That may smooth out the reed switch pulses, but the reed switch won't last long switching at double the line frequency.

I presume "long" means, the Reed might fail ?

 

Hello,

How about the hall sensor as mentioned, with an amplifier and a schmidtrigger circuit to create square wave pulses.
When these pulses are fed into a missing pulse detector, it will very fast react on the switching off.

Bertus

 

Here is a schematic that I recommend. Some of the the other members of this forum may have some changes. With the optocoupler the output when the relay is energize is like a full wave rectified wave. The 74LS132 is a schmitt trigger and the output becomes a gate at a TTL level, and the input to the opto coupler is in series with one lead of an existing relay.

 

I found an interesting document comparing reeds to hall effect sensors. They seem to favour reeds. As seen in the specs below, reeds can handle over 1 billion operations.

 

I found an interesting document comparing reeds to hall effect sensors. They seem to favour reeds. As seen in the specs below, reeds can handle over 1 billion operations.

View attachment 130091

Wow, that is a very heavily slanted comparison. Some of the points are true, and if a reed switch meets your needs for any given project, it will probably be easier to implement, but this chart greatly exaggerates the situation. There are lots of hall effect sensor and switch options that require little, if any, additional circuitry other than requiring a constant supply voltage.

My first "real" electronics project was designing a hall effect switch to replace reed switches in a machine's user interface. In order for the switching to feel right, the trip points needed to be at relatively precise locations. The reed switches we used had huge tolerances (~40% variation in trip point if I remember right) and so each machine required tedious mechanical calibration. Using hall effect sensors with surprisingly tight tolerances I was able to design a very simple board that cost less than the two reed switches it replaced and requires no calibration. Just bolt it down and it works.

Sorry for the mini rant. It's not especially relevant for your situation where sensitivity tolerances probably wouldn't be an issue. I just happen to love hall effect sensors and hate to see them slandered! What can I say; I'm a fan-boy. They just made our build process and service procedures so much easier, it's hard not to love them!

 

Is it acceptable to remove a wire from a relay coil, feed it through a current transformer (like feeding it through a small doughnut), and reconnecting the wire to its original position? This will give you an excellent signal to turn into a trigger for your effects, while retaining complete galvanic isolation from the original circuits. This would be a much more reliable way to create a signal than wrapping wire around the relay coil.

ak

 

Is it acceptable to remove a wire from a relay coil, feed it through a current transformer (like feeding it through a small doughnut), and reconnecting the wire to its original position? This will give you an excellent signal to turn into a trigger for your effects, while retaining complete galvanic isolation from the original circuits. This would be a much more reliable way to create a signal than wrapping wire around the relay coil.

ak

This sounds like a winner to me!

 

those are open frame and pretty large relays. microswitch should also be able to do the job and without any conditioning

 

I presume "long" means, the Reed might fail ?

Yes long, as in period of time.
I'm concerned about mechanical fatigue causing the reed to break if it's being operated at a switching rate of 120Hz.
Even if it lasts a billion operations, as the TS stated, that's less than a 100 days of continuous operation.

 

If all the relays have one side of the coil connected directly to the same side of the 50 v AC supply then why not just solder a wire to the switched end of the relay coil. This could then be connected via a suitable value resistor to the type of opto coupler that uses two reverse connected LEDs on the input side (So that it works with either input polarity) You would need a small amount of filtering on the output to get rid of the break in output near the zero crossing point of the AC waveform.

Les.

 

Then a Hall sensor and a couple of transistors can do the trick I believe.
One transistor to keep a cap discharged as long as there is a pulse (120Hz) and another to switch when there is no pulse (when cap charges)

 

so many options

 

Thanks for the excellent suggestions. This forum rocks! I have a lot of experimenting to do which the fun part

 

Question not answered-- Is one side of 50V AC supply at ground ? If true, then a V divider, diode cap feeding a Schmitt trigger.

 

Another question : Is there a small relay driving the big 50VAC relay ?
Or a small triac ?

There should be something to switch on the 50VAC to the relays.

 

For those who are interested I have uploaded a schematic of a typical late 60's EM game. Hopefully it may answer some questions members had.
View attachment t3.pdf

 

Looks promising. One end of relay coil connected to 50 V AC, other end to switch. Values to be worked out, but start with hi Z V divider, 50 k & 6.8 k, diode, 10 uF cap. shunted by 10 k, fed into Schmitt trigger. A mA load on relay should have no neg. effect. Suggest 74C914 Schmitt trigger.

 

Looks promising. One end of relay coil connected to 50 V AC, other end to switch. Values to be worked out, but start with hi Z V divider, 50 k & 6.8 k, diode, 10 uF cap. shunted by 10 k, fed into Schmitt trigger. A mA load on relay should have no neg. effect. Suggest 74C914 Schmitt trigger.

Now we are getting somewhere.

 

Agree, but - somewhere the TS said something about leaving the original electrics in the original states, as in no physical connections. Then he softened and was ok with removing a wire, feeding it through a CT, and reconnecting it. If all of the relays always are switched to GND, that is perfect for a 100K impedance voltage divider and a 2N3906.

ak

 

AK might be right again as I can find no retraction from " no modification". Would it be to distructive to just add a single wire to
each relay- even by a miniature alligator clip ?

As to winding some turns, it does not look promising. I wound 7 turns on a 120V AC relay & only read .07 V. 50 V coil would be around .02 V / turn. Even 10 turns would be tedious to wind.