I have a question about voltage-to-current conversion.

current signals are good for long runs where voltage drop across the conductor becomes a problem. 10m with a properly gauged conductor a

Victor hyppolito wrote:

Victor hyppolito scrobe on the papyrus:

Why not convert it to a number with an ADC and then send an ASCII string over a serial link? That way you can send it as far as you want.

The choice has a lot to do with degradation of the signal as seen by the computer. This depends on the source and load impedances, noise sources, and choice of conductors. A 0-5V signal in a two conductor shielded cable should be fine, especially if the conductors are internally twisted, and the load impedance is high. Another problem with voltage is when there is a ground loop from the computer to the sensor, but this is not a problem if differential means are used.

Current signals (typically 4-20 mA) are used in industrial monitors and controls where the load can have electronics which are powered from loop current, and it adjusts the load current which is read by the sender. This allows power and signal to share a single pair of wires, and current is not affected by wire resistance.

Paul

But how do you input this signal into the computer? Do you have a data aquisition card or do you use your soundcard for it. And what kind of signal is it and how do you amplify it? Tell a bit more about what sensor you are using, maybe it isn't even necessary to amplify or causes more errors. A

Hello Victor,

If the computer does not have a good AD converter built in but does have a sound card, I'd say no. I would convert to a frequency. Reasons: The game port AD conversion is often not that good (and none of mine have a game port to begin with). With the sound card you can use a simple FFT routine to determine the incoming frequency. There is an abundance of such SW as freeware or shareware.

A good voltage to frequency converter chip is National's LM331.

Hello Ban

I=B4m working with strain gauges. My goal is to amplify the output signal of a wheatstone bridge(0-10mV) and send it to another room(10 meters away). Afterwards, I would input it in an aquisition card (basically a adc converter and a PIC). Thats why I think i need to amplify it and maybe send it as a current signal. The data will go to the computer via rs232 using an IC max232 to convert TTL to -12, 12V. Thanks for the reply! Ciao! Victor D'ippolito

Ban escreveu:

nal

be

Have you considered an AC excitation?

...Jim Thompson

Hello Jim,

Preferably at a frequency outside the harmonics of 60Hz or in Victor's case probably 50Hz.

Victor hyppolito scrobe on the papyrus:

.. ..

I get bad karma about this idea. Why have your ADC/PIC card 10 metres away from the signal source? It should be as near to the strain gauges as possible. The RS232 signal can travel considerably more than 10 metres with minimal degredation.

a) 10mV IS very strong signal as far as strain gauges are concerned. b) in your place would just use 2 pair shielded cable and "Damn the torpedoes!" c) The conversion to current is very convenient when working in very noisy enviroment or working on bodies where weight is of concern. The weight of wires and their connectors on properly "telemetered" plane or missile is to be considered. As long as the connection exist you have a current flow. In lab conditions, not worth the effort.

Have fun

Stanislaw Slack user from Ulladulla.

Hello Jim

I haven't considered AC excitation... Actually I don't know how to use AC excitation. What signal should a use? A sine wave, square wave? Do I use RMS valures? Sorry about my ignorance! Thank you for the reply! Victor D'ippolito

Just drive your bridge with a sine wave instead of DC.

I've always been fond of sine wave excitation since it makes getting rid of OpAmp offset voltages a snap.

...Jim Thompson