I'm building an old school distributor machine for testing vintage motorcycle ignition systems. I want to flash LEDs to indicate when the ignition pulse is happening and would like to use 12V to trigger them to avoid needing the high voltage coil and spark plug. One hurdle I'm running into is the dwell time of the ignition pulse is far too long. The LEDs stay on too long to see any timing marks.
I've been searching for edge trigger circuit designs that will work for me but every one I've in come across the minimum on time is at least as long as the trigger pulse. I can make a microcontroller see a rising or falling edge and flash the LEDs for a few microseconds but it seems to go erratic at the upper end of the RPM range.
Does anyone know of an edge trigger circuit that can have an output that's shorter than the input pulse?
I've been searching for edge trigger circuit designs that will work for me but every one I've in come across the minimum on time is at least as long as the trigger pulse. I can make a microcontroller see a rising or falling edge and flash the LEDs for a few microseconds but it seems to go erratic at the upper end of the RPM range.
Does anyone know of an edge trigger circuit that can have an output that's shorter than the input pulse?
Do you mean something like an edge-triggered monostable multivibrator, a 4538 for example?
https://www.mouser.com/datasheet/2/149/cd4538bc-294672.pdf
https://www.mouser.com/datasheet/2/149/cd4538bc-294672.pdf
It can. The output pulse width just depends on the RC time constant and not on the width of the input pulse, as long as the input pulse is wide enough to trigger the CD4538 - which means wider than about 25 ns to 70 ns (depending on supply voltage and on whether you are interested in typical or worst-case numbers).
Thanks. That looks like it'll work with the voltage I'm using without any problem too. It looks like the CD4538 is obsolete. I'm looking at CD14538 now.
Input pulses will be in mS at low speed. The ignition system operate at relatively slow speeds compared to the rest of the electronics world. 10000 RPM is only 167 Hz (waste spark sytem) and the ignition modules are all designed to extend the dwell time to allow the coils to charge as much as possible.
Input pulses will be in mS at low speed. The ignition system operate at relatively slow speeds compared to the rest of the electronics world. 10000 RPM is only 167 Hz (waste spark sytem) and the ignition modules are all designed to extend the dwell time to allow the coils to charge as much as possible.
What about a typical 555 timer circuit?
You can adjust rising edge or falling edge pretty easily, and the 555 simply sets whatever pulse width you want.
You can adjust rising edge or falling edge pretty easily, and the 555 simply sets whatever pulse width you want.
Quote from the datasheet "Monostable operation is initiated when TRIG voltage falls below the trigger threshold. Once initiated, the sequence ends only if TRIG is high for at least 10 µs before the end of the timing interval."
You stated, "Input pulses will be in mS at low speed"
When the 'switch' is closed, pin 2 is immediately drawn to 0V since the capacitor sits with 0V across it. This immediately triggers the 555.
Then, the capacitor charges as a single time constant up to 5V through the 100k. You can select the R and C time constant as needed for your application to get the charge back up to 5V.
The 0.1uF capacitor generates a negative pulse. You can hold the switch closed forever, and will only generate one pulse out of the 555. Since you have a slow requirement, this should work just fine.
The circuit has been built and tested and working for years, this is not theoretical.
When the 'switch' is closed, pin 2 is immediately drawn to 0V since the capacitor sits with 0V across it. This immediately triggers the 555.
Then, the capacitor charges as a single time constant up to 5V through the 100k. You can select the R and C time constant as needed for your application to get the charge back up to 5V.
The 0.1uF capacitor generates a negative pulse. You can hold the switch closed forever, and will only generate one pulse out of the 555. Since you have a slow requirement, this should work just fine.
The circuit has been built and tested and working for years, this is not theoretical.
IIRC, you want an edge-triggered/non-retriggerable one-shot. If so, have to look at datasheets to find one. https://www.mouser.com/c/semiconductors/logic-ics/monostable-multivibrator/?logic type=Monostable Multivibrator&instock=y
Maybe one like this: https://www.mouser.com/datasheet/2/609/LTC6993_6993_1_6993_2_6993_3_6993_4-1270601.pdf
Non-retriggerable means it can't be triggered again until after it times out.
Maybe one like this: https://www.mouser.com/datasheet/2/609/LTC6993_6993_1_6993_2_6993_3_6993_4-1270601.pdf
Non-retriggerable means it can't be triggered again until after it times out.
I built this circuit already, that's why I read the datasheet to figure out why it wouldn't work. It will not end it's cycle if the input is still active, just as the datasheet clearly states. The next issue with the 555 is the minimum output pulse time is 10 Ms. I need around 10 uS.
You can always capacitor couple the input so that only the edges will matter. Other things can be done too. First you need to know enough about the inputs and outputs. Input risetime? Input switch bounce? Input maximum frequency? Input voltage range? Maximum closed-switch dwell time? Output pulse width, drive current requirements, etc. Then it should be fairly straightforward to get down to specifics. A digital scope could probably do most of the measurements.
Last edited:
The CD14538 actually looks perfect for my application. I'm using a 12V supply so I should be able to directly connect the input to my ignition module with some protection of course.
This'll work too. You'll get a pulse for each edge. You can adjust the length of the pulse by adjusting the number of inverters. With fast enough gates you can get a pulse width below 1 ns.
Tom
Tom
AND gate and inverting delay line (N>4 inverters for CD4000 or 74LS) to get narrow high going output pulse on each rising edge input
OR gate and inverting delay line to get narrow low going output pulse on each falling edge input
However you will maximize economy (achieving lowest chip count) if you apply deMorgan's Laws, and avoid pure-AND and/or pure-OR gates.
OR gate and inverting delay line to get narrow low going output pulse on each falling edge input
However you will maximize economy (achieving lowest chip count) if you apply deMorgan's Laws, and avoid pure-AND and/or pure-OR gates.