The mod does not work unless you are
using parallel wound flipper coils (which
only exist in 50volt versions). so the short
answer is, don't waste your time because
you can't get parallel wound 28 volt flipper
coils, only series wound.
now it begs the question, can you convert
a series wound flipper coil to a parallel
wound coil? hmmm....
IMHO, I would imagine a cap of that value was placed there to keep
noise spikes off of a bus line, not necessarily to prolong the life of
the EoS switch, The best way (once again IMHO) to reduce arcing is to
clean the switch well. A dirty or pitted switch will arc. The arcing
accelerates the degradation of the contacts and will shorten the life
of the switch.
If a simple cleaning doesn't do the trick, I would suggest replacement,
or (if you have a lot of time on your hands) burnishing the contacts.
Happy New Year!
I'll admit, there's alot I don't know about these things, but from a
purely electronic standpoint, I know caps are used frequently to
"debounce" a switch.
To reduce a spark, it would be limited essentially to 5 RC time
constants, which in this case using solely the resistive component of
the coil (we'll call it 10 Ohms just to keep the math easy) R*C*5 would
be .1mSeconds (110uSeconds) before the cap would be fully charged.
I would imagine the duration of those sparks, due in part to the
ionization of the surrounding air, to be much longer tha 1/1000th of a
second. Just 'cause it's pinball, doesn't mean it can defy the laws of
physics (Although most of my SDTM shots seem to. ;-0 ).
Maybe I'm coming at this from the wrong angle, if so, I'm quite
interested in understanding the physics behind this one.
Thanks for bringing this one up!
There's the real question! Nope, no harm that I can think of. ;-)
Enjoy your game!
Well, to put it frankly, He's Wrong!
And i think if you ask him in email,
he will confirm that using the cap
with a series wound flipper coil
will NOT help decrease the spark.
if you goggle I do believe he admits
to that error.
The cap just does nothing with a
series wound coil. Now if you have a
parallel wound coil, YOU CAN SEE
the effect the cap has! It really does
make a big difference. But with a
series wound coil, you're wasting your
We can go into a long discussion about why a capacitor across a contact
reduces the arc, but clearly to me, the capacitor prevents the rapid
buildup of voltage across the contact as it is separating. This
prevents the initial arc, and the creation of the subsequent longer one
that sustains electrical current.
Edward Cheung CARGPB26
Since both are just two wrappings of wire around a plastic bobbin, I
don't see why rearranging the taps on the back would make a difference.
So, unless there's something obvious I'm missing, yes, you should be
able to rewire a "series" wound coil in to a "parallel" one.
| David Gersic http://www.zaccaria-pinball.com |
| Cows in orbit: The herd shot around the world. |
| Email address is a spam trap. Visit the web site for contact info. |
Isn't the series wound one the high power is a tap on the whole
(single) strand of wire on the coil, and the parallel wound is one wire
of shorter length/higher strength, and there is a seperate winding for
the hold (longer/lower)?
If so it would be difficult to make a series a parallel. If not I'm
not understanding it right.
Game is Black Knight LE. On the far left are the two lower playfield
flipper assemblies with the capacitor attached, on the far right are
the upper playfield flipper assemblies with no capacitor attached. All
are brand new flipper assemblies except one of the left flippers is
using an old coil (wanted to test a knackered coil whose resistance is
only 2.8ohms). The right flippers can only work if the left work (two
stack leaf switch) ie if the right fires the left must have fired also.
The top right hand flipper has the EOS obscured by the coil but if you
look closely you can still see the reflection of the flash, lol you
have to blind to miss it but try and spot a flash from the left hand
As they say, a picture proves a thousand words.... without a doubt
adding a cap to your series wound coil makes a MASSIVE reduction to the
For those of you that think this doesn't work or isn't worth doing then
I pity your EOS's. Clay, your feedback will be interesting. Thanks Ed
for taking the time to post a picture on your web site, all it took was
a little spark......
Wow that's really cool.
But ya gotta do me a couple favors here
before I give-in on this.
With the game off, measure the resistance
of the two flipper coils used.
But more important, could you please
refilm this?? I need you to disconnect
the 2nd set of flipper coils from *both*
Because Black Knight is a 4 flipper
game, I'm not quite believing your
video. And i don't quite understand
what you are saying below about the
second pair of flippers.
Is the spark i'm seeing the EMF from
the 2nd pair of flippers? Or is it the
lower flipper EOS switch? It is rather
The EMF happens because as the
2nd pair of flippers (the upper flippers)
lose power, and the flipper coils'
magnetism collapses, you can get a
backward of high voltage thru the
switch on the lower set of flippers next
to the EOS switch. I just want to make
sure your video does not reflect that.
So if you don't mind, could you repeat
that with the 2nd set (upper flippers)
completely removed from the circuit?
An even better idea would be to film
this on a game with just 2 flippers.
That would really remove all doubt.
My comments about series versus
parallel flipper coils and the spark
suppression caps reflect what Williams
said in their documentation on the
subject. So i didn't just pull this out
of the air.
Also in my experience when i have added
the cap to series wound coils, visually
I saw no spark supression.
See also automotive "points" style ignitions, and their use of a
"condensor" (ie: capacitor) to prolong the life of the switch points.
| David Gersic http://www.zaccaria-pinball.com |
| I am NOT a cynic - I just remember last time too well! |
I'll remeasure the resistance later today, from memory the new coils
are around 40ohms each.
I'll run a series of tests. I'll remove the caps and film all flippers.
Next I'll disconnect the upper playfield flippers and film the lower
flippers. I'll then add the cap to one of the lower flippers and
compare them side by side (still with the upper playfield flippers
disconnected). This will give an example of a two flipper game. I'll
post later on with another link.
I have a Williams System3 at the clubhouse.
It's a Phoenix, so it only has 2 flippers.
I have some good video cameras too.
So i will film the flippers tonight with
and without caps. And i'll show in the video
how i have the caps wired, and the resistance
of the coils.
I think this will be a much better test than
using a black knight. The 4 flipper issue
and how it powers the additional flippers
just confuses the issue. it's not a very
good test vehicle for this.
40 ohms could be the resistance of the
hold side of the coil. but that's why too
high for the power side. I'll measure the
resistance tonight on the Phoenix, and
show it all in the video.
Good point. This might be how the practice orignated.
Here is my version (using a BK but with the upper playfield flippers
taken out of the circuit), details as below:
right (older coil) has 0.7 ohms on the primary and 24.5 ohms on the
secondary. EOS gap is 1/8 inch.
left (new) coil has 1.0 ohms on the primary and 26.0 ohms on the
secondary. EOS gap is 1/8 inch.
Both EOS's are new, as is the rest of the flipper assembly.
The big difference between our approaches is that I, and Ed, placed the
cap across the secondary winding. I initially favoured the cap across
the EOS switch (as you have) but after a long discussion with an
ignition expert on old cars covering matters such as how collapsing
magnetic fields work in dual coil windings, secondary effects and
backwash, it went there. This is especially noticeable on the flipper
without the cap, you can see sparks thru and around the secondary
It is also why I have noticed no difference in flipper power.
Unfortunately as I compressed the video the sound got out of sync
(sound is late) but you will get the idea.
Edward Cheung CARGPB26
But is there a chance you could put up a still
photo of how you have the Cap soldered in place?
I mounted my caps just like Williams did. I don't
quite understand how you did it.
I don't know either. I think the effect is very minimal
on series wound flip coils. Now on parallel it's
dramatic. But it may be more dramatic because
as seen in the video, I'm using an Earthshaker
as the parallel coil test bed, which is a 50 volt
game. (The spark on a 50 volt coil should be
greater than a 28 volt coil.)
I'm not taking the caps off the series wound
Phoenix game either. Not because they are
really doing much (they aren't), but mostly
because they aren't hurting anything. And it
would just be more work to remove them.
To further muddy the waters, again I got my
info directly from Williams about the cap and
how it has little to no effect on spark suppression
on series wound coils. I figure they know something
about this stuff. But again, it's not hurting anything
to have the caps, so if you have them, put them
on and forget about it. There really is no
This may have some impact on EOS spark
suppression too. I do know that older
flipper coils CAN be weaker (in flipper power)
than the exact same coil which is new.
Again this is due to the inductance of the
coil (given the exact same resistance).
It's not talked about much here, but
inductance is a factor in flipper coil strength,
at least to some degree.
Why mess with the original design? I thought that spark was necesary to
clean the oxide off the contacts (and thus provide a stronger power stroke).
<c...@provide.net> wrote in message
The cap is supposed to reduce contact pitting. Maybe as Clay says it's
not needed on the older machines because they're 28volts instead of
50v.... the wear may happen at an acceptable rate vs. the 50volt
flippers which will have a stronger arc.
I think the other key is that the EOS switches being adjusted furthur
out, causing a larger spark will do more damage. This is purely from
one of those "eye of the storm" type globes where there's "lightning"
inside - I figured out a while ago you could get the spark to happen
OUTSIDE the globe, if you lick your finger, rub it on the globe, then
pull the finger back slowly. It was really cool until I was trying for
a longer spark, got it to go about an inch and it burned a hole in my
finger. (A small one, but a hole just the same..... boy did that
So, maybe the EOS is the same way?
The cap is soldered across the lugs, middle to ground or across the
secondary winding. Same as Eds and you can see his clearly in his photo
on his page. No difference to yours, just the installation point is as
close as possible to the windings, not to the EOS. I know, this is at
odds to the suggestion that it should be as close to the contacts as
possible but I was given a reason for this, although the very long
technical explanation went slightly over my head but it was specific to
series wound coils and collapsing magnetic fields.
I'm going to dig out my old books and have another look at this. The
cap in use was intended for a 50V circuit, not 28V. Also there is an
optimal resistance one should aim for in the circuit before the cap,
but this is dependant upon an RC network aiming for a contact voltage
of X at X time after the contacts have opened. What this was meant to
be only a Williams engineer could tell us. I'll have to make some
assumptions unless anyone out there knows what this should be.
I also vaguely recall reading somewhere about the use of zener diodes
in line with the cap, again I'll have a dig and post anything I find of
interest in a new thread.
1.The inductor part of the coil(s) tries to keep the current flow the
same as the steady state; the 'ON' state being more interesting and
the cause of the arc.
2. The voltage supply of the coil sees a voltage reversal of twice the
input value. This means a 50 volt supply is being drawn to -100 volts.
This is why the diodes are there. They are pulled into forward bias
and shunt the current to prevent this from happening. It's why a game
with damaged or missing flipper coil diodes can act so badly.
3. The circuit model of the flipper EOS switch opening looks like a
huge spike of infinite amplitude and zero duration. In the real world
it has some finite maximum value and a short duration. If you look at
the voltage component of that spike, you can see why it causes arcing.
50 volts in a normal environment will not jump very far, and 100 volts
not much more. But an arc of 500 MHz frequency and 10k volts will arc
further than the gap of most EOS switches. It is the high voltage of
the pulse repsonse the heats and ionizes the surface of the contacts
and causes pitting.
4. A capacitor tries to keep the voltage of a system constant, and
resist changes. So when the pulse response occurs, the cap has the
effect of lowering the pulse response amplitude from thousands of volts
down to something more managable. I have not measured it, but I bet
it's like 50-60 volts in a 28 volt system, and 100-150 volts in a 50
volt system. That's why you have a capacitor rated at 250 or 1000
volts on a 28 or 50 volt flipper coil. It is these lower transient EOS
voltages that cause a smaller arc, not the cap absorbing the energy
directly. The caps Williams used are obviously too small to totally
counteract the pulse repsonse, but they make it good enough so that a
$5.00 EOS switch can last years. You wouldn't use a $30.00 cap to save
a $5.00 switch, right?
5. The amount of energy a capacitor has to absorb is different based on
if you have two inductors in series or in parallel. Inductors,
capacitors and resistors all hooked together are called tank circuits
and require frequency ranges to more accurately determine how they
work. I would suspect some improvement in a series wound system is
believable, but I have not done the circuit models to take a stab at
Unless there is a great outcry for more geekness, I'll stop here and
say the inductance of these coils is really the issue, not the
> Another thing that we didn't consider
> is INDUCTANCE. We all look at coils
> as resistors, but in reality, they are more
> than that. They are inductors too. And as
> a coil gets older, and the enamel coating
> on the wire breaks down, the inductance
> (Henries) of the coil changes.
> This may have some impact on EOS spark
> suppression too. I do know that older
> flipper coils CAN be weaker (in flipper power)
> than the exact same coil which is new.
> Again this is due to the inductance of the
> coil (given the exact same resistance).
> It's not talked about much here, but
> inductance is a factor in flipper coil strength,
> at least to some degree.
> Clipped for length
This time constant, in pinball coil applications is going to be in the
1000th's or 10,000th's of a second. Once 5 of these time constants (5mS
at most) have transpired, 98% of the magnetic field has been derived
from the current and the coil appears to be a purely resistive load.
Shortly after this point (within 200mS) the flipper mechanism connects
with the EoS switch. At this point another purely inductive load (the
secondary coil) is introduced into the cicuit.
Once again, for five time constants (now of the secondary coil) current
is pulled from the circuit, partly from the primary coil's collapsing
magnetic field to develop a new magnetic field around the secondary
Once the flipper button is disengaged, current ceases to flow in the
circuit, both magnetic fields collapse the counter EMF (surge created
from the magnetic field collapsing, generating an opposite voltage
difference across the coil) will be shunted by the coil diodes
preventing these spikes from damaging attached components.
IMHO, The only time the difference between parallel or series wound
coils would make a difference is in the discharge cycle since they were
not charged simultaneously. During the discharge cycle, parallel coils
would have a lower inductive reactance than either segment and
discharge current at a far greater rate than the series wound
Nice post also.
Ron -- CARGPB7 -- (Change hot to ice to email)
In article <1168060278.2...@q40g2000cwq.googlegroups.com>,
As the EOS switch is separating, the separation distance builds rapidly
from zero (closed) to the final value. Air's dielectric resistance is
proportional to this gap's distance. Thus the voltage at which an arc
will start also builds more or less linearly. Thus, at the initial
very tiny separation, the arc voltage is low, and increases.
As Al and Ron have pointed out, the current through the inductor does
not change instantaneously (it is a fly wheel to current), and instead
of through the EOS switch, this current now flows into the EOS
capacitor. This current and the capacitor's value sets the slew rate
of the voltage (V is the integral of the current divided by
capacitance). Bottom line is: If this slew rate is slower than the
rising arc tolerance voltage, there will not be an arc. That is what I
was referring to in my previous post that it makes technical sense that
the capacitor is a good idea.
So the larger the capacitor, the slower the slew rate of the voltage
upon opening of the EOS switch. The drawback of too big a capacitor is
that the cap is shorted as soon as the flipper button is released, and
the EOS closes. So normal arc suppresion networks use a small series
resistor with the capacitor.
Edward Cheung CARGPB26
AAAGH! You guys are making my brain hurt!
I'm gonna solve this once and for all. I've just taken a coil with the
cap mod out to my car, and have hooked it up to my car battery. I'm
going to pinch the EOS switch together and see if it sparks or not.
Okay, thumb on one blade, index finger on the other and here we
I was still scratching my head on the cap issue, but you're right! The
EoS switch in the early moments of engagement will appear capacitive. A
parallel capacitor (regardless of coil composition) will assume the a
part of the available current and reduce the potential of a spark. I
would still be concerned (as you pointed out) about the collapse of the
magnetic field and the counter EMF current surge being absorbed by the
cap until the threshold voltage of the diode was breached. Perhaps the
cap would also delay the diodes conductance.
However, I think the most intuitive statement in this thread to date
was the gentleman that stated sometiing along the lines of
"Hey....we're talking about a $2 EoS switch ...right?
That was fun, huh?
Ed, my little blurb in the middle was just giving Al a ribbing. I did
however sleep at a Holiday Inn Express last night :-)
Ron -- CARGPB7 -- (Change hot to ice to email)
In article <1168111472.7...@42g2000cwt.googlegroups.com>,
Edward Cheung CARGPB26
A series wound coil however is turning 'ON' when the EOS switch is
opened. For high power operation, the high impedance series part of
the coil is bypassed by the EOS switch. When the switch is opened, the
high resistance coil is added to the circuit. Since this is not the
high energy side of inductor operation (remember the inductive spike
response is NOT present), then the only voltage available to overcome
the dielectric properties of the air (and create an arc) is the 28 or
50 volts of normal flipper power. It will still cause pitting and
burning, but not nearly to the degree parallel coils will. So adding a
capacitor may have the effect of lowering EOS arc in series wound
coils, but Williams would be right in saying the improvement is so
small vs. cost it was not worth doing in a production environment.
We're back to spending an extra $5.00 on each game when the original
EOS will likely outlast the plunger and link without a capacitor.
For the mosst part, it will not hurt to put one on, so go for it if you
like. If the capacitor you add shorts out though, it will have the
effect if bypassing the EOS switch, so don't use one that's too small
or cheaply made.
I have seen many series wound games that have original EOS switches,
but I doubt many System11 and newer games have still their original EOS