What is the best IC for programmable pulse width control with a wide range?

[Jdo300]

Hello All,

I'm looking for a nifty chip that I can use to pro grammatically set the pulse width of a square wave signal in my circuit. The pulse width range I'm looking for is anywhere between 10 ns and 500 ns and so far, the only thing I have found is the DS1014 series ICs. They have the range that I need but they only have 5 discrete pulse widths in each chip. I was wondering if anyone here may know of another IC that has a similar range but with more settable pulse widths?

Thank you,
Jason O

 

[chroot]

10 ns is going to be rather tough for older chips like the 555.

What kind of resolution do you need on the pulse width?

- Warren

 

[Jdo300]

I'm pretty flexible in that matter. Since I would be programming that in. But just as a rough guess, maybe 5 ns steps?

- Jason O

 

[Jdo300]

I tried the 555 and the best I could get out of that was about 5 uS. Also, it will puke out on me if I try to drive it over 1 MHz.

- Jason O

 

[chroot]

5 ns steps... so you want many hundreds of distinct steps in there. Probably something like a 10-bit code?

I'd probably use a serial DAC to produce a aribtrary DC voltage, controllable by your computer. I'd use this voltage as a reference for a comparator.

I would use a current source to charge a capacitor, and, when the voltage of the capacitor exceeds the voltage set by the DAC, the comparator would trip.

The output of the comparator would then go to the reset input of an SR latch. The output of the latch is the pulse you're trying to generate.

So, in total: to trigger the thing, you apply a pulse on the set pin of the latch. The latch output goes high, enabling a current source which charges a cap. When the cap reaches the voltage you've set by the DAC, the latch resets, and the pulse stops.

You could get theoretically any resolution or range you could possibly want with this, and it'd only take perhaps five components.

- Warren

 

[Jdo300]

Hmmmm that sounds doable. Thanks for the help . OK I have another unrelated question but I'll make a new thread for it.

- Jason O

 

[Mindscrape]

There are these "single shot" multivibrators that national sells, but it sounds like the same thing that chroot described, though the comparator stuff is all in one.

 

[Jdo300]

That sounds even better. For my application, I'm really cramped for PCB space so something like that chip would be nice. I looked it up and I think it is the DM54123 IC.

If I were to go with this, would I just pick a capacitor and then use the DAC in place of the resistor? could someone give me a suggested wiring diagram for this?

Thanks,
- Jason O

 

[Jdo300]

Oh by the way, maybe I could use a digital resistor in place of the DAC to drive this?

 

[chroot]

If you look at Figure 3 in the datasheet, they show a network using a potentiometer. The potentiometer is serving the same purpose as a DAC: to provide an adjustable voltage.

I don't know much about "digital resistors," but I assume you could them. You can also use digital potentiometers, like those made by Intersil.

- Warren

 

[Jdo300]

Hi Warren,

Thanks again for the tip. I'll definitely look into this method .

- Jason O

 

[Jdo300]

Hello All,

I found the ULTIMATE solution for the programmable pulse width control. Check this out:

http://pdf1.alldatasheet.com/datasheet-pdf/view/165446/DATADELAY/3D7622.html

The 3D7622 is a 22-bit programmable pulse width generator which has a range of 10ns to 1.05 ms or greater!

- Jason O