Creating a pulse

[jwcornell]

I'm working on a device which uses a latching solenoid. The solenoid requires a 10 millisecond pulse of five to ten volts at 5 watts to operate the solenoid (the polarity of the pulse determines the direction of operation)

The trick here is it's to be used in a remote location operated by a battery so low power consumption is critical.

I will trigger the pulse event by way of a reed switch attached to a vacuum gauge. Actually, there will be two reed switches with one to close the circuit and operate the switching solenoid in one direction and the other to close another circuit to operate the solenoid in the other direction. The reed switches will be placed on the gauge at the points where the user wishes to open or close the solenoid.

Thoughts?

 

[Edd]

5 watts? If it's rated as 5W @ 5V, then it will draw 1A when supplied 5V, or 2A when supplied 10V, by ohm's law. That's not low power at all.

5V power supply + 555 timer + 1A NPN transistor on output

I was going to suggest a low power CMOS 555, but if you are drawing so much current from this supply for your solenoid, it's probably a miniscule proportion of the total current drawn either way. Still, I spose 1A of current for 10ms isn't too bad if you aren't triggering it frequently. Then you could greatly reduce reduce the standby current draw to less than 1mA.

 

[jwcornell]

Edd

Thanks for your thoughts. You are correct that 1 or 2 amps could not properly be considered low power but given the short timeframe needed of only 10 milliseconds, I thought I might get away with using that term.

Anyway, your thoughts are much appreciated. I can set the timing of the RC circuit on the low power version of the 555 (I'm thinking maybe a 556 instead as I need to trigger the event and will use the first timer on the 556 running astable to trigger the second timer which will be monostable) to allow for only two or three triggering events under typical conditions as the other challenge is to anticipate how long the reed switch will remain closed under typical circumstances. In this case, the vacuum gauge will not likely move quickly (perhaps a few psi per hour) so a long RC time is needed. Since the reed switch will also serve to close the circuit to supply power to the entire device, most of the time the power would be off given your suggestion.

Thanks for your thoughts on the NPN transistor. I think that's a much better solution than a relay which I was considering.

Any other ideas?

 

[Hero999]

What about using a power MOSFET?

Also, as the pulse is for such as short duration, a large capacitor could be connected across the supply, which will eliminate the need for it to be rated to 1A continious.

 

[colin55]

What about using a 74c14. It will take 1microamp and use less components.

 

[jwcornell]

Wow you guys are all great! Thanks Hero999 and colin55 for the ideas. I will do a bit of research and get back to you after I've had a chance to evaluate these suggestions.

 

[jwcornell]

Gang,

I did a bit of research and liked the idea of using the 74c14 as it requires virtually no power and appears to be able to be configured to fire only once regardless of how long the switch is closed. I ran across this schematic from **broken link removed** Does this configuration appear to be a working solution?

Please excuse me if the image below does not appear correctly as it's my first time trying this. hopefully the entire page does not show up but if it does, the image I am referring to is the ninth one down from the top which is captioned "To produce a pulse of constant length, (no matter how long the button is pressed), the following circuit is needed:" just above the schematic.

**broken link removed**

 

[jbeng]

I'm definitely a fan of the 74hc14 and have used it in several astable/monostable circuits lately...