How to limit the voltage of a signal pulse going into an IC?

ElectricAye

Hi all,

I have an "ultra fast" comparator AD8561 that is powered at +5 volts and therefore its absolute maximum input signal voltage is +5 volts (at least that's how I understand the data sheet). Problem is, the comparator will be looking at a train of pulses that sometimes exceeds +5 volts. I don't want these >+5 volt signals to damage the AD8561, so how can I limit inputs to the comparator without dragging down the magnitude of my other (< +5 volt) signals? I'm afraid of using a voltage divider because that will not only reduce my other signals but it might also introduce other problems, capacitance, impedance mismatches, ringing, etc. All the signals will be pulses that are about 1 microsecond wide and I don't want anything that might stretch them out or deform them too much.

Suggestions or comments are greatly appreciated.

thanks,
Mark

SRLM

What about a resistor on the line, and a 5 volt zener diode on the signal input side of the resistor? If the signal is less than 5 volts, then it simply goes to the chip via the resistor, and if it's greater than 5 volts the zener shorts it to ground. I've never tried this, but it seems somewhat logical. Just try it on something other than your chip first...

Phil Pilgrim (PhiPi)

Why not power the comparator from a +10V supply and apply voltage division or clipping to its outputs? This would give you an input common mode range from 0V to 7.8V and eliminate any disruption of the input signals' input impedance. (BTW, with a +5V supply, the usable input common mode range is 0V to 2.8V.)

-Phil

Carl Hayes

SRLM's idea, resistor-plus-zener, wuld work. Or several ordinary diodes in series to ground (using the sum of their forward drops the same way you'd use a zener). Another way would be to buffer the pulses with some device that wouldn't be bothered by voltage overshoots.

If your comparator can actually stand a diode-drop MORE than the supply rails, then a diode to the positive rail might serve to limit input pulses acceptably. One might choose a very-low-forward-drop diode in that case.

But I wonder why you're using a comparator (an analog device) rather than a gate or other digital device.· You express an interest (as I understand it) in maintaining the pulse-length, and gates are designed to do that, having much faster slew rates at the output.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
· -- Carl, nn5i@arrl.net

Post Edited (Carl Hayes) : 2/3/2009 12:03:43 PM GMT

ElectricAye

Phil Pilgrim (PhiPi) said...
Why not power the comparator from a +10V supply and apply voltage division or clipping to its outputs?

This thought crossed my mind but I was hoping to avoid adding another power supply.

SRLM said...
What about a resistor on the line, and a 5 volt zener diode on the signal input side of the resistor?

SRLM, okay, that seems reasonable to me. But what is this kind of design called? Is there a name for it? I'd like to read more about it but I don't know what to search for. (Warning: I'm a MechE, you know. )

Carl Hayes said...
Another way would be to buffer the pulses with some device that wouldn't be bothered by voltage overshoots.

Carl, yes, I thought about using some kind of amp at Gain = 1, but I don't know how to start searching for "parts that can take my bad design techniques." And I wonder if any parts can take input signal voltages above the power supply???

Thanks, you guys, for the inputs. You are the lanterns of hope in my cavern of darkness.

ElectricAye

Carl Hayes said...

But I wonder why you're using a comparator (an analog device) rather than a gate or other digital device. You express an interest (as I understand it) in maintaining the pulse-length, and gates are designed to do that, having much faster slew rates at the output.

Hmmm.... well, I just used what I've seen other people use on similar applications. That doesn't mean I know what I'm doing. The pulse coming into the comparator is from an analog device (a photomultiplier via an amp x10) and it uses a fast latch. If you know of something that might be better and that comes in a DIP, please do tell.

thanks,
Mark

Bean

Carl Hayes said...
If your comparator can actually stand a diode-drop MORE than the supply rails, then a diode to the positive rail might serve to limit input pulses acceptably. One might choose a very-low-forward-drop diode in that case.

You have to be careful doing this...I've seen this actually raise the supply rail if the supply·is not very low impedance.

Bean.


▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
·The next time you need a hero don't look up in the sky...Look in the mirror.




Post Edited (Bean (Hitt Consulting)) : 2/3/2009 5:51:01 PM GMT

SRLM

Sorry, I don't know what the zener-resistor setup is called or if it even has a name, I was simply thinking of electronic building blocks and how to put them together.

Edit: amazingly, I think I choose the right words to search for: Zener Limiter. Here are some of the results:

http://www.eng-tips.com/viewthread.cfm?qid=99268
http://hyperphysics.phy-astr.gsu.edu/HBASE/Electronic/limiter.html

Carl Hayes

SRLM said...
Sorry, I don't know what the zener-resistor setup is called or if it even has a name, I was simply thinking of electronic building blocks and how to put them together.

Edit: amazingly, I think I choose the right words to search for: Zener Limiter. Here are some of the results:

http://www.eng-tips.com/viewthread.cfm?qid=99268
http://hyperphysics.phy-astr.gsu.edu/HBASE/Electronic/limiter.html

It's often called a clipper.· Or clamp.· Or limiter.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
· -- Carl, nn5i@arrl.net

Carl Hayes

Bean (Hitt Consulting) said...
You have to be careful doing this...I've seen this actually raise the supply rail if the supply·is not very low impedance.

Bean.

Ah, true.· But we strive to make our supplies have low impedance.· That's what·capacitors are for.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
· -- Carl, nn5i@arrl.net

Carl Hayes

ElectricAye said...

Carl Hayes said...

But I wonder why you're using a comparator (an analog device) rather than a gate or other digital device. You express an interest (as I understand it) in maintaining the pulse-length, and gates are designed to do that, having much faster slew rates at the output.

Hmmm.... well, I just used what I've seen other people use on similar applications. That doesn't mean I know what I'm doing. The pulse coming into the comparator is from an analog device (a photomultiplier via an amp x10) and it uses a fast latch. If you know of something that might be better and that comes in a DIP, please do tell.

thanks,
Mark

Gosh, Mark, I think I'd attack that in that "amp x10".· It may be as simple as adding a zener or a diode from its output to its negative input (if it's an op-amp) to limit its output.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
· -- Carl, nn5i@arrl.net

ElectricAye

Carl Hayes said...


Gosh, Mark, I think I'd attack that in that "amp x10". It may be as simple as adding a zener or a diode from its output to its negative input (if it's an op-amp) to limit its output.

Hi Carl. That sounds cool but unfortunately the amp is already built into the photomultiplier's module and I can't get to it.

Right now it looks like I'm going to try a combo of PhiPi's method of cranking up the supply voltage on the comparator and SRLM's zener clamp idea. I think that will give me a margin of safety while at the same time not causing me to lose too much range.

Thanks everybody!

ElectricAye

Phil Pilgrim (PhiPi) said...
(BTW, with a +5V supply, the usable input common mode range is 0V to 2.8V.)

Thanks, Phil, until you pointed that out to me, I never noticed that before. I just blithely assumed I could pump 5 volt signals into this thing and not pay a price.


cheers,
Mark