Amplifier for Sensor to AD

Amplifier for Sensor to AD

Postby Tom Maier » Tue May 27, 2014 12:27 am

Here is a schematic for a general purpose non-inverting amplifier with diode input protection. This amp is only for positive input signals of less than 0.5V. For negitive going signals you would need an offset. Maybe I'll post one of those later.

Use a rail-to-rail opamp like the mcp6001. A opamp that is not rail-to-rail will normally top-out at Vdd-0.7 or at Vdd-1.5 Volts, and if you are using a Vdd of only 3.3 V, then the opamp can't span very far.

If you don't want to build your own amp board, microchip sells a demo board for the mcp6xxx series of opamps and you can configure it to how you like.

Anyway... here it is.
Attachments
amplifier_11x.gif
amplifier_11x.gif (94.93 KiB) Viewed 8027 times
User avatar
Tom Maier
Verified identity
 
Posts: 179
Joined: Mon May 26, 2014 2:37 pm
PIC experience: Professional 5+ years with MCHP products

Re: Amplifier for Sensor to AD

Postby Tom Maier » Tue May 27, 2014 12:31 am

Oh, yeah...

What if you had a signal of less than 1 Volt? You could put the protection diodes in series and use two up and one down (if you know what I mean). The diode voltages add in series.

Use fast diodes like the 1N4148. The 1N4004 is very slow and a spike can go whizzing right past it before it decides to turn on. You can use your favorite Schotky if you have one, but make sure the leakage is fairly low.
User avatar
Tom Maier
Verified identity
 
Posts: 179
Joined: Mon May 26, 2014 2:37 pm
PIC experience: Professional 5+ years with MCHP products

Re: Amplifier for Sensor to AD

Postby ric » Tue May 27, 2014 1:09 am

Shrunk a little for those with small screens :)

amplifier_11x.png
amplifier_11x.png (63.62 KiB) Viewed 8023 times
Latest test project, an LED matrix display made from one reel of addressable LEDs. here
User avatar
ric
Verified identity
 
Posts: 659
Joined: Sat May 24, 2014 2:35 pm
Location: Melbourne, Australia
PIC experience: Professional 5+ years with MCHP products

Re: Amplifier for Sensor to AD

Postby SeanD » Fri Jun 20, 2014 1:12 pm

Tom
thanks for this. I am wondering why you are using a +/- power supply rather than a single positive rail and ground? The PIC ADCs I use can't go below 0V (or -0.6V with the diode drop), but there may be other parts which allow negative ADC voltages? I tend to use rail to rail op-amps powered from a single 5V supply which powers both the PIC and the op-amp. attached a typical ADC from the 16F1x series
Sean

ADCimage.png
ADCimage.png (19.64 KiB) Viewed 7946 times
SeanD
 
Posts: 3
Joined: Mon Jun 02, 2014 5:53 pm
PIC experience: Professional 5+ years with MCHP products

Re: Amplifier for Sensor to AD

Postby ric » Fri Jun 20, 2014 1:26 pm

SeanD wrote:...
am wondering why you are using a +/- power supply rather than a single positive rail and ground?
...

He's not. The op-amp negative supply is labelled "V-", but it is connected to ground.
Latest test project, an LED matrix display made from one reel of addressable LEDs. here
User avatar
ric
Verified identity
 
Posts: 659
Joined: Sat May 24, 2014 2:35 pm
Location: Melbourne, Australia
PIC experience: Professional 5+ years with MCHP products

Re: Amplifier for Sensor to AD

Postby Tom Maier » Fri Jun 20, 2014 2:48 pm

My cad software uses V- as the designation for the negative supply when I use a generic opamp. That's not uncommon. It could also be labeled Vss, Vee or on parts designed for unipolar supply, "GND". It's the negative rail and for a sungle supply, the ground is the negative rail.

One "gotcha" that hits some people is that they might select an opamp with rail-to-rail output, but the input is not rail-to-rail. Then if they use it in a single supply application like the one I show above, they do not get linearity of the signal because when the sensor signal is near ground the opamp output will have a positive offset, since the ground is the nefative rail and the opamp can't "see" the input signal at that voltage.

A lot of people like to use the old LM324, but if they do then they have trouble at the positive rail because that old chip can handle inputs down to ground, but can't drive to the positive rail. In that case they have to use a V+ that is at least 1.4 volts higher than the Vdd of the ADC. And then it's good to use a clipper to avoid overdriving the ADC.

There is a new version of the LM324 that has rail-to-rail inputs and outputs, but I forgot the part number.

And some people say "hey, you don't need that 4.7 K resistor at the ADC input"", but I like to have something there to protect against accidental misprogramming or run-away code. It that ADC pin flips to an output, then you would have the opamp and ADC pin trying to drive into eachother and possible blow a pin or cause cyclical resets. I don't like having circuits that could self destruct or latch up due to an unlikely accident. Bad design philosophy. That resistor could be lower, like 220 Ohms or something if you need the increased bandwidth, but I made it 4.7 K so you use the same resistor as the input protection resistor and don't have to chase more parts.
User avatar
Tom Maier
Verified identity
 
Posts: 179
Joined: Mon May 26, 2014 2:37 pm
PIC experience: Professional 5+ years with MCHP products

Re: Amplifier for Sensor to AD

Postby IDEngineer » Tue Dec 05, 2017 11:18 pm

Tom Maier wrote:And some people say "hey, you don't need that 4.7 K resistor at the ADC input"", but I like to have something there to protect against accidental misprogramming or run-away code.

Old thread, I know, but....

The analog reason for having that resistor is to reduce errors due to bias currents on the opamp inputs. For maximum accuracy the inputs should "see" equal impedances.
IDEngineer
 
Posts: 1
Joined: Tue Dec 05, 2017 11:13 pm
PIC experience: Professional 5+ years with MCHP products


Return to ADC & Comparators

Who is online

Users browsing this forum: No registered users and 7 guests

cron