Amplifier circuit, need more sensitivity or gain?

Thread Starter

Nevalite

Joined Feb 20, 2012
45
I get the idea, I found the offset circuit in the data sheet, I will give all this a go tomorrow back at school. Thanks for your help.
 

Thread Starter

Nevalite

Joined Feb 20, 2012
45
My only concern is that in the datasheet the circuit given for offset adjustment is in the order of mV, I am thinking off offsetting in the order of a couple Vs, will this work?
 

Thread Starter

Nevalite

Joined Feb 20, 2012
45
I decided against the offset circuit. I added a 12v regulator and I am now running the V+ of the INA131 from the 12v, this way I can take the output from the gauges and amplify it even further.

Is there any reason why 5v on the gauge and 12v on the amp will cause inaccuracies or other problems?
 

Ron H

Joined Apr 14, 2005
7,063
The INA131 amplifies the difference between the two inputs. If your bridge is balanced, it will try to read 0V on the output. It can't do that with a single supply. You need positive and negative supplies.
 

Thread Starter

Nevalite

Joined Feb 20, 2012
45
The difference between the two inputs (Vin) or the difference between the V and GND?

If I used an INA122 would I be able use a single supply and read 0v at no difference?
 

Ron H

Joined Apr 14, 2005
7,063
The difference between the two inputs (Vin) or the difference between the V and GND?

If I used an INA122 would I be able use a single supply and read 0v at no difference?
The main reason for instrumentation amplifiers is to amplify the difference between the two inputs. The voltage that the difference is riding on (+2.5V, in your case) is called the common mode voltage. If you look at the datasheets, you will find a spec for common mode rejection. The bigger this number is, the better the amplifier will "ignore" the common mode voltage.
With a single power supply, the INA122 will typically swing to within about 50mV of ground (100mV guaranteed). Divide this by the gain, and you will get the value of the difference voltage required before the output will begin to respond.

If your deflection can be in both directions (positive or negative difference voltage), you will need positive and negative power supplies, unless you can tolerate having the output ride at a positive voltage when the bridge is balanced. If you can, you can offset the output by applying a voltage to the reference input.
Is your deflection in both directions?
How are you going to process the output of the amplifier? Does it go to an A/D converter?
 

Thread Starter

Nevalite

Joined Feb 20, 2012
45
The deflection is only positive, the gauge is on a pull rod which is always in tension. The output goes to a DL1 data logger which has a built in 12 bit A/D converter.

If I understand correctly the INA122 running off my 12v battery should work for positive readings. I am going to try and hook it up now.
 

Thread Starter

Nevalite

Joined Feb 20, 2012
45
So the circuit is working, I have to turn the gain to about G=2200 to get good results...the output is at 9V unloaded and I get up to 11V with 500lbs of load. I can't go any further because the output will max out under 500lbs. Is there any way to get better resolution?
 

Ron H

Joined Apr 14, 2005
7,063
So the circuit is working, I have to turn the gain to about G=2200 to get good results...the output is at 9V unloaded and I get up to 11V with 500lbs of load. I can't go any further because the output will max out under 500lbs. Is there any way to get better resolution?
You should not be getting 9V if your bridge is balanced. Are you using precision resistors in your bridge?
 

Ron H

Joined Apr 14, 2005
7,063
I believe the bridge is a strain gauge, meaning that all four resistors in the bridge are strain gauges.
I missed that. The comment about SG1 andSG3 being the active gauges threw me, but I suppose that means that the other two are oriented orthogonally to the direction of strain.
I suspect he still has a ground missing on the amplifier.
 

Thread Starter

Nevalite

Joined Feb 20, 2012
45
As stated, the bridge is the strain gauge. Two of the resistors are oriented in the axis of strain and the other two are perpendicular dummy resistors.

When you excite the bridge un-amplified you get in the range of 3mV unloaded-10mV loaded coming out of it, why would I be expecting 0V unloaded from my amplified output? Shouldn't I expect 3-10mv times my gain to give me some voltage unloaded?
 

Ron H

Joined Apr 14, 2005
7,063
I don't see anything wrong with the circuit, but you are likely getting oscillations from the long wires to your gain pot, or whatever is on the end of those long wires.
EDIT: You can use a single turn or multiturn trimpot, mounted very close to pins 1 and 8.
You also need a 0.1uF ceramic cap from pin 7 to pin 4, bridged across the IC, with the leads as short as possible. This is necessary to prevent oscillations.
Please post a link to the datasheet of your strain gauge, or at least the part number.
 
Last edited:

bertus

Joined Apr 5, 2008
22,270
Hello,

I also see no capacitors near the regulators.
Those are needed to prevent the regulators from oscillating.
(see the datasheet for more info).

Bertus
 

Thread Starter

Nevalite

Joined Feb 20, 2012
45
Yeah the gain pot was just temporary to see what resistance I needed to get the right gain. I also calculated it theoretically but I was just testing.

As for all the capacitors, I will add those now.

The strain gauges I'm using are omega SGT-2/350-FB11. There is no data sheet for just that gauge but here is one for a whole bunch: Strain Gauges

They are full bridge, 350 ohm +/- 15%, gage factor of 2.13
 

Thread Starter

Nevalite

Joined Feb 20, 2012
45
Is there any way for me to get a better range than 9V-11V under the same loading conditions? I can't increase the gain anymore without hitting 12v. Could I inject -9V into the ref pin and offset the entire thing and then turn up the gain some more?
 

Thread Starter

Nevalite

Joined Feb 20, 2012
45
The thing I don't understand is that when my gain is set to 5 (no resistor Rg) I get 18.6mV, With my 80ohm resistor, G=2358 I read 8.81V. Do I need to design an offset circuit for 18.6mV or 8.81V?
 
Top