2
\$\begingroup\$

For my internship, I was asked to design an LED driver circuit with <100 ns pulse width and <20 Hz frequency for driving an IR LED. It will be used for testing some photodiodes. Therefore, I am trying to find a pulse generator circuit with very low duty cycle however I could not find any. There are many pulse generator circuits with ns pulse width but all of them has very high frequency. The closest thing I found is a retriggerable monostable multivibrator IC and it looks a little complex(http://www.ti.com/lit/ds/symlink/sn54ls123.pdf). How can I find examples of such circuits? What should I look for?

\$\endgroup\$
  • 1
    \$\begingroup\$ Here's an idea to get you started: linear.com/solutions/1324. One approach, like this one, is to use an arbitrary length pulse, such as one generated by a microcontroller having flexible and even adjustable inter-pulse timing, to trigger a one-shot short-pulse generator. \$\endgroup\$ – scanny Aug 9 '16 at 8:14
  • \$\begingroup\$ ha, @scanny, only an order of magnitude better than necessary! \$\endgroup\$ – Daniel Aug 9 '16 at 8:52
  • \$\begingroup\$ @scanny That looks like a good answwer to me. Why not post it as such? \$\endgroup\$ – JRE Aug 9 '16 at 10:27
  • \$\begingroup\$ Reminds me of laser jamming/blinding circuits used to defeat police speed measurements... \$\endgroup\$ – JimmyB Aug 9 '16 at 20:41
2
\$\begingroup\$

When you're a programmer, every problem can be solved with a microcontroller...

Perhaps overkill, but an ATtiny4/5/9/10 (SOT23-6) running at 12 MHz can be programmed to pulse any of its three (four) I/O pins for as low as 83.3 ns at any time interval from 166 ns to over 5 seconds.

\$\endgroup\$
  • \$\begingroup\$ Some ATtiny's even have an internal 64MHz PLL enabling them of 32MHz "PWM" output (16ns pulse)... \$\endgroup\$ – JimmyB Aug 9 '16 at 20:51
1
\$\begingroup\$

There is a possible solution using delay lines. 10-TAP, TTL-INTERFACED FIXED DELAY LINE from data delay devices, inc http://www.datadelay.com/datasheets/ddu7c.pdf The part DDU7C-100 has 10 taps with a delay of 10 ns per tap. You connect a square wave signal with < 20 Hz to the input of the delay line. To get a pulse with a length of 20 ns, you connect two taps with a delay of 20 ns between them to the inputs of an a AND gate, the first tap directly, the second tap inverted by a fast inverter. You can vary the pulse length by using different taps.

There are also programmable pulse generators: http://www.datadelay.com/datasheets/3d7608.pdf. A pulse length between 10 ns and 100 ns may be selected with a resolution of 0.25, 0.5 or 1 ns. The pulse width is programmed using a parallel interface with 8 or 12 bit.

\$\endgroup\$
0
\$\begingroup\$

The SN74LS123 is very simple to use, but it is to slow for your problem. You want pulses from 10 ns to 100 ns, but the typical minimum pulse width of a SN74LS123 is 116 ns. As far as I know, all other TTL multivibrators are to slow (minimum puls width) for your problem. Sorry, but I can't recommend another monostable multivibrator for your problem.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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