LF347 as instrumentation Amplifier

Thread Starter

viju

Joined Sep 28, 2008
120
Ron:

I have been buiding this circuit in my office where camera is not allowed.Basic idea is to find the resistance of a welded joint.I pass a known current to the welded joint and measure the voltage drop across it.Depending upon the value of the resistance ( typical 10~14 mΩ for my application) voltage drop will vary.I need to amplify the milllivolts getting across the welded joints to the gain of 1000 so that I can get Vout of 0~10 Volts( since I am getting in millivolts,my PLC input is 0~10Volts amplification is must with accuracy and linearity).

These type of applications are working in my office in more numbers but all of them are supplied by my suppliers.
 
Last edited:

Ron H

Joined Apr 14, 2005
7,063
The reason I was asking for a photo is I still suspect you have a grounding problem. Maybe you could provide a drawing - and I don't mean a schematic, but a drawing, with dimensions and wire gauges, if possible.
 

Thread Starter

viju

Joined Sep 28, 2008
120
Let me a days' time to draw the circuit as I don't have the software.But the system works with some linearity with OP07 as an amplifier.But the gain suffers the linearity for 7mV and above.I have set the gain for 1000.

Kindly suggest me:

I think gain of 1000 need to be build in two parts. Gain of 100 fed to Gain of 10 in the next stage.Am I right?
 

Ron H

Joined Apr 14, 2005
7,063
The AD620A does not seem to have a power supply rejection spec. In light of that, I would not use switch-mode power supplies. Switch-mode supplies are generally not a good idea for sensitive (high gain) analog applications.
 

Thread Starter

viju

Joined Sep 28, 2008
120
Will I have to power the circuit through transformer rectifier circuit?Can I introduce ISO 120 isolation amplifier between input and output?So that noise can be eliminated?Data sheet says that no other components are required for isolation other than ISO 120.

Ron:

Please advice me if I can go for two series OP-amps with the gain of 100 and 10?I found this idea in this forum suggested by you members.
 
Last edited:

Ron H

Joined Apr 14, 2005
7,063
Bear with un-professional way drawing. I have also tried connecting ground to the "REF" terminal.But it didn't respond as I said earlier.It works well for 1 V input signal or more with the lesser gain.
I would do it like this.
I think the AD620A has adequate linearity (from the datasheet) to use at a gain of 1000.
 

Attachments

Thread Starter

viju

Joined Sep 28, 2008
120
Ron:

I also tried with the connection as you suggested.But the outcome was not encouraging. Could you please explain the Linear SMPS. I have a SMPS from Meanwell.
 

Ron H

Joined Apr 14, 2005
7,063
Ron:

I also tried with the connection as you suggested.But the outcome was not encouraging. Could you please explain the Linear SMPS. I have a SMPS from Meanwell.
A power supply is either linear, or it's a Switch Mode Power Supply (SMPS). There is no such thing as a linear SMPS. SMPS's are generally noisy, and may not be suitable for powering sensitive, high-gain circuitry. If I were having your problems, that's the first thing I would change, at least until I had absolutely eliminated it as a source of errors.
Please post a link to the datasheet of your ±12V SMPS.

How are you measuring the input voltage? How are you varying it?

I would suggest that you Google "instrumentation amplifier application note" and read several of them carefully, if you haven't already done so. Grounding and shielding are absolutely critical. When you actually start measuring welds, issues like unintended thermocouple junctions, common-mode voltage on the pipe, etc., will probably arise.
 
Top