Only positive/negative numbers when using AMP in ADC

[nadabro]

Hi.

I think this is a very simple issue, but my knowledge is very limited, so here it goes my issue:

-I´m using ADC to connected to a sensor, with the sensor connected directly to the ADC inputs, i´ve positive and negative values.
-I´ve added a OPA227 OP-AMP to do some tests with the sensor, and connected the output of the amp to one ADC input, the second input of that ADC channel i´ve connected to GND (also i tried to +2.5V).

But i only have positive or negative values (depending the INPUT i use in ADC), i tried some types of calibration on the ADC (ADS1211) but no luck.

What can i do to solve this?

Thanks.

 

[mgate]

I think that after signal re-scale (amplification/attenuation), you need to shit the common mode voltage to match the ADC mid scale voltage.. By other words, you need to translate the bipolar signal to unipolar signal... https://masteringelectronicsdesign.com/design-a-bipolar-to-unipolar-converter/

---------- Post added at 01:45 ---------- Previous post was at 01:37 ----------

check also:
https://www.edaboard.com/threads/245197/#post1049199

---------- Post added at 01:49 ---------- Previous post was at 01:45 ----------

check also:
https://www.edaboard.com/threads/249191/#post1066105

 

[nadabro]

It makes sense on "converting" the bipolar signal from the amp to match the unipolar signal in the ADC. But before try some of the schematics you gave, i looked in the ADC datasheet and it looks like it has a function (Vbias) that should solve my issue:

The VBIAS output voltage is dependent on the reference input (REFIN) voltage and is approximately 1.33 times as great.
This output is used to bias input signals such that bipolar signals with spans of greater than 5V can be scaled to match the input range of the ADS1210/11. Figure 12 shows a connection diagram which will allow the ADS1210/11 to accept a ±10V input signal (40V full-scale range).

So using "Vbias" in the ADC should work?

Thanks (Obrigado! Tb sou de Portugal :smile

 

[mgate]

Yes, it should work...

I suggest you to post the the interfacing circuit diagram and source code, so that we can be able to help you on this

 

[nadabro]

I´m testing it in a breadboard, following the datasheet pag18 circuit.

But my main issue is to change the code in the microcontroller (Arduino) to activate VBIAS..

The code i use can be found here (**broken link removed**), the part to write to the ADC:

void ADSx::CMRwrite(byte channel,byte mode,byte gain,byte TMR,int DR)
{
  byte Byte[3];
  
  int _gain;
  switch (gain) {
    case 2: _gain = B001; break;
    case 4: _gain = B010; break;
    case 8: _gain = B011; break;
    case 16: _gain= B100; break;
    default: _gain = 0;
  }
  Byte[0] = mode<<5 | _gain<<2 | channel-1;
  
  byte _TMR;
  switch (TMR) {
    case 2: _TMR = B001; break;
    case 4: _TMR = B010; break;
    case 8: _TMR = B011; break;
    case 16: _TMR= B100; break;
    default: _TMR = 0;
  }
  Byte[1] = _TMR<<5 | DR>>8;
  Byte[2] = DR;
  write(B0101,3,Byte);
}

In datasheet
The BIAS bit controls the VBIAS output state—either on (1.33 • REFIN) or off (disabled), as follows:
0 disabled (default)
1 enabled

But i dont understand how i´m going to register the bit that activate Vbias.

Any help i appreciate.

Thanks.

 

[mgate]

Well,

looks like pretty much forward,

According to page 19 of the datasheet, to set the VBias on you need to set BIAS on Command Register (CMR) , witch is at bit 7 of CMR 's byte 3.

The CMDwrite routine that you are using is writing only byte2, byte1 and byte0, as it starts on the byte 2 address (b0101)

Now I have two suggestions... either you recode the CMDwrite routine to include bias handling, witch means write the four command bytes, and considering command byte 3 default values, it would be something like this:

---

Code C - [expand]
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
void ADSx::CMRwrite(byte channel,byte mode,byte gain,byte TMR,int DR, byte vBIAS)  // vBIAS =1 => ON;vBIAS =0 => OFF
{
  byte Byte[4];
 
  Byte[0] =  ((vBIAS & 0x01) << 7) | 0x40;  // Command Byte3
 
  int _gain;
  switch (gain) {
    case 2: _gain = B001; break;
    case 4: _gain = B010; break;
    case 8: _gain = B011; break;
    case 16: _gain= B100; break;
    default: _gain = 0;
  }
  Byte[1] = mode<<5 | _gain<<2 | channel-1; // Command Byte2
  
  byte _TMR;
  switch (TMR) {
    case 2: _TMR = B001; break;
    case 4: _TMR = B010; break;
    case 8: _TMR = B011; break;
    case 16: _TMR= B100; break;
    default: _TMR = 0;
  }
  Byte[2] = _TMR<<5 | DR>>8; // Command Byte1
  Byte[3] = DR;                      // Command Byte0
  
  write(B0100,4,Byte);  // B0100  is the address for command byte 3 
}

---

or, since you are not performing any write operation to command byte3 in your source code, could code a sub just for control the Vbias, for example:

---

Code C - [expand]
1
2
3
4
5
6
7
8
void ADSx::CMR_bias_control(byte vBIAS)  // vBIAS =1 => ON;vBIAS =0 => OFF
{
  byte Byte[1];
 
  Byte[0] =  ((vBIAS & 0x01) << 7) | 0x40;  // Command Byte3
 
  write(B0100,1,Byte);  // B0100  is the address for command byte 3 
}

---

note: I don't know if the adc allows a write operation only for this byte, but I couldn't found anything saying that we can't do this ...

Regards,
Ricardo

 

[nadabro]

Hi.

The first suggestion works great! It allows to enable/disable Vbias, and i can measure voltage (1.76?) from Vbias pin

But unfortunately did not resolve my issue..i followed the schematic on page 18. But it still gives positive numbers, it stables at -4000 (example) but i only get values up to 32768 (16bit).

So my only solution is to try one of the schematics you gave above?

Thanks

 

[mgate]

Post your reading code