11
\$\begingroup\$

I have a complex circuit, but my problem only relates to one small part.

I have a relay, which is energised by pulling the low side low using an NPN. I would like a pushbutton to operate the transistor (and therefore the relay) for a time (500ms ish) then deactivate it and wait for the next push (ignoring the button if held). I understand how I can achieve this using timers (or a MCU), but I was hoping there might be a more simple and compact solution...

The image shows a simplified schematic of the starting point...

relay transistor and button circuit

\$\endgroup\$
2
  • 1
    \$\begingroup\$ Thank you...I understand how to achieve this with a 555, I was hoping there might be a simpler smaller solution with an R/C couple or some such. \$\endgroup\$
    – Jon
    Commented May 17, 2013 at 10:15
  • 2
    \$\begingroup\$ Oh, sorry, it says without NE555. Never mind. This is not a duplicate, my bad. \$\endgroup\$
    – user17592
    Commented May 17, 2013 at 10:15

4 Answers 4

Reset to default
5
\$\begingroup\$

If your question is simply can you do it without a 555 or a MCU, absolutely! You could use a 74HC123 retriggerable monostable multivibrator, or any of a dozen similar .

Alternatively, you could do it with discrete transistors, but that would almost invariably take more parts then a 555, and probably take more board area.

No matter what you do, you're not going to get that much in terms of size-reduction over a 555. The whole point of an IC is to reduce the parts count.

Really, the best you could hope for in terms of board/part complexity reduction is a fancier timer IC. The LTC6993 TimerBlox: Monostable Pulse Generator is a pretty cool part, and it only requires 2 external components.

enter image description here

Note: in this situation, it's being used as a missing pulse detector. That is why the input is labeled "modulated carrier".

\$\endgroup\$
2
  • 1
    \$\begingroup\$ I was hoping there would be a simple circuit with a few discrete components but it's looking like a timer is the best way forward for simplicity's sake. \$\endgroup\$
    – Jon
    Commented May 17, 2013 at 12:17
  • \$\begingroup\$ @Jon - You're not going to be able to reduce the parts-count of a 555 one-shot by going discrete. The entire point of the 555/any IC is to reduce the overall part-count. As such, pretty much any solution that reduces the total number of parts will invariable involve an IC. \$\endgroup\$
    – Connor Wolf
    Commented May 18, 2013 at 5:06
2
\$\begingroup\$

Here is a circuit that will come close to what you want that uses discrete components.

enter image description here

This circuit will have the relay coil in place of the 120 ohm resistor. Remember to add a diode in reverse bias across your relay coil to protect the transistor from over voltage spikes when the transistor goes off.

There are a few things about this that you should be aware of. The relay drive will persist as long as the the switch is until the 100uF capacitor no longer supplies current to the NPN base. This behavior will make the relay turn off even if the switch is held closed for a long time. If the switch is held shorter than the capacitor charging time then the relay will shut off when the switch is released.

\$\endgroup\$
7
  • 1
    \$\begingroup\$ This is definitely what I am looking for, but I think I would struggle to make that smaller than a 555. Looks like a timer is going to be the way forward. OR...someone should produce a physical one-shot pushbutton (if there is already such a thing I can't find it) \$\endgroup\$
    – Jon
    Commented May 17, 2013 at 12:19
  • \$\begingroup\$ @Jon - I agree that you should be looking at a 555 for this application. As you can see from the posting that I made there are some downsizes to the minimal circuit. The capacitor used this way also causes a soft slow turnoff of NPN transistor and could lead to a not so sharp characteristic of the output on/off. In the case of the relay it will drop out OK with this type of drive waveform although the timing could end up being variable. \$\endgroup\$
    – Michael Karas
    Commented May 17, 2013 at 20:04
  • 1
    \$\begingroup\$ I think so, although by instead using a small MCU I can also gain some useful capabilities elsewhere so I think I will (as so often) resort to code. Thinking about it, it will probably end up smaller (as it can replace a few other bits of the circuit too) and not much more cost in the end... \$\endgroup\$
    – Jon
    Commented May 17, 2013 at 21:52
  • \$\begingroup\$ @MichaelKaras R2 doesn't seem to be necessary; base current is enough to charge capacitor (for the transistor I'm using, anyway). Does R2 serve any other purpose, or can I just drop it? \$\endgroup\$
    – John H.
    Commented Jul 16 at 1:56
  • 1
    \$\begingroup\$ @JohnH. - R2 ensures that (a) the C1 is fully discharged when the switch is open and (b) ensures that the Q1 base goes fully to GND and not left at or just below the Vbe threshold. \$\endgroup\$
    – Michael Karas
    Commented Jul 16 at 2:33
2
\$\begingroup\$

It appears there is no simple ciruit to do this, it requires a timer or a complex(ish) circuit that's larger than a timer chip. In the end I decided to use a small MCU as it was worthwhile overall, as it gives me capabilities I can use elsewhere to replace other bits of circuit (not shown).

\$\endgroup\$
1
\$\begingroup\$

How about a discrete monoflop like this:

schematic

simulate this circuit – Schematic created using CircuitLab

Not entirely sure about the sizing of the components (especially R4 and C1) as these are very dependent on the power supply voltage.

  • C2 ensures that pressing the button too long does eventually release the load;
  • Either R3 or R1 can be replaced with an relay (in parallel with a flyback diode, cathode at Vcc);
  • R4×C1 defines the delay time;
  • Either transtor Q1 conducts or Q2;
  • Q1 is conducting while in rest.
\$\endgroup\$

Not the answer you're looking for? Browse other questions tagged or ask your own question.