Voltage limits

up

put

Sounds reasonable, given your loose limits.

You could set the top limit as the supply rail of an op amp with rail- to-rail swing--that's pretty stable. Run the op amp on +5v, 'n Bob's yer uncle.

-- Cheers, James Arthur

That will slurp a lot of -15 current when it clamps low.

How about this?

GND

Do you mind the opamp winding up? Some opamps do strange things when they rail.

You can do a feedback clamp, depending on what the opamp input circuit looks like.

r up

=92m

utput

Ahh a R-R I didn't think about that. I've not used them much. Got a favorite R-R opamp? I'd like to keep the +/-15 volt supply rails... which seems to limit my choices.

There's the LM7332.

George H.

What kind of load on the OpAmp which must stop at the limits? Configuration... feedback, source, etc? ...Jim Thompson

up

ut

I could move the diode further down the resistor chain...

=A0R1

Hmm, use the current limit of the opamp to limit the voltage? (talk about slurping current...) It looks simple but does it work? (I'll have to play a bit.)

No I tried it already and it works fine... there's a bit of 'crud' when the opamp leaves the rail, but the clamp is so that the input down stream doesn't get over driven. It's all slow stuff, freq max of maybe 10-100kHz.

The opamp is a unity gain inverter. I'm adding two signals. A DC level and then an AC signal from a signal generator. I add them in an inverter and then flip them over to get the polarities=92 correct. I looked at some feedback type clamp circuits, but didn=92t see a way to get the +1V limit without more opamps.

(There was this old bounding and clamping EDN article by B. Pease)

George H.

to

=A0 =A0 ...Jim Thompson

=A0 =A0| =A0 =A0mens =A0 =A0 |

=A0 | =A0 =A0 et =A0 =A0 =A0|

=A0|

=A0 =A0 =A0 |

ide quoted text -

Hi Jim, I'm summing two signals an adjustable DC level (I can change the range to suit the circiut) and an AC signal from a signal generator. These get added and feed to the VCO input of a 4046. (operted from a 6 volt supply). At the moment the input is two opamps (both inverting) the first sums the signals and the second flips 'em back to positive. I can load the second opamp down as much as I want. (I don't think the VCO takes much current.. but I'd better check.)

Max freq. on the signal generator will be less than 100kHz.

George H.

up

ut

=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0|

Hi John, What's the purpose of R2 in the above circuit.

Playing with numbers, If my opamp has a +/-25mA current limit, then this would work with R1=3D400 ohms, R2=3D0 and R3=3D100 ohms. Was that you= r intent?

George H.

George Herold a écrit :

Depending on the performances you target, have a look at clamping amplifiers: AD8036/8037 for example.

or up

I m

utput

at your

Interesting! Thanks Fred, I'd never heard of those. Do they sell anything slower (and cheaper!)

I'll go see at AD.

George H.

or up

I=92m

output

I actually meant to run the R-R buffer on +5v, but you've come up with an interesting variation.

Here's a version using an R-to-R op amp on +5v:

--|+\ |

Here you go...

I didn't show your summing stage in front. Tweak resistor dividers to tailor break-points. Depending on your OpAmp, you may experience the need to add some compensation ;-)

If you had a quad OpAmp, there's a very nice way to do it with two diodes and a few resistors. It also has no feedback consequences... it's not in a feedback loop. ...Jim Thompson

I stuck a few opamps in and they all clipped at same level. (with the same load and supply voltage.) 13.8 volts.

So something like this,

|\ | >-330-+----1k--+---out |/ D | ^ 1k | | +2.6V---+ GND

Is using the max. output voltage of the opamp in this way a bad idea? (There will be some variation in the supply voltage, but not enough to mess things up.)

The 2.6 volts will come from a LM317. (I?m not sure the 2.6 is the right number...)

Thanks, George H.

No, the opamp rails at +-14 or whatever, and the resistor network just scales that.

Variations on these, maybe...

ftp://jjlarkin.lmi.net/Clamp2.jpg

ftp://jjlarkin.lmi.net/Clamp.JPG

John

ts or up

at

o I=92m

ut

he output

a?

to

's

Wow, I like it! I was about to ask about D1, but that protects the npn from being zenered by the opamp.

Thanks James.

Right, and the transistor b-c junction clamps the output to +5.6v max.

Really, the simplest would be using John's 3-resistor network on the output of a R-to-R amp.

-- Cheers, James Arthur

Here you go, one more time...

Thinking thru the OpAmp implementation I realized I'd tried too hard, and made the first version unnecessarily complex.

Also this latest version doesn't rely on feedback. ...Jim Thompson

tput

put

dea?

gh to

il-

Wow, "my cup runneth over"! Thanks Jim.

It's going to take me a bit to puzzle out how it works. (But don't tell me!)

Yeah I've got a version with diode's, 2 more opamps and resistors.

I was hoping to make it a bit smaller.

George H.

=A0 =A0 ...Jim Thompson

=A0 =A0| =A0 =A0mens =A0 =A0 |

=A0 | =A0 =A0 et =A0 =A0 =A0|

=A0|

=A0 =A0 =A0 |

ide quoted text -

Check out my latest post, you'll like it ;-) ...Jim Thompson

or up

tput

=A0 =A0R1

Do you need those 2 opamps? Just capacitor coupling the ac one then adding the 2 via resistors would do in some cases.

NT