Does exist on the market a differentiator device?

[tireminnanzi]

Does exist on the market a differentiator device? Basically connected to a multiway switching to output a pulse of defined length everytime the deviator changes position on/off?

 

[Lucien Nunes]

I understand what you want to achieve and I would have expected to find a product but cannot immediately see a single unit that does this out of the box. To clarify:

• A single output is to be pulsed on both the rising and the falling transitions of the input?
• What should the output do if multiple input transitions occur within the pulse duration?

You can certainly get edge-triggered relays that will output a pulse on either the rising or falling edge, and connect one of each in parallel. But there would be no control of the behaviour for multiple transitions - a single longer pulse would be the unavoidable result.

What is the application?

 

[James]

6ED1057-3BA03-0AA8 | Siemens LOGO! Series Starter Kit for Use with LOGO! 8.3, 230 V ac Supply | RS

uk.rs-online.com

can be configured to do what you require.

 

[Lucien Nunes]

I was thinking more in the £1 category! Obviously it's easy to design and build with discrete logic if one is electronically minded, but perhaps if we know the application we'll get a clearer idea of what level of engineering the OP is looking at.

 

[tireminnanzi]

very fast answering, this blog is really amazing.

Yes something simple in the budget. We are using this new line of products from Bticino brand where the push button is missing, only the multyway switch (great aestetics but poor technics...) so I was evaluating to setup a small differential circuitry, but first I wanted to double-check what is on the market. No problem on multiple pulses management because it's a house-gate opener impulse that doesn't need accuracy in pulse length. I think the edge-triggered relay is a great idea and can fit this application, now I search for one. @James: I think a PLC it's too much in complexity/cost, but thanks for the answer too.

 

[Lucien Nunes]

As an alternative, perhaps you could take a standard Bticino switch mechanism and modify it to include a momentary contact? Or modify the 'click' mechanism or install a spring inside to return the original contact to the 'off' position? Modifications would invalidate the safety approval for electrical power usage, but if you are only operating an extra-low voltage trigger for the gate that might not be a problem.

 

[tireminnanzi]

it's not low voltage it's 230Vac, I found some candidates, but I think they only trigger at the rising edge not at the decreasing edge:

http://www.veltragyu.com/wp-content/uploads/TRE_time_relay_701-706.pdf

https://www.schalk.de/timer-relays-clock-generators/articles/uzr-u4.en.html?file=files/pdfs/db_uzr4_uzru4_en.pdf

Also it seems that the input signal is a dc, and can't be the 230Vac, I think I need to build it up analogically.

 

[Lucien Nunes]

The circuit can be very simple if there is a suitable DC supply of a few milliamps available. Is there a 5, 12 or 24V DC supply on the unit e.g. for a radio receiver, that can be used for the pulse circuit? Can you build small electronic circuits on perforated board? It's going to need about 10 components.

 

[DPG]

The circuit can be very simple if there is a suitable DC supply of a few milliamps available. Is there a 5, 12 or 24V DC supply on the unit e.g. for a radio receiver, that can be used for the pulse circuit? Can you build small electronic circuits on perforated board? It's going to need about 10 components.

I'm thinking a good old fashioned 555 would probably be suitable.

 

[Lucien Nunes]

The trick I think the OP is looking for here is to trigger on both the opening and closing of the switch, so that you just flip it once to open the gate. The 555 is a useful collection of bits that can be assembled to do many things (including what it does best, short the supply rails together when changing state) but I can't immediately see how to make it respond to both switch edges without extra active components.

Does the switch range from Bticino include intermediate switches (4-terminal switches for lighting control with more than two switch positions?) If so, take a look at this possibility:

C needs to be large enough for the unit to reliably respond.
R1 might not be needed, depending on minimum pulse length and threshold voltage of the input. E.g. if the trigger current is much less than the short-circuit current, R1 can be used to extend the pulse available from a given capacitor size.
R2 is only a nicety to prevent the caps being dead shorted by the switch, R2C = 0.1 seconds.

 

[DPG]

The trick I think the OP is looking for here is to trigger on both the opening and closing of the switch, so that you just flip it once to open the gate. The 555 is a useful collection of bits that can be assembled to do many things (including what it does best, short the supply rails together when changing state) but I can't immediately see how to make it respond to both switch edges without extra active components.

Does the switch range from Bticino include intermediate switches (4-terminal switches for lighting control with more than two switch positions?) If so, take a look at this possibility:

C needs to be large enough for the unit to reliably respond.
R1 might not be needed, depending on minimum pulse length and threshold voltage of the input. E.g. if the trigger current is much less than the short-circuit current, R1 can be used to extend the pulse available from a given capacitor size.
R2 is only a nicety to prevent the caps being dead shorted by the switch, R2C = 0.1 seconds.

View attachment 88675

Must admit I didn't read the full detail of the question.

 

[Lucien Nunes]

Using the above capacitor gadget to trigger a monostable (which could be implemented using a 555) would make the control circuit independent of the characteristics of the controller. Using passives only, the trigger input resistance and threshold are important, whereas if the monostable pulls solidly up or down it will work with any input. OTOH a transistor driving a relay might be even more universal still.

 

[marconi]

Here is my attempt at a differentiator or dual edge triggered detector. If I can find some mains rated opto coupled isolators in my man shed I will swap the red spst dip switch for a mains switch operating a lamp. This would be tomorrow though.

It uses three quad 2 input nand ics. I had to hand some 4011s.

?

 

[Lucien Nunes]

Excellent toolbox parts solution. I have in mind a single-package solution using an LM339 - a=input buffer, b=inverter, a couple of diodes and c + d=window detector.

I have to admit that many of my quick solutions these days are just PICs. As much as I like to design circuits, I can usually get there faster by programming. On one occasion I had so little time to make a gizmo to send abroad, I built the hardware and sent an in-circuit programming cable with it, then while it was in transit built a copy, programmed and debugged it and then sent the hex file for the end user to upload. At the moment of dispatch I hadn't actually worked out any of the logic or math, and there was quite a lot of floating point that I didn't have a hope of programming in time. By building generic hardware, I didn't need to think about it at that stage.

But I do like tricky hardware solutions. I am seeing an opto-isolator inside a bridge rectifier, maybe a unijunction transistor. How about a couple of ECC82s? What about triggering on the transit time of a C/O relay contact? Two mercury switches connected by a spring?

 

[tireminnanzi]

What do you think of such acircuit:

where V1 represent the full rectified and buffered sinusoidal and R2 is a simple relay coil.
From simulation the lenght of the pulse of the coil is 0.5s, I hope it's enough, or the capacitor value will need to be even higher than 470uF.

 

[tireminnanzi]

or better a solid state relay that has a very low control current. The draw back is that accepts only positive dc voltage not negative, so the output should of the circuit should be rectified too.

 

[Lucien Nunes]

if V1 is rectified and unipolarity there will need to be a DC path around the capacitor.

It's not now clear to me what supply you have available - AC or DC and at what voltage - and what the nature of the control input is; logic level, a relay coil or what? How far is the switch from the controller? Is it preferable to locate the circuit near the switch or the controller?

Although it is possible to construct a general-purpose circuit that is independent of most of the above parameters, the easiest and smallest implementation will probably benefit from being matched to the specific operational environment.

 

[tireminnanzi]

if V1 is rectified and unipolarity there will need to be a DC path around the capacitor.

It's not now clear to me what supply you have available - AC or DC and at what voltage - and what the nature of the control input is; logic level, a relay coil or what? How far is the switch from the controller? Is it preferable to locate the circuit near the switch or the controller?

Although it is possible to construct a general-purpose circuit that is independent of most of the above parameters, the easiest and smallest implementation will probably benefit from being matched to the specific operational environment.

Ok, I explain it better:
The input line at 230Vac (+/-20%) 50Hz is switched on/off, with this signal has to be generated a pulse of time in the range of 0.5 to 2sec to act on a 230Vac load (the gate opener).
Any other voltages needed (i.e. relay coil at 24V) should be taken by voltage step down from 230Vac.
I seek the easiest solution: from market or custom designed.

The 230V control switch can be several meters from the load (gate opener).

 

[marconi]

Re Lucien’s #14 I have in mind a single-package solution using an LM339 - a=input buffer, b=inverter, a couple of diodes and c + d=window detector.

I had one LM339N quad comparator ic open collector output left over from an earlier project I did. Lucien's idea - or my version of his idea - is attached. It uses one LM339, one npn transistor, two diodes and a few resistors. You can observe the brief pulse off of the left hand LED whenever the SPST dip switch transitions.

Two of the comparators detect input voltage excursions produced by the red spst dip switch at a high threshold of 2/3 Vcc and a low threshold of 1/3Vcc. The third comparator combines/compares the outputs of these two comparators and produces a very brief pulse whenever the input voltage is in the window ie in transition between switch states Viz Vcc 5.5V to 0V Or vice versa. The fourth comparator - a monostable- stretches this very brief pulse to make it visibly wink the left hand LED. The transistor does a signal inversion between the third and fourth comparator because the monostable requires a negative going pulse to trigger it.

 

[marconi]

Ok, I explain it better:
The input line at 230Vac (+/-20%) 50Hz is switched on/off, with this signal has to be generated a pulse of time in the range of 0.5 to 2sec to act on a 230Vac load (the gate opener).
Any other voltages needed (i.e. relay coil at 24V) should be taken by voltage step down from 230Vac.
I seek the easiest solution: from market or custom designed.

The 230V control switch can be several meters from the load (gate opener).

I understand your requirement is to produce a timed on pulse whenever the switch transitions on to off OR off to on.