Sensor Op Amp

[rc-pilot-freak]

My sensor outputs 3.4V (min) to 3.9V (max). I would like to have the sensors output to connect to my ADC (0 - 5V).

Any ideas on the amplification circuit?

[vintage1]

Have you more rail voltage than 5v available?

If so an LM324 will do...otherwise it is hard to go right up to the rails...you may end up building something a bit 'custom' out of transistors instead.

Typically to get 0-5v output, you might use +- 12v..and a diode clamp to make sure it doesn't go much below ground or above 5v..

Then one op amp 4 resistors and 2 diodes and maybe a zero pot, will get you all you need.

[rc-pilot-freak]

Well, I forgot to mention the supply is limited to 4-battery cells (4.8V).

[Acetronics]

Hi, The freak ...

Want 5 v from 4.8 v .... looks interesting !!!

Better use 0 - 4.095v ( 10 bits ) ... or 0 -2v 55 ( 8bits )

Alain

[rc-pilot-freak]

Better use 0 - 4.095v ( 10 bits ) ... or 0 -2v 55 ( 8bits )

Do you mean this voltages for PIC Vref? The sensor is allready powered through AD1584BRT - 4.096V Micropower, Precisions Series Mode Voltage References.

[vintage1]

well here is something that will do it given enough supply rails.

All resistors 1% or better please..

You can't get 0-5v from a 4.8v supply..better to try for e.g. 1-3v or something, and use an LM324 type chip. And do the rest in software.

You will need a stable voltage reference for the circuit as well.

[rc-pilot-freak]

Thanks Vintage, but I can only supply the circuit with 4-cells. I don't have the possibility to supply it with +/-12V.

Does LM324 need 12V supply voltage?

I have a single supply AD623ANZ Instrumential Amp on stock. Can I do it with this?

[Acetronics]

AD 623 Works great with a 5v supply ... note Zero won't be a real 0.000v !!!

for that you'll need a negative supply !!!

But you could keep it aside for much better uses ...

By 4.095 v ... I mean the max voltage to use from your AOP Output...to ADC Input.

Alain

[rc-pilot-freak]

By 4.095 v ... I mean the max voltage to use from your AOP Output...to ADC Input.Alain Yes.


How do i make an amplification circuit for my project with this amplifier?

[vintage1]

If you use a 0-5v rail opamp, just take what I have shown and connect the +5 and +12v together, and the 0v and -12v together.

The gain will be correct - times 10 - but it will never quite get to ground, nor up to +5v, or even +4.8v

I fail to see how you expect to get +5v out of a 4.8v battery..

[rc-pilot-freak]

Yes I see what you mean. 4-cells are 5.28V fully charged. I should better use 5-cells and a LM2940 5.0 LDO Voltage regulator to supply the circuit with constant 5V.

I don't understand your "10 gain" amplification? Can you tell me a bit more?

[pmackenzie]

Vintage,
my op amp stuff is rusty, but do you have the + and - inputs reversed?

freak-
you need an overall gain of 10 since your sensor swings by 0.5 volts and you need to expand it to a 5 volt range.
Slightly lower gain would be safer to avoid clipping at the extremes, at the expense of some resolution if the A/D.
You can't just do a gain of 10, you first need to subtract the 3.4 volts otherwise your output will want to be over 30 volts.
V1's circuit looks like it will do both, but as I said I think the inputs might be reversed.
Pat MacKenzie

[vintage1]

Yes I see what you mean. 4-cells are 5.28V fully charged. I should better use 5-cells and a LM2940 5.0 LDO Voltage regulator to supply the circuit with constant 5V.

I don't understand your "10 gain" amplification? Can you tell me a bit more?

Yup. You want a difference of 3.4v to 3.9v to come out at 0-5v..3.9- 3.4 is 0.5v so a gain of ten is required.

That circuit should work OK-ish on +5v, but it will be better if you can drive the opamp from the 5 cell pack direct..

Like this. I've added a 1k pull down resistor that will allow it to drive down to ground or very near.

I think a 324 style opamp would work here, or what was that thing you have? Maybe that as well. You want to get to 1v off the 6v rail.

Yup I did have the + and - reversed..fixed now

[Chippie]

Use a 7660 to generate a negative rail.....this would overcome the problem of needing a 3 rail supply from a 5 cell pack....

10 gain amplification means exactly that.....your 3.4v-3.9v from your device has a signal amplitude of 0.5v to get you to 5v level, it needs amplifying...by a factor of 10...


hope this helps :)

[Nightz]

Should be no need for a negative supply. Many single source rail-to-rail op-amps are available. Especially since your input is well above ground and your measuring a difference of only 1/2 a volt.

Use a differential amplifier to give you a voltage from 0 to 0.5V, then use another op-amp to scale your signal. Scale 0V-0.5V to 0.5V-4.5V (using amp gain with a 500mV offset). This way you stay away from the op-amp rails and you stay away from any possible issues with linearity on your A/D. Some A/D's have issues being rail-to-rail.
Google Op-AMPs. Texas Instruments has an free application note called "Op-Amps For Everyone". Check it out.

Linear tech has free spice software to simulate your circuit.
Multisim 9 is also a free download (45 day trial).
Load one of these programs, and play with the simulation. But don't forget, just because your simulation works does not mean your circuit will....op-amps are fun like that :)

-Rocko

[vintage1]

Use a differential amplifier to give you a voltage from 0 to 0.5V, then use another op-amp to scale your signal.

I see you must have worked for Marcioni Radar.

When I was an apprentice, I was servicing boards from their radar racks. These boards were about the size of a Letter sheet of paper, with gold plated connectors, and slid into a large 19" rack.

Theer of them came to me from one rack. The first had a single op amp in the middle of it. It was being used to buffer the signal. Of course it had a DC offest pot to set it up.

The second was used to invert the signal. Of course it had an offset pot and a gain adjustment pot - all these expensive multiturn Mil Spec trimmers.

The third added two signals together. Of course that also had two gain pots and two offest pots and another op amp to de-invert the signal.

All in all 3 boards, about 6 expensive trim pots, 6 gold plated edge connectors...to do what?.. take a high impedance signal, invert it and add it to another one.

When I evenrually saw the specification, I understood why.

The original design guru had drawn a block diagram with a buffer, inverter and adder. And the dimwit in charge of the design had assigned each of these blocks to individual junior engineers, who had duly laid out a PCB for each of them.

If you examine the circuit I posted, you will see that the offset, and gain are all handled withing one chip, and the output swing will be as good as the chip is.

Now I do admit that I had to do some maths to get it all right, and to find three resistors that were in the correct ratios from the E24 range...but even that is surely not beyond the bounds of human comprehension?

Or is it that today, if its not in the 'book of circuits' no one does it at all?

I also remember we had a project in at one time for a laser rangefinder. It wasn't good enogh. Us young first year trainees were given the task of 'getting it to work to spec'...try as we might te best we could do was picck out a house with white walls a mile away. It needed to do two.

Eventually one of the senior people turned up on a visit, and we unburdened our troubles to him. "Ah", he said "What power is the laser? and what is the noise figure on the photodiode?" ..we duly told him..there was a shuffling of the back of an envelope, the sqeaking of his pocket slide rule, and he said "I make the theoretical range just about one mile, if you have a nice white painted target on a clear day: You will never get more than that out of it"

It was at that point I decided to leave the high tech world of missiles and military avionics, in search of something more meaningful than wasting the taxpayers money on things that could be better and more cheaply done by commercial companies. :D

[Nightz]

Vintage...
Yup....you can use the same op-amp to do both stages that I wrote about however, since rc-pilot-freak will be purchasing the op-amps (sounds like he does not have any yet) he may as well buy some duals to play around with. I doesn't really matter. You can do it a dozen different ways. Sometimes it helps a beginner to see each step before pulling it all together. I was trying to keep it simple (by explaining each stage) for the guy.

rc-pilot-freak,
Looks like you have more than one good method to try. I suggest trying all of them on a bread board so you experience how op-amps work. Go with the solution your most comfortable with and fits your application best.

Regards,
-Rocko