555 one shot delay next trigger

Thread Starter

bobgroger

Joined May 9, 2017
6
First let me say I am NOT an engineer, I have tinkered with mostly analog electronics for years but I am starting to get my head around simple digital projects. Here is my scenario: I have a 555 one shot outputting a 100ms pulse when pin 2 is brought low. This works fine. What I need is a way to delay or lock out the next trigger for a period of time, around one second. Would another 555 one shot with a pulse time of ~1 second driving pin 2 of the second 555 through an inverter produce what I am looking for? Just looking for affirmation this may work or another path to follow, thanks in advance
Bob G.
 

dl324

Joined Mar 30, 2015
13,092
Welcome to AAC!

It would be simpler if you just used a non-retriggerable one shot like SN74121.

EDIT: Just reread your post and noticed that the one shot is 100mS. It would require two independent one shots. Do you realize that the trigger needs to return to a voltage above 1/3 VCC before it can timeout?

It would be easier to read your posts if you used paragraphs and we always appreciate schematics (of the well drawn variety).
 
Last edited:

AnalogKid

Joined Aug 1, 2013
9,353
Yes, a 2nd 555 can be used to inhibit the first one.

There are some logic relationships you need to address. For example, what do you want to happen if the input signal is in the trigger state when the 1 second inhibit ends - trigger the first 555 immediately, or wait until the input goes to the non-trigger state, then back to the trigger state so the output leading edge is timed to the trigger input edge?.

As mentioned above, the 555 is not a true monostable (no negative feedback). What is the pulse width of the input signal?

ak
 

Thread Starter

bobgroger

Joined May 9, 2017
6
Yes, a 2nd 555 can be used to inhibit the first one.

There are some logic relationships you need to address. For example, what do you want to happen if the input signal is in the trigger state when the 1 second inhibit ends - trigger the first 555 immediately, or wait until the input goes to the non-trigger state, then back to the trigger state so the output leading edge is timed to the trigger input edge?.

As mentioned above, the 555 is not a true monostable (no negative feedback). What is the pulse width of the input signal?

ak
The input signal is a manual switch. I am having problems with contact bounce or triggering when the switch is opened if it is held down more than 100ms. After the first trigger it will be at least one second until the next trigger is needed, I want that to happen when requested but not less than one second from the previous trigger

I need to improve the reliability, this is for an airshifter and engine kill for a drag bike. If it triggers at the wrong time you lose.

What do you recommend for drawing circuits other than a scanned drawing? Current circuit is a textbook 555 oneshot. I have tried to debounce the switch with a .01 cap between the switch and pin 2 as in attachment 1 but still have false triggering problems. What I propose is as in the second drawing, a 1 second MMV in front of a 100ms MMV. I have ordered some parts to breadboard this but does it seem reasonable as shown?
 

Attachments

dl324

Joined Mar 30, 2015
13,092
Images from the PDFs for the convenience of others:
upload_2017-6-9_15-2-24.png
upload_2017-6-9_15-2-52.png
I have tried to debounce the switch with a .01 cap between the switch and pin 2 as in attachment 1 but still have false triggering problems.
Try putting the cap in parallel with the switch.

I think you want to use the approach suggested by @Bernard if the lockout can be after the 100mS time period. In that case, you'd use the inverted output of the 1 second one shot to force reset LOW on the 100mS one shot. That would cause it to ignore any triggers during that period.

What you have won't work. You can't direct couple to the trigger of the 1mS OS and expect it to work.

You also need a snubber diode across the coil of whatever the 100mS OS is driving. What is the coil resistance and voltage?
 

dl324

Joined Mar 30, 2015
13,092
What do you recommend for drawing circuits other than a scanned drawing?
I use the free version of Eagle. Some say it's too difficult to learn; I didn't find that to be the case.

Here's something for you to consider:
upload_2017-6-9_16-27-55.png

The first OS triggers the second when the 100mS time period times out. The second OS will hold the reset of the first OS low, preventing it from triggering.

The output of the first OS drives Q2 which energizes the solenoid. If solenoid current is higher than 100mA, use a different transistor; or a MOSFET.
 

Thread Starter

bobgroger

Joined May 9, 2017
6
Images from the PDFs for the convenience of others:
View attachment 128587
View attachment 128588
Try putting the cap in parallel with the switch.

I think you want to use the approach suggested by @Bernard if the lockout can be after the 100mS time period. In that case, you'd use the inverted output of the 1 second one shot to force reset LOW on the 100mS one shot. That would cause it to ignore any triggers during that period.

What you have won't work. You can't direct couple to the trigger of the 1mS OS and expect it to work.

You also need a snubber diode across the coil of whatever the 100mS OS is driving. What is the coil resistance and voltage?
OK, I will give that a try. It is driving a 12v relay, it has a built in snubber diode. I will update later when I work through using the reset pin. I have to figure out how to trigger the 1 second OS after or with the falling edge of the 100ms one, correct?
 

AnalogKid

Joined Aug 1, 2013
9,353
Does the circuit drive a relay as shown? If so, what are the coil requirements?

And - are you stuck with a SPST switch? A SPDT switch gives the best combination of debouncing and recovery time, since it is perfect debouncing and there is no recovery time. The drawback is that the switch has to go through a 2-gate flipflop, but the output signal is perfectly clean.

Back to my original question, If someone has pushed the button a few milliseconds before the 1 second inhibit time has ended, or maybe pushed it at the 1/2 second point and still is holding it down, do you want the output pulse to be triggered, or to wait until the switch is released and then pushed again?

Depending on the answer, I think I can get flipflop debouncing, a 100 ms output pulse, a 1 s inhibit timer, and the inhibit logic in a single hex inverter, plus one transistor as a relay driver.

ak
 

Thread Starter

bobgroger

Joined May 9, 2017
6
I currently have a SPST switch and it was fairly expensive. I could however find a suitable SPDT if it would be worthwhile since I already have the bracketry and mounting stuff. As far as if the button is down when the delay is up, I think I would prefer to have to press the button again to make it trigger. It does drive a small 12v relay, coil is 380 ohms which if my Ohm's law is correct is 376ma?
 

dl324

Joined Mar 30, 2015
13,092
I forgot to AC couple the switch to prevent a long press from preventing the OS from timing out. I added C7 and R8 to take care of that.
upload_2017-6-9_19-55-13.png

Other than R2, C2, R4, and C5, component values aren't very critical.
 

AnalogKid

Joined Aug 1, 2013
9,353
The circuit I was thinking of in post #11 will work, but a more simple solution is to have the switch trigger the 1 second timer, and have that output's leading edge trigger the 0.1 s output pulse. In this way, the 1 s monostable does the debouncing without a SPDT switch and flipflop. This works only if the timer is a true monostable.

R2-C2 - differentiate the switch input so that if the switch still is depressed at the end of the 1 s inhibit, it is ignored until it is released and pressed again.

U1A-R3-C3-U1B - 1 s inhibit. This is a true monostable; once triggered, its output pulse width is independent of further activity at the input. It ignores switch bounce.

R4-C4-U1C - 0.1 s output pulse, formed from the *leading* edge of the inhibit signal. This is a "pulse-stretcher" or "edge-stretcher", sometimes called a boxcar circuit. It is not a true monostable, but its input is controlled by the clean output of the inhibit timer so it is well-behaved in this circuit. Total delay from a button press to the start of the output pulse should be less than 1 us (3 gate delays). This is much shorter than the relay activation time (a few ms).

Q1 - relay coil driver. Use any appropriate small power MOSFET rated for at least 25 V and 1 A.

The timing R's and C's are shown as 5%, but 1% will give better accuracy and repeatability over temperature. Also, the CD4093 transition levels are not as tightly controlled as those in a 555, but should give good repeatability in this application.

No matter what you build, it will need transient protection; cars and bikes are nasty electrical environments.

ak
Bike-Shift-Pulse-Inh-1-c.gif
 

Attachments

Last edited:

Thread Starter

bobgroger

Joined May 9, 2017
6
The circuit I was thinking of in post #11 will work, but a more simple solution is to have the switch trigger the 1 second timer, and have that output's leading edge trigger the 0.1 s output pulse. In this way, the 1 s monostable does the debouncing without a SPDT switch and flipflop. This works only if the timer is a true monostable.

R2-C2 - differentiate the switch input so that if the switch still is depressed at the end of the 1 s inhibit, it is ignored until it is released and pressed again.

U1A-R3-C3-U1B - 1 s inhibit. This is a true monostable; once triggered, its output pulse width is independent of further activity at the input. It ignores switch bounce.

R4-C4-U1C - 0.1 s output pulse, formed from the *leading* edge of the inhibit signal. This is a "pulse-stretcher" or "edge-stretcher", sometimes called a boxcar circuit. It is not a true monostable, but its input is controlled by the clean output of the inhibit timer so it is well-behaved in this circuit. Total delay from a button press to the start of the output pulse should be less than 1 us (3 gate delays). This is much shorter than the relay activation time (a few ms).

Q1 - relay coil driver. Use any appropriate small power MOSFET rated for at least 25 V and 1 A.

The timing R's and C's are shown as 5%, but 1% will give better accuracy and repeatability over temperature. Also, the CD4093 transition levels are not as tightly controlled as those in a 555, but should give good repeatability in this application.

No matter what you build, it will need transient protection; cars and bikes are nasty electrical environments.

ak
View attachment 128619
Thank you so much for taking the time to post that, and dl324 as well. Good stuff, and I will try both versions ASAP and report back. I hear you about interference, had to switch to resistor spark plug wires to make ignition work consistently....
 
Top