Page 1 of 2

pcb impulse relay? ...or 'stepper', 'ratchet', 'sequencing'

Posted: Sat Jun 07, 2008 11:51 am
by geb
It's the nomenclature that's messing with me I'm sure.

I want to operate a dpdt relay, in both states, with one, single, solitary momentary switch. No separate electronic circuits. No programmable pcb varieties. No microcontrollers whatsoever. With respect to Occam, I just refuse to install another dedicated pcb in this project merely to do what I could essentially get away with using a little low voltage barrel solenoid driving an off the shelf dpdt latching pushbutton. With that arrangement I could switch dpdt back and forth with a single subsequent impulse of the same polarity each time.

I know I need a latching (probably magnetic if it's going to be pcb sized) dpdt relay for this. Every one on the market needs two separate switches for control. But none of the pcb sized ones will 'ratchet' (allow the use of a single input of same polarity to move the arm back and forth).

This quest is just about driving me insane. I really hate the idea of having to stick a noisy little barrel solenoid and noisier little pushswitch in this project just so that I can swith dpdt with a momentary 'tact' switch on the user panel.

Anyone have any better ideas? I've been to about two dozen other forums so far with this and everyone has come up blank. I've asked dozens of engineers and salespeople at the major manufacturing and supply places.. to be told that it is impossible to purchase anything for this purpose. I'm so demoralized I could almost cry.

Thank you so very much.

Posted: Sat Jun 07, 2008 12:12 pm
by Robert Reed
I'm not sure exactly what your physical limits are, but at first glance it looks like you may need a little electronics to do this. My first thought would be a reed relay driven by a 'D' type flip-flop wired for toggle operation. Not much room needed here - just (2) 14 DIPs.

Posted: Sat Jun 07, 2008 4:06 pm
by Externet
Do not cry, will take you nowhere...:sad:
If I understand what you want, you only need to canibalize the 'turbo' switch from an old PC case.

That's it. DPDT (a few are SPDT) No relay, no electronics, no magnetics, no nothing. One push is latched on, one more push of the same button is latched off, quiet to please you, and for pennies if you have to pay for it. :grin:

Miguel :smile:

Posted: Sat Jun 07, 2008 4:15 pm
by geb

Is a turbo switch different from a latching pushbutton switch that I can purchase from digikey? I can't use a latching pushbutton switch on this panel. I need to use a momentary spst. I have to do this. This is a parameter that cannot be gotten around.

What is a 'turbo' switch? Is it a momentary switch? Is it spst?

I explained that I need to do this with a single momentary switch.

If I could do it with a latching dpdt I would have been done with this three minutes after I started.

Please tell me that the turbo switch is actually able to do this and I will weep with joy.

Posted: Sat Jun 07, 2008 4:27 pm
by Externet
Start weeping :smile:

Could also be installed under a membrane panel if the membrane is flexible enough, as the travel is about 2mm

We do not need the first garbled line, saw that at the Audio DIY forum too. A bunch of people will come up with very convoluted solutions.


Miguel :grin:

Posted: Sat Jun 07, 2008 6:06 pm
by MrAl
Hi there geb,

Switching circuit often have a multitude of solutions. I will
suggest a few here and you'll have to pick one you like.

What you probably want is a "ratchet and cam" relay. These have
(as one would guess) a ratchet and cam inside, so that the first
pulse turns the relay on and the second pulse turns the relay off.
Where to purchase these i dont know offhand, perhaps a search.
You might also build one if you are into constructing mechanical
things like this. A manufacturer called "Omron" does make them.

If that doesnt sound good or you have difficulty finding one, then
perhaps a solenoid operated by the momentary push switch,
and the solenoid is mounted in close physical proximity to a
second switch which is a push switch too but this one is the type
"push on push off" ie latching. The solenoid and second switch
could be mounted inside the case, while the small mom push on
the front panel with no problem. The advantage of this technique
is that the parts should be available in many places, all you need
is one solenoid and one momentary push switch and one push on
push off switch that doesnt take a huge force to operate.

Next, this might also be done with more than one relay, perhaps
with four relays, but possibly even more (six, eight). This would be
the most complex wiring of the three.

If you can add simple electical elements such as capacitors and
diodes and possible resistors there are other solutions too.

Posted: Sat Jun 07, 2008 7:00 pm
by dyarker
I think you mean "PUSH-ON, PUSH-OFF" switch. Works like ball point pen. For example Alcoswitch part number MPE106D. The latching mechanizm is in the switch, switch contacts to coil of normal relay.

Posted: Sat Jun 07, 2008 8:31 pm
by Lenp

I think this is the minimum you can get away with to do your push on/push control with a 'simple' switch. These work with a solenoid coil that operates microswitches via a clever toggling cam. A quick search on line, turned these up. They are (were) made made by PB and I used them in the past for remote power control of custom audio systems.

Hope it helps!


Posted: Sat Jun 07, 2008 8:45 pm
by Lenp
Quick note
Anybody remember the Alco FRE-103 series of 'isolation relays'?

It was a 120V plug-in cube, with a NO/NC outlet and a pair of pins for the control. It was, in effect, a low voltage transformer/relay but it was all done in one assembly. It was a a two coil assembly, with a box like lamination that had an armature at one end. 120V went to one winding and the control to the other winding. When the user's control pins were shorted the coil was shorted and the magnetic field increased attracting the armature that switched the contacts.

Or has anyone ever done ac switching by putting the load in series with the primary and shorting the secondary winding?



Posted: Sat Jun 07, 2008 9:33 pm
by MrAl
Hi again,

In my previous post above i suggested cam relays as the best choice,
but i also mentioned using standard relays alone. This turns out
to be a bit complex, but it does work.

Here's an example of a circuit that does PUSH ON, PUSH OFF,
using seventeen (ha ha) normally closed relays. The relays only
need one set of contacts each.

Im sure there are simplifications possible.

This would work with either DC or AC relays too, or normally
closed reed relays.


Posted: Sat Jun 07, 2008 11:58 pm
by Bob Scott
Lenp wrote:Or has anyone ever done ac switching by putting the load in series with the primary and shorting the secondary winding?

I have only thought about it. It should work. Just check that the primary winding has enough current capability, and that the primary winding resistance is not too high in series with the load, and that the primary leakage current (inductive current) is low enough when off.

I think that using the secondary low voltage winding for control of a 120V circuit means you can use class 2 low voltage wire for the switch.

Posted: Sun Jun 08, 2008 2:25 pm
by Engineer1138
Why don't you want to use electronics?
The system will need to maintain state; that will be done either electrically or mechanically. I'm sure that this could be done with a few relays, but then you've just added more parts and that is significantly more complex than the simple electronics that would accomplish the same thing.

Now if it were just an academic exercise, I'd say you could use a solenoid and a "heart" cam, but you can apparently find those already in one unit.

You could always build the rotary mechanism in a ball-point pen -- that uses a momentary input to control two states -- and attach contacts to it, but that seems a bit...much.

So, asking again, why are you so dead set against something as simple and cheap as a 2-transistor multivibrator?

Posted: Thu Jun 12, 2008 9:51 am
by geb
The skinny...

I am adding a process to an inverter ac/dc tig machine. The machine uses very powerful HF for arc start and maintainence. Some of the electronics that I've already used in this project, my led meters for instance, are already giving me headaches as I seek to isolate/shield them from this HF.

The inverter welder itself is chock full of microcontrol; ICs, pcbs with hundreds of smds, etc. I don't know how the engineers accomplish the protection of all of this, but it must be some feat. I read somewhere that if you remove the plastic covers on one of these machines and initiate the HF in a totally dark room you can actually see a faint glowing aura surrounding every metal component, wire, and trace within.

So, never having studied EE in college, but with just enough skill to get myself in trouble, I'm adding a water cooling circuit to my welder. I really don't want to add a separate pcb microcontroller, with battery backup, to enable me to make a single pushbutton into an 'impulse' or 'flip flop' switch unless there ends up being no electromechanical means left on earth for me to do it. I don't know how to shield it 100 percent from this HF problem, and beyond that, it's just philosophically painful to use a dedicated pcb microcontroller to do something so 'simple'. At least I thought it was going to be simple when I began.

It would be difficult to explain exactly why this must be accomplished with one momentary instead of two. If I could use two I'd be home free. That is all I'd need to switch a 12vdc magnetically latching dpdt.

It seems absurd that I would have to either make one of these out of a ball point pen mechanism attached to a standard relay (I had already considered that) or a 12vdc push solenoid attached to a latching dpdt pushbutton (a more practical route perhaps). I know I need to back off here and regain my sanity. It feels a little unreal - like being told you can't purchase a flat bladed hand tool to turn a slotted screw.... you'll have to build a microcontroller for it.

I have seen a solenoid and a heart cam in one unit. I even found it for 9.95 at NOS surplus. But these are monsters. I've only got a couple of square inches maximum to spare. I've gone to great lengths to keep size and weight down on this project. If I could find a solenoid and heart cam (or similar) miniaturized in a pcb power relay, I'd be there.

Posted: Thu Jun 12, 2008 9:58 am
by geb
sorrry.... duplicate message accidentally

Posted: Thu Jun 12, 2008 4:34 pm
by Externet
OK, let's see if we can fine tune the challenge... :smile:

You have a couple of square inches on a PC board,
The contacts needed are DPDT,
You want a single momentary normally-open switch to flip them,
You do not want a latching DPDT push-on push-off switch,
It should be activated under a overlay panel, as a single tact switch,
The switch motion travel should be about 1mm as tact switches are,
What is the current the contacts will carry ?
What is the coil operating voltage ?
What voltage is available as supply if a circuit becomes mandatory ?
What is missing or wrong from the list ?

As preliminary, is there anything here you can evaluate?, Some are larger than desired, but, would a shrunk one of these work ? :

In 24V: ... rodID=1459

In 12V ... 990_DS.pdf

At Digikey: ... c1-12.html

NTE numbering: ... nkCode=asn

Check "711" and "712" here: ... 500256.PDF

Miguel :smile: