Hi,
I have a strain gauge going to an OPAMP (INA125P) with a BIG gain, and I am thinking that I need to put a high pass filter on the line.
My question is should I do this before it goes to the OPAMP or after? If i do it before will it not just cause more noise on the circuit.
Also, I'm thinking it should be a high pass filter because on the datasheet it seems to be worse on lower frequencies.
Thanks,
Cameron.
cameronasmith:
Hi,I have a strain gauge going to an OPAMP (INA125P) with a BIG gain, and I am thinking that I need to put a high pass filter on the line.
My question is should I do this before it goes to the OPAMP or after? If i do it before will it not just cause more noise on the circuit.
Also, I'm thinking it should be a high pass filter because on the datasheet it seems to be worse on lower frequencies.
http://www.ti.com/lit/ds/symlink/ina125.pdf
Thanks,
Cameron.
You definitely don't want a high-pass filter on a DC measurement
since you'll lose the DC component.
A low-pass filter makes sense. What are you trying to achieve?
When I increase the gain to get the readings I want from the strain gauge the signal is too varied. Even when I use code to smooth the signal.
So, I was thinking using a filter before the amplification might help a lot. Also, moving my circuit off a breadboard onto a PCB as well.
The reason I thought high pass filter was because the OPAMP (as i understand it) works best when the frequency is higher rather than lower.
Hi Pelleplutt,
Thank you for taking the time to draw that for me, but I'm sorry but I don't know what are OP:s
PLease can you explain in more details.
Thanks
The reason I thought high pass filter was because the OPAMP (as i understand it) works best when the frequency is higher rather than lower.
You understand it incorrectly.
If you filter before the amplifier when the signal to noise ratio is very low then yes you can end up with more noise.
However, improving the layout and decoupling around your circuit can greatly reduce the noise. Also using the correct component values, not just ratios will help.
So you need to post a schematic of what you have built for further help.
but I don't know what are OP:s
Operational amplifiers
The reason I thought high pass filter was because the OPAMP (as i understand it) works best when the frequency is higher rather than lower.
Quote the opposite really, opamps are designed principally as DC amplifiers, the frequency
response can be quite poor, especially for instrumentation amplifiers like the INA125,
which are optimized for DC precision. For stabilitiy many opamps start losing
open-loop gain at a few 100 Hz.
A high pass filter removes the DC information completely which is all you care about in
a strain guage amp really!
Okay,
So high pass filters are a no-no. But, what about low-pass filters or will it just add more noise to the signal, and it's not worth it.
But, what about low-pass filters or will it just add more noise to the signal, and it's not worth it.
It depends on the nature of your signal and where the noise is coming from.
If as I suspect it is due to poor layout and circuit then adding more poor circuits only make things worse.
Hi Grumpy_mike.
Please see the attached photo for reference. But please note that I now have the E+ connected by wire to pin 4,
and E- connected by wire to ground.
What do you think? I'm new to electronics, so I don't know what is messy or what is not.
Something I have been thinking about is moving it to a solderboard, which will allow me to solder the Strain Gauge wires right next to the pins (strain gauge wires are currently connected with those blue things)
Thanks,
Cameron.
Initial thoughts are that you have no decoupling anywhere.
You need 0.1uF caps across the supplies and across the vertuial ground resistors as close to the chip as you can.. Also put a large cap about 47uF as well.
Yes a solderd board will improve things.
Thank you for your repsonse Grumpy Mike, I have some questions:
You need 0.1uF caps across the supplies
So anywhere there is a positive connection place a decoupling capacitor.
across the vertuial ground resistors
What is the vertuial ground?
Also put a large cap about 47uF as well.
Where should I place this large capacitor?
BTW: Your website: http://www.thebox.myzen.co.uk/Workshop/Introduction.html is VERY cool!
Virtual ground is a name for a mid-rail reference potential (ground means
reference potential really).
Decoupling goes across the supply, but can here be from virtual ground to each
supply rail. Its purpose in analog circuits is to reduce noise on the supply rails
and reduce the likelyhood of unwanted feedback from amplifier output to inputs
via supply-voltage variation.
White noise is broad-band and reducing bandwidth (such as with a low pass filter)
will reduce noise power in proportion. Noise amplitude is proportional to the square
root of noise power, hence voltage noise is commonly measured in nV/root-Hz
Noise can also be picked up - here shielding of cables and rf-filtering is important.
cameronasmith:
Please see the attached photo for reference.
Do you have your 5V and GND lines shorted together at the GND pin of the ADS1115 module?
So anywhere there is a positive connection place a decoupling capacitor.
Yes but the closer to the chip you can get the better.
