Noise Reduction on PRESSURE SENSOR SIGNAL CONDITIONER

[witssq]

I'm designing high precision pressure sensor signal conditioner.

And I would like to reduce noise.



Although many issue about the attached schematic exist, initially noise at between TP7/TP8 and 24PC sensor.





<Test 1>



When 3Pin(GND), 1Pin(+10V) is established, then meaure at TP7, TP8 position with 1.5M Shield Cable connected with J3



Peak to Peak Value



TP11(+10V) : 50mV

TP13 : 100mV FIG. 3_13

TP15, 16 : 100mV FIG. 3_15_1

TP15, 16 (Opened/Not Connected) :

65mV Signal Shape is same as FIG. 3_15_1

TP7,TP8 Peak to Peak : 80mV FIG_3_7, F_3_8

TP9,TP10 (Opened/Not Connected) : 300mV Signal Shape is same as FIG. 3_7





<Test 2>
When 3Pin(GND), 1Pin(+10V) is established, then meaure at 4Pin, 2Pin(Signal) without connection to J3(DIN Connector)


1Pin(+10V) Peak to Peak : 40mV
4Pin , 2Pin Peak to Peak : 30mV(Signal Noise size is reduced)


I found initial power supply ripple 50mV to signal 80mV ripple.
But Not connected signal ripple is smaller than power ripple.


I found that FIG_3_15_1 hum ocurrs power on Fluorescent light plugged on common AC multi-outlet receptacle with system. When this lamp is off,
hum is disapeared.

Please advise me how I can make not susceptable to this lamp-on
and how do I reduce this noise.

Best regards,

 

[chuckey]

Do you want your sensor to respond so fast that you need the a bandwidth to encompass 50Hz? Most likely not, just connect capacitors across you amplifiers thar will rid you of the 50 Hz, but will slow down the amplifiers response to your transducer. I could not open your air_con circuit, it closed my browser down.
Frank

 

[witssq]

Do you want your sensor to respond so fast that you need the a bandwidth to encompass 50Hz? Most likely not, just connect capacitors across you amplifiers thar will rid you of the 50 Hz, but will slow down the amplifiers response to your transducer. I could not open your air_con circuit, it closed my browser down.
Frank

I can see all the attachment here.
Attaching new schematic file, Please review my schematics.

Best regards,
SunSung.

 

[mtwieg]

Simplest solution is to, as chuckey said, lower your measurement bandwidth below 50/60Hz. Like maybe down to 5Hz or something (if this is for environmental measurements, then even less than 1Hz would be fine). You'll want to put your capacitor between the outputs of the bridge (since it's a differential signal). Technically power supply ripple shouldn't be a problem since it will only induce common mode voltages, which should be ignored by you in amp.

But you probably want to think about shielding anyways. Is your sensor located remotely, or is it right on the PCB with the other components? If remote, get some shielded twisted pair wire or something. If not, just throw the whole system in a shield for now.

 

[witssq]

Many thanks for response.

Although more calculation needed, Sampling speed to need would by 400Hz or more.
Distance between sensor and PCB is 1.5M. But Currently I connected shield cable for signal line, opened
wire for exc line.

The resonance with Fig. A_3_8, is found when flourescent lamp is on, and not found when flourescent lamp is off. Then lamp(220V AC) is connected on the same multi-outlet receptacle with this system.

And, please kindly let me know the details why measurement bandwidth is important in this situation.

Best regards,
SunSung