voltage clamp

[elMickotanko]

Hi,

My uni project is a line following robot. Thanks for all your help on the motor circuit before! I've got a problem with the line detection. Ive attached a jpg of my circuit as is which works well. Its about 0.5-0.8v on the black line and 3.6-3.7v on white (lm324 saturation). the problem is i need to limit this to 3.3v to feed into the ADC of the microcontroller. ive tried working the sensor off 3v3 but its no good. I got 3v3 zener diodes aswell but not sure how to use them. Ive googled it but to no avail.
Is there an easy way to do this maybe using diodes in the feedback or the zener somehow?

Thanks,

Mick.

 

[ericgibbs]

hi,
If you are confident that the opa follower output never exceeds +3.7v
then you can put your diode [not the zener] is series with your output
and connect a 1K0 resistor from the output of the diode to ground.
The voltage at the junction of the diode and resistor will be limited to
approx +3volts.

The anode of the diode goes to the output of the op amp
Refer attachment

Regards
EricG

If its still too high, connect another diode in series

 

[Roff]

Here's a precision limiter. If your A/D max input is equal to the VCC on the microcontroller, you can use that voltage as the reference input (V3) to the active clamp.
The waveforms speak for themselves.

 

[Nigel Goodwin]

If you ADC has inbuilt protection diodes (as a PIC does) all you need is a simple current limiting resistor - if not?, then just use two resistors as an attenuator!.

 

[ericgibbs]

hi elmick...

Just to add to your personal database.
As stated you don't need in this instance a semiconductor solution, just resistors.

If in future if you have the same type of problem, where you must use a zener. I'll use your present problem as an example.

You want to drop 0.4V and clamp with a 3V3Z, if your zener requires say 4mA
to operate correctly you would use ohms law to calculate a series resistor
Rs= 0.4/ 0.004 = 100R

Connect one end of the resistor to the output of your opa and the other end to the anode of the zener and the cathode of the zener to ground.
The anode is the 'bar' and the cathode is the 'triangle' of the zener.

The 'clamped' voltage is at the junction of the resistor and zener.

You can apply this method to most zeners, just look at the zener spec, see how much current it requires, subtract the zener voltage from the supply voltage and apply ohms law to find the resistance value.

Regards
EricG

 

[elMickotanko]

Thanks for all your replies.

I tried Erics method first, of the diode in series with output. It works. output is reduced by diode drop and never goes above 3v. tried it with the adc and it works well. Thanks!

The anode is the 'bar' and the cathode is the 'triangle' of the zener.

Is it not the other way round? Thanks for the zener example anyway i'll remember it.

If you ADC has inbuilt protection diodes (as a PIC does) all you need is a simple current limiting resistor

This would be good but can you explain how it works? im not sure how to analyse it? I've attached the input circuit for the ADC showing protection diodes (its an HCS08 im using, our project is funded by Freescale so weve got to use their stuff!). Is there anything i should watch using erics circuit? im not sure how the protection works. Here is some of the parameters:

Input voltage VADIN — VSS (min) VDD (max) V

Input capacitance CADIN — 4.5 (typ) 5.5 pF (max)

Input resistance RADIN — 5 (typ) 7 (max) kΩ

Analog source resistance RAS — 10(max)kΩ

Here's a precision limiter...

Thanks for doing that aswell. I might build one anyway as a backup/learning exersize.

 

[ericgibbs]

Mick,
woops, you are correct re bar is cathode, sorry about that.
I must remember not to stand too close to the microwave!.
Eric

 

[Roff]

Low voltage zeners have a very soft knee. If you were trying to accurately digitize data, and tried to use this type of clamp, the results would be very nonlinear. This is almost certainly not a problem in our OP's application, but it's something to be aware of.

 

[ericgibbs]

hi Ron,
You are right about the knee voltage and I am sure you know about the temp/drift curves +/-. Often in the past to get a 'steady' almost flat temp tracking say, 3.3V I've had to resort to a high quality 6.20V zener, resistor divide into stable opa follower.

Regards
Eric

 

[elMickotanko]

oops i forgot to attach the file with my last qoute

 

[Nigel Goodwin]

The protection is simple, Ras limits the current entering the input pin, and the top diode shunts it to the positive supply if it's more than 0.7V higher than the positive supply. The bottom diode does the same if the input goes negative.