have hundreds scanned and a couple of
thousand waiting scanning. And there is a great resource site -
https:/ipdb.org for pinball photos, service info, and photos

Post by john larkin
The downside is that he plays drums too. I don't understand why anyone
would do that. Lotta noise, both cases.

Hope he wears earplugs! And solid state pinball games have volume controls.

John :-#)#


--
(Please post followups or tech inquiries to the USENET newsgroup)
John's Jukes Ltd.
#7 - 3979 Marine Way, Burnaby, BC, Canada V5J 5E3
[phone] (Pinballs, Jukes, Video Games)
www.flippers.com
"Old pinballers never

[Lasse Langwadt]

Post by Don Y

I assume you have the normal back diode across the coil for clamping
the ringing voltage.

If you care about the life of the relay (and how well it "opens"),
a better approach is to add a zener rated at the coil voltage in
series with that diode.  The diode, by itself, delays opening of
the contacts and leads to more opportunities for arcing.  The
zener speeds up this transition.

or use a fet that is avalanche rated or has build in clamping

[john larkin]

Post by Pimpom
Is it reasonable to assume that most small general-purpose DC relays
will not engage at 50% or the rated coil voltage? If not, how about 40%?
(This is *not* about the release voltage).

Most data sheets have a must-operate voltage and a must-release
voltage, but typicals are well, typically, very different.

Test a few! I'd expect that it would be very unlikely for an em relay
to operate at 40%.

We test parts and document whatever we learn.



FTR-B3 relay non-latch 12 volts JL June 2015

Coil measures 1.0K

Typical pickup = 7.7 volts spec is 9
dropout = 2.1 spec is 1.2

The 5 volt version pulls in at about 2.5 and drops at 1.4.

One non-energized relay measured 47 and 44 mohms on its NC contacts,
measured at both 1.3 amps and 100 mA.

JL Jan 2024