Voltage signal (-0.40mV ~ 7.60 mV) to NI card and LV

[Lathe_Biosas]

Hi

I'm trying to plot a voltage signal (-0.40 mV ~ 7.60 mV) on LabView
unfortunatelly I haven't achieved reading the signal with mV
resolution, my conclusion is that the signal needs to be amplified.

Could anyone please tell me if amplifing the signal would help? If yes
how can it be amplified?

I have an Analog Devices AD595 (Monolithic Thermocouple Amplifier)

http://www.analog.com/UploadedFiles/Data_Sheets/421725987AD594_5_c.pdf

and I was wondering if it would help


Any help or hints would be kindly appreciated
Best Regards

 

[Martin Riddle]

The AD622 works for me in situations like this.

Cheers

 

[Walter Harley]

Hi

I'm trying to plot a voltage signal (-0.40 mV ~ 7.60 mV) on LabView
unfortunatelly I haven't achieved reading the signal with mV
resolution, my conclusion is that the signal needs to be amplified.

LabView is software, not hardware. From your title I guess you're using a
National Instruments data acquisition card. They make quite a few models.
Which one are you using?

Could anyone please tell me if amplifing the signal would help? If yes
how can it be amplified?

But anyway, yes, amplifying it would probably help; alternatively, you need
data acquisition hardware with appropriate resolution (for instance, a
digital voltmeter with a mV scale and a computer output of some sort).

Accurately amplifying a DC signal that small is not trivial. Some things
you'll need to specify are:

- what resolution do you need? E.g., 10uV, 1nV, ...?

- how much drift, over time and temperature, is acceptable? (Are you
trying to measure over a timespan of milliseconds or days?)

- what speed do you need? E.g., is this a slowly-varying DC signal, or a
1MHz waveform, or ...?

- what is the output impedance of the signal source? (Alternately, what
input impedance do you need from your amplifier?)

I have an Analog Devices AD595 (Monolithic Thermocouple Amplifier)

http://www.analog.com/UploadedFiles/Data_Sheets/421725987AD594_5_c.pdf

and I was wondering if it would help

Are you measuring the output of a thermocouple? If not, then probably not.

But anyway, achieving fractional-mV resolution in a DC amplifier depends as
much on good circuit design and layout skills as on the parts you use, and
just having the right IC (even if it were, which it probably isn't) is not
sufficient.

 

[Lathe_Biosas]

LabView is software, not hardware. From your title I guess you're using a
National Instruments data acquisition card. They make quite a few models.
Which one are you using?

LabView 6.1 with a PCI-6024E

But anyway, yes, amplifying it would probably help; alternatively, you need
data acquisition hardware with appropriate resolution (for instance, a
digital voltmeter with a mV scale and a computer output of some sort).

A Fluke 192B Scopemeter was used with success. Now the goal is to see
a waveform in Labview

Accurately amplifying a DC signal that small is not trivial. Some things
you'll need to specify are:

- what resolution do you need? E.g., 10uV, 1nV, ...?

Steps of .70 mVols aprox but it doesn't need to be extremly accurate.

- how much drift, over time and temperature, is acceptable? (Are you
trying to measure over a timespan of milliseconds or days?)

It could be nice to graph on Labview using for example a waveform.
Going from aprox. -0.40 to 7.60 mV takes more or less 5 minutes.

- what speed do you need? E.g., is this a slowly-varying DC signal, or a
1MHz waveform, or ...?

As much as I know it is a slow varying DC signal

- what is the output impedance of the signal source? (Alternately, what
input impedance do you need from your amplifier?)

Unfortunatelly I don't know that.

Are you measuring the output of a thermocouple? If not, then probably not.

Yes, I'm measuring the output of a thermocouple.

But anyway, achieving fractional-mV resolution in a DC amplifier depends as
much on good circuit design and layout skills as on the parts you use, and
just having the right IC (even if it were, which it probably isn't) is not
sufficient.

Thank you very much for your answer, so what would you recomend me to
do?

Best Regards

 

[Lathe_Biosas]

Hi,

Thanks for your answer, I will look for the AD622 (cool device), if I
don't find it I will try with a 741

Regards

 

[Martin Riddle]

The 622 is an instrumentation amp, not an OPamp like the 741. It has very good offset voltage typ 125uv.

You should amplify the signal to ~4v for the NI 6024. I believe this is a 16bit card. For the +/-5 volt range the resolution is
150uv. Which is adequate.

Cheers

 

[Walter Harley]

[...]
Yes, I'm measuring the output of a thermocouple.

So, the simplest and best thing you could do would be to get a DAQ card with
a thermocouple input, such as the NI USB-9211, which they list at $395, lead
time <1wk. I don't know your budget, but I know that building a homebrew
thermocouple amp, if you've not done it before, is going to cost you at
least $50 in parts plus at least 2 days of work, and you won't end up with
anywhere near the accuracy, nor will you end up with something resellable;
so you can do the cost-benefit analysis yourself. (To the naysayers: yes,
if you've got some prior experience and a well-stocked junk drawer, you
could do better; but if you haven't done it before, then no.)

Barring that, given that you've got the thermocouple amp chip, starting with
its app notes seems like a reasonable approach. If they make no sense to
you, that's informational.

 

[Walter Harley]

Hi,

Thanks for your answer, I will look for the AD622 (cool device), if I
don't find it I will try with a 741

Don't bother with a 741.

Look at the DC offset voltage spec.

 

[Pooh Bear]

Hi,

Thanks for your answer, I will look for the AD622 (cool device), if I
don't find it I will try with a 741

Forget the 741.

If you just need reasonable DC precision the LM11 may fit the bill.

Graham

 

[Lathe_Biosas]

Hi

I have an LM311, would it help?

Regards

 

[Spehro Pefhany]

Hi

I have an LM311, would it help?

Regards

Not at all.



Best regards,
Spehro Pefhany

 

[Lathe_Biosas]

And the AD620?

http://www.analog.com/UploadedFiles/Data_Sheets/37793330023930AD620_e.pdf

Regards

 

[Walter Harley]

Not at all.

Here I must respectfully disagree with Spehro. The LM311 is just what you
need. Put it in a box - ideally, you might want to solder it to a printed
circuit board, and then mount the printed circuit board in a box, so that it
doesn't rattle. Also put some audio connectors in the box, and wire the
input connectors to the output connectors. Now, sell the box to an
audiophile, for $395 plus tax and shipping. Use the proceeds to buy a
thermocouple card to run LabView with.

 

[Jim Thompson]

Here I must respectfully disagree with Spehro. The LM311 is just what you
need. Put it in a box - ideally, you might want to solder it to a printed
circuit board, and then mount the printed circuit board in a box, so that it
doesn't rattle. Also put some audio connectors in the box, and wire the
input connectors to the output connectors. Now, sell the box to an
audiophile, for $395 plus tax and shipping. Use the proceeds to buy a
thermocouple card to run LabView with.

Sno-o-o-ort

...Jim Thompson

 

[Richard the Dreaded Libertarian]

Sno-o-o-ort

Oh, my GOD! I laugh at the same stuff as Thompson? Will somebody please
euthanize me?

;-D
Rich

(By the way, this is a joke.)

 

[Fred Bloggs]

Could anyone please tell me if amplifing the signal would help?

Not as much as reading the specifications- the 6024E can be programmed
for FS of +/- 50mV with 0.008mV resolution and 0.106mV absolute accuracy
using 100 point averaging.

 

[Lathe_Biosas]

Hi,

I choose the AD620 but I don't know if I wire it right

-IN and +IN are connected to the Thermocouple
-Vs and +Vs to power supply -12 and +12 Volts
REF is connected to ground
OUT directly to the voltmeter

The gain was set to aprox. 1000 with a 47 Ohm resistance

Is that ok? Unfortunatelly with the power supply -Vs=-5 Volts and +Vs=5
Volts the output
doesn't represent a gain of 1000.

Any recomendation?
Best Regards