DC voltage converter 0-110mV to 0-5V

--- Since:

you'll need a gain of 50, and you don't need an inversion, so you can do this:

Vin>------|+\\ | >--+-->Vout +--|-/ | | | +--[R2]--+ | [R1] | GND>---+----------->GND

An easy way to work out the resistors is to decide how much current you want in the divider and then choose R1 to drop the maximum voltage on the + input with Vout at its maximum output also.

In your case, if you wanted to have 1mA in the divider, then with

E 0.1V R = --- = -------- = 100 ohms I 0.001A

and R2 would have a value of:

E 4.9V R = --- = -------- = 4900 ohms I 0.001A

Looking at the divider like this:

E1 | [R2] | 0.1V +----E2 | [R1] | GND

and choosing 1% resistors for R1 and R2, the opamp output voltage with 0.1V on the + input will be:

E2 (R1+R2) 0.1V * 4970R E1 ------------ = -------------- = 4.97V R1 100R

So, you've got a 30 millivolt error.

If you diddle around with the resistors you can probably do better than that, but the easy way out is just to use a decent multi-turn trimpot, put 100mV on the + input, and crank away until you get precisely 5.00V on the output. BTW, the classical formula for the circuit is:

R1 + R2 Vout = -------- Vin R1

-- John Fields Professional Circuit Designer

...Jim Thompson

Thank you John. You are the man. What op-amp would you recommend that would preferably not need to be mounted on circuit board's.

Depends on the supply voltages you can provide. I usually take the old LM324 for this but it's pretty high in offset, a few mV. If you need better precision look for a rail-to-rail opamp with less offset. TI, National etc.

Regards, Joerg

Given that the input is from a current sensing shunt, it pays to feed it into differential amplifier to nulify the voltage present on the sensor. Very few current sensors connect directly to some zero reference point. Depending on your expertise such can be made from a set of dual input NAND CMOS gates. _VERY low power_!

Good luck

Stanislaw Slack user from Ulladulla.

I'm afraid that thing is pretty easily blown up, depending where you put your current shunt. If it is on the high side of a power supply and both the to be measured supply and your data recorder are grounded, then I just hope your supply has a good current limit or you will see your great opamp circuit smoke really fast. :-) Better to buy one of these multimeters with built-in interface and read-in and multiply with matlab or whatever works on linux.

In article , Stanislaw Flatto wrote: [....]

C1A -----!!---- ! ! R1A ! R2A !

-in ---/\\/\\------+-+----/\\/\\/--+-- ! ! ----!-\\ ! ! >--------+----------- Output ----!+/ R1B !

Gain = R2/R1 (C1 explained below)

Even if (you think) the current shunt is connected to ground, the wire that actually makes this connection will have some resistance. This resistance will vary with temperature and the phase of the moon so you really want to keep it out of the measurement.

Most shuts have two big screws for the current connections and two small ones for the "meter" connection. The "-in" hooks to the "meter" connection on the ground side of the shunt.

To see that this circuit works, assume no current flows on the inputs of the op-amp, the output is zero and "-in" and "+in" are both at 1V. The series combination of R1A+R2A will have exactly the same current in it as R1B+R2B. As a result the two input of the op-amp will be at exactly the same voltage so, the op-amp is happy.

About C1:

If you use the LM324 op-amp (or many like it) there can be a problem with high frequency signals and noise. If the op-amp can't flow the fast changing signal, you will start to see a non-zero difference voltage at the input of the op-amp. If this difference is large enough it can cause the input to "rectify" this high frequency signal making a low frequency output that tracks the amplitude of the high frequency stuff.

Getting 1% capacitors is hard so you want them to be small enough values that they don't effect what you are attempting to measure. The frequency the capacitors start to matter at needs to be low enough that they do prevent the inputs from having much AC voltage on them. For an initial guess, I'd pick 10 times the highest frequency you wish to measure.

C = 1 / (2 * PI * R2 * F)

The power supply will be a six SCR 3 phase xfmr/rectifier. Frequency is a non issue and if I can get the average of the last 0.1 second's, I would be fine. The output of my "voltage multiplier" is an input to a SCR controller board that requires 0-5V signal.

If I can get, say 5% accuracy, I would be fine. So I am looking for something basic.

There is a need to put in some sort of a differential function in, as the DC output of this power supply is not supposed to be tied to any ground, even though it is indeed on the low side. So, logically speaking, I need this

Input voltage floating +/- ===> differential amplifier with unity gain ===> your schematic to convert 100 mV to 5V

OK, but I meant what kind of a supply is available to operate the converter?

Look at the LEM LA55-P Hall Effect isolated current sensor.

Full Scale input is 100 Ampere-Turns, so wind on whatever turns are required for your sensitivity. Output current is Input A-T/1000, so load with whatever resistor value is needed to get 5V.

I am sorry, there is no need for integration, I only meant that there are no requirement's for particularly fast response of high frequency event's here, only rectified DC that needs to be measured.

On Tue, 18 Apr 2006 16:04:21 GMT, snipped-for-privacy@emporium-inc.com wrote:

--- OK.

Here's what I'm assuming you've got on the secondary side of the transformer: . .+--------+ .| S1|--[>|]--+----+ .| | | | .| S2|--[>|]--+ [LOAD] .| | | | .| S3|--[>|]--+ | .| | +---- .| | | .| | [SHUNT] .| | | .| | +---- .| | | .| COM|-------------+ .+--------+

and here's how the converter fits in:

. . +--[10K]--+ . | | . | +12 | . | | | +12 . +-----+--[10K]--+---|-\\ | | . | | | >--+--|+\\U2 . | [SHUNT] +---|+/U1 | >---+-->0-5V OUT . | | | | +--|-/ | . | +--[10K]--+ GND | | | . [RS] | | | GND | . | [LOAD] | +--[10k]--+ . | |+ [10K] | | . | [SUPPLY] | [204] | . | | | | | .COM>---+-----+---------+---------+---------+

RS is the resistance of the wiring connecting the low side of the shunt to the supply output common, and U1 is the difference amplifier which Ken Smith noted needed to be in there to keep the drop across RS from appearing in the output.

If you use Linear LT1218's (or any other rail-to-rail opamps with low input offset voltages you won't have to trim anything.

Here's the LTSPICE circuit file if you want to simulate it:

Version 4 SHEET 1 880 680 WIRE -688 272 -688 176 WIRE -688 544 -688 352 WIRE -688 576 -688 544 WIRE -512 272 -512 80 WIRE -512 544 -688 544 WIRE -512 544 -512 352 WIRE -400 80 -512 80 WIRE -400 112 -400 80 WIRE -400 240 -400 192 WIRE -400 304 -400 240 WIRE -400 416 -400 384 WIRE -400 544 -512 544 WIRE -400 544 -400 496 WIRE -336 80 -400 80 WIRE -336 240 -400 240 WIRE -208 80 -256 80 WIRE -208 176 -208 80 WIRE -208 240 -256 240 WIRE -208 240 -208 208 WIRE -208 272 -208 240 WIRE -208 544 -400 544 WIRE -208 544 -208 352 WIRE -160 80 -208 80 WIRE -160 176 -208 176 WIRE -160 208 -208 208 WIRE -128 544 -208 544 WIRE -128 544 -128 224 WIRE -32 80 -80 80 WIRE -32 192 -96 192 WIRE -32 192 -32 80 WIRE 16 544 -128 544 WIRE 16 544 16 128 WIRE 80 336 80 224 WIRE 80 400 80 336 WIRE 80 544 16 544 WIRE 80 544 80 480 WIRE 112 336 80 336 WIRE 128 192 -32 192 WIRE 128 224 80 224 WIRE 160 128 16 128 WIRE 160 176 160 128 WIRE 224 208 192 208 WIRE 224 336 192 336 WIRE 224 336 224 208 FLAG -128 160 +12 FLAG -688 576 0 FLAG -688 176 +12 FLAG 160 240 +12 SYMBOL res -224 256 R0 SYMATTR InstName R1 SYMATTR Value 10K SYMATTR SpiceLine tol=.1 SYMBOL res -240 64 R90 WINDOW 0 -34 55 VBottom 0 WINDOW 3 -33 53 VTop 0 SYMATTR InstName R2 SYMATTR Value 10K SYMATTR SpiceLine tol=.1 SYMBOL res -240 224 R90 WINDOW 0 64 65 VBottom 0 WINDOW 3 64 65 VTop 0 SYMATTR InstName R3 SYMATTR Value 10K SYMATTR SpiceLine tol=.1 SYMBOL res -64 64 R90 WINDOW 0 -31 56 VBottom 0 WINDOW 3 -31 61 VTop 0 SYMATTR InstName R4 SYMATTR Value 10K SYMATTR SpiceLine tol=.1 SYMBOL voltage -688 256 R0 WINDOW 123 0 0 Left 0 SYMATTR InstName V1 SYMATTR Value 12 SYMATTR SpiceLine Rser=1E-6 SYMBOL voltage -400 400 R0 WINDOW 123 0 0 Left 0 SYMATTR InstName V2 SYMATTR Value .3 SYMATTR SpiceLine Rser=1E-6 SYMBOL res -416 96 R0 SYMATTR InstName R5 SYMATTR Value 1 SYMBOL res -528 256 R0 SYMATTR InstName R6 SYMATTR Value 1 SYMBOL res -416 288 R0 SYMATTR InstName R7 SYMATTR Value 1 SYMBOL Opamps\\\\LT1218 -128 128 R0 WINDOW 0 34 88 Left 0 WINDOW 3 12 120 Left 0 SYMATTR InstName U2 SYMBOL Opamps\\\\LT1218 160 272 M180 WINDOW 0 33 86 Left 0 WINDOW 3 13 120 Left 0 SYMATTR InstName U1 SYMBOL res 64 384 R0 SYMATTR InstName R8 SYMATTR Value 204 SYMATTR SpiceLine tol=1 SYMBOL res 208 320 R90 WINDOW 0 0 56 VBottom 0 WINDOW 3 32 56 VTop 0 SYMATTR InstName R9 SYMATTR Value 10k SYMATTR SpiceLine tol=1 TEXT -650 578 Left 0 !.tran 0 .1 0

-- John Fields Professional Circuit Designer