How's this?
ftp://66.117.156.8/Clamp.JPG
I need clean clipping, and can't overdrive some downstream mux's that run off +5 and ground.
John
How's this?
ftp://66.117.156.8/Clamp.JPG
I need clean clipping, and can't overdrive some downstream mux's that run off +5 and ground.
John
Just guessing... but how about a TL431 across the Rf.
+----R1--+--R2-----+ | | | +-------> K--------+Ikmin is 0.4mA. Iref current is 2uA
R1 and R2 become part of the op amp gain. D from BC
Or ...how about a rail to rail output op amp with +V at 5V.
D from BC
Your scheme "works", but how precise do you have to be, and how fast?
Being a preciseness nut I would tend to use a multi-op-amp scheme myself.
...Jim Thompson
-- | James E.Thompson, P.E. | mens | | Analog Innovations, Inc. | et |
This schematics is no better then the classic diode clamping to +/- rails.
VLV
The clamp voltage, and indeed the signal, needn't be super precise. This is for a tachometer front-end, max bandwidth 100 KHz maybe. But we can have signals with a huge voltage range and need to clip clean. Downstream, it's convenient to use cheap mux's/amps/dacs/comparators that run 0 to +5, and some of them will freak if the signal gets past the rails.
The thing I did seems to be solid, and has low parts count. Of course, the common-base stage has high gain, so I'll have to make sure it's stable inside the opamp loop.
Oops, BAV99 is a series pair; I'll have to use the common-cathode version here.
John
Maybe so, but we'd have to be sure it didn't wind up in saturation, and take a long time to recover. Some opamps can take milliseconds, or even seconds, to come out of overload. A closed-loop clamp avoids that particular hazard.
John
Classic diodes where? To which rails?
Schematic?
John
This is probably better: more tunable, and no added gain inside the loop.
ftp://66.117.156.8/Clamp2.JPG
John
100KHz, and what is the loading? Where is the OA output going with the unclamped signal, what kind of impedance, R +C? And what kind of gain did you want out of that circuit?
Downstream, the load is pretty light, a bit more signal conditioning, some switching, and some comparators. But that stuff doesn't like being driven beyond the rails. If the customer grossly overloads it, like connects 120 volts RMS, we'd still like it to clip cleanly.
Gain=1 is about where it will run. Rf and Ri will be ballpark 50K, so we can't stand too much capacitance in the clamp circuit.
John
how about one of these ?
Jure Z.
I've used that one, and the original HFA1130, which is somewhat faster. But it's overkill for my application, too fast, high bias currents, high power, and expensive. It needs Rf values in the hundreds of ohms, and I need 50K or so.
It is handy in other places, like driving fast adc's.
John
Investigate some CMOS opamps. They tend not to "stick" by design.
That would work for the positive clamp level, but not zero volt clamp level; one would get about -600mV on the low end and violate the given conditions.
At least if the clamping is done as shown, the output can remain within the stated criteria.
I would suggest a silicon to -12V and a schottky to the NI of the op-amp; that would guarantee a positive and non-zero "low" level.
It's ikky but how about..
+----R1--+--R2-----+ | | | +------+-------> K--------+ | TL431 | | | Vin---R--|
If memory serves me correct, there are opamps with built-in clamps. The min and max voltages are applied to external pins.
-- Reply to nico@nctdevpuntnl (punt=.) Bedrijven en winkels vindt U op www.adresboekje.nl
On a sunny day (Mon, 24 Dec 2007 19:07:02 -0800) it happened John Larkin wrote in :
You need to limit BEFORE the gain stage to prevent opamp overdrive:
in ----------- R ------------------------- R --- CMOS opamp on 5 V supply
----------- |k \\ \\ Zener 5.1V +5.1 to -.7 V
+5 to 0 V |a ///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.