Input voltage clamp

So I want to clamp the input to a comparator to keep it below ~2V or so. I'm thinking of just a string of diodes (3 or 4) to ground. Current is limited up stream. Is there some better way? Is there any advantage to using an LED?

George H.

Reply to
George Herold
Loading thread data ...

If you could use a stiff enough rail you'd get away with just one diode. Plus another one in reverse to GND if ESD-conscious.

It would visually indicate when it clampeth :-)

--
Regards, Joerg 

http://www.analogconsultants.com/
Reply to
Joerg

If your accuracy requirements aren't too severe, a series clipper is always good. The drawback is that the slight difference in V_f causes an offset.

LEDs are good if they're operating in the dark, but you'll get probably tens to hundreds of nanoamps of photocurrent in room lights and a great deal more in sunlight. They also have fairly high capacitance in general.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

Three series diodes have lower capacitance than one, if that matters.

Cheers

--
Clive
Reply to
Clive Arthur

OK, I guess the advantage there is when not near clamping the diode(s) are reversed biased, and not dribbling in little bits of current. (though that's not a problem here.)

George H.

Reply to
George Herold

Yeah besides the larger capacitance I was a little worried about the speed. I'd like it to turn on/off is a few ns.

George H.

Reply to
George Herold

Well two or three diodes in series was too slow (or something). One diode worked... I'll try a rail. I'm slamming on a ~10-20V and some of that is leaking through. It's hard to tell how much or how fast, (x10 probe loads the node and no problems then. :^)

George H. (gotta run.)

Reply to
George Herold

Should not be slow and as Clive wrote three in series are lower in capacitance than one against a rail. Make sure to use diodes without a PIN characteristic, for example do not use a 1N4007.

How fast do you need it to be? If in the nsec range you have to use Schottky. Those will leak a bit more.

I had the same problem after eating way too many wild blackberries off a bush :-)

--
Regards, Joerg 

http://www.analogconsultants.com/
Reply to
Joerg

This works pretty well. The clamp voltage is easily tweaked.

formatting link

It's still a diode junction - just one - so it's a little soft. A string of diodes would be a lot softer.

--
John Larkin   Highland Technology, Inc   trk 

jlarkin att highlandtechnology dott com 
 Click to see the full signature
Reply to
John Larkin

Now you tell us. ;)

Something like a BAS70 series clipper might be the ticket. They come in CA and CC, which are good for clippers, and the offset should be lower than with separate diodes. (There's no actual guarantee that the dice are from the same wafer, but you can make a metric buttload of them from one wafer, so the odds are excellent I expect.)

They have a PN guard ring, so don't push them too hard or you'll get reverse recovery artifacts from the PN junction.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

Yeah..I mostly just need a blanking pulse to 'turn off' my comparator... as I blast ~20 V... through my 3 pF spad into ~X pF comparator input when I'm resetting it (spad). X is about 20-30 pF, and 500 ohms.

I think this thread has to turn into, "How to do my comparator front end." (Both sides.) I'll post pics and schematics of what I'm doing tomorrow.

74hc14 mickey mouse logic.

George H.

aside; how much current can you pull (load) from a single stage? (74hc14) When I stuck 100 ohms on, the output dropped from 5 to 4V about 20 ohms. Is that right? (I didn't turn it on for very long.) low duty cycle. GH

Reply to
George Herold

Wot's a spad? Something in a space ship warp matrix?

That's about right but there will be high lot-to-lot tolerances. Better follow that with a CMOS bus driver. If on the same chip you could parallel some.

--
Regards, Joerg 

http://www.analogconsultants.com/
Reply to
Joerg

Yikes, that's some seriously obese comparator. Is it on a cable or something?

Cheers

Phil Hobbs

Reply to
pcdhobbs

The spec sheet says Vcc and ground pins can handle up to 50 mA continuous, so (with six sections) that'd be 8 mA on each section if you want to be safe. I've seen more than half an amp (but that was at abnormally low temperature, and only briefly), transient current.

HC can operate from 2 to 6V, useful current output probably no more than 4 mA if you want to keep logic margins with a 5V supply.

Reply to
whit3rd

Hi all, I've been testing this with a dummy capacitor as spad. (Joerg, SPAD* is single photon avalanche diode, an avalanche photo diode biased above the breakdown voltage.) And I had to try the real spad in there... much fussing around.

Anyway, As I said I mostly need to make a blanking pulse for the comparator input. I was raising the inverting input to ~2V but that doesn't seem to be enough. Here's a messy schematic of what I'm doing now.

formatting link

I'm going to try a Schottky diode instead of the 1n4148. And then maybe tie to inverting input to the +5V with the blanking pulse.

Other ideas welcome.

George H.

*or the airplane, Snoopy flies.. your choice.
Reply to
George Herold

Thanks, any guess how long it takes to turn on? (maybe I need a little inductance in my spad signal path to slow things down a little?)

George H.

Reply to
George Herold

OK the comparator is an LT1016, I see no input C spec. But say 5 pf, I've got a toggle switch on the input ~another 5 pF... then some other strays.. and if I stick my 'scope probe on it. :^)

George H.

Reply to
George Herold

George Herold wrote

What is wrong with the good old 74121 one shot as pulse 'stretcher'? Or if you want lower supply (and thus more sensitve input) 74HC123? And it has internal protection diodes AFAIK. Used the 121 many times.

Single chip solution... one half spare for other games.

Reply to
<698839253X6D445TD

With a fast PNP, sub-ns. If you need speed, bypassing the base to ground would help.

With a slow PNP, the transistor capacitance will clamp your signal!

A faster clamp would be a schottky diode or two into a stiff node that can sink current. More work.

--
John Larkin   Highland Technology, Inc   trk 

jlarkin att highlandtechnology dott com 
 Click to see the full signature
Reply to
John Larkin

Sure that would work... I happen to have the HC14's in stock. My problem is how to turn the comparator off. The top half of the spad is getting blasted with ~5-10V step when I reset it and that get's capacitively coupled into the comparator, which causes it to fire again.. Actually I think it's a little more complicated than that.

george H.

Reply to
George Herold

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.