Load cell amplifier IC required or a high resolution ADC

[arun]

Hi,

Can anyone suggest some Load cell amplifier ICs, I tried using a MCP 3208 ADC directly with the load cell bridge arms , I find the output very low , as low as 1mV.

A higher resolution ADC might do , Microchip seems to have reached the limit with 12 bit ADCs.

I found some like CS 5520, CS 5516, but these are very expensive and sophisticated to use.



thanks
arun

[Will Reeve]

I am using a AD8551 opamp as a load cell amplifer into a 10bit ADC (don't need the higher resolution as you seem to). Are you sure 12bits is not high enough?

Will

[sar]

Check out this page using INA125P instrumentation amplifier.

http://psas.pdx.edu/LoadCellAmp

[Guest]

We use both LTC2408 24bit A/D or INA125U (nice as it has built in programmable reference).

Woz

[Humberto]

Hi,


I used the MCP3204 (4Chnnl) 12 bit ADC) wich is the young brother of the MCP3208
(8Chnnl) 12 Bit ADC and it function like a shame, but I noticed that for a good behaviour,
it needs a low impedance signal source. (say < 1K). Reading your comments,
I guess that you have a big impedance mismatch. I would think in an instrumentation
ampliffier as front end to get a good improvement matching the Load Cell with a low
impedance ADC.
I had to work around the MCP3208.C driver because it doesn´t work properly.

Best wishes,

Humberto

[Kenny]

Maxim have cheap single-supply instrumentation amplifiers suitable for strain gauge
applications like load cells.

MAX4462HESA has a fixed gain of 100.
I have used lots of these to amplify the signals from strain gauges on composite materials.
http://pdfserv.maxim-ic.com/en/ds/MAX4460-MAX4462.pdf

In your case if you really only have a 1mV signal from the load cell you will need a gain
of at least 1000 to get into the volts range. Gains greater than 1000 are generally not used
because of problems with noise and drift.
I suspect that you are using only a limited part of the load cell's range.
The MAX4194 is another one I have used with metals.
http://pdfserv.maxim-ic.com/en/ds/MAX4194-MAX4197.pdf
The gain can be set with a single resistor.

Both of these have a reference input that can be used to offset the output if the load cell
is used for both tension and compression.

I would recommend that if possible use a load cell more appropriate to the expected load
so as to get a bigger signal, and therefore a better signal/noise ratio.

[Will Reeve]

The reference design for the AD8551 (from the datasheet) gives 0 to 40mV from the strain guage (running off 4v) and gives a 0 to 4V output. Perfect for the PIC. I guess you need anothe x4 gain to get your 0 to 1mV up to 4V which might give you some noise problems, layout would be quite critical.

Will

[kender]

I had a similar problem recently. My load cell had a differential output (S+ and S-) of +/- 200 mV full swing. I ended up using an instrumentation amplifier LT1167. Keep in mind though - most in-amos do not have a rail-to-rail output. If you need the R2R output - make you own "in-amp" out of R2R op-amps (such as LMC6462, LM6132).

Just my $.02.

Nick

[dave]

As you will probably know most load cell bridges are 350 ohm impedance, the usual standard for load cells is 2mV/volt, and most instrumentation has its own onboard 10 volt load cell bridge power supply so for full working range output of the load cell with 10 volts is 0-20mV.
Regards Dave

[rRolf Meier]

[kender]