Mcp4725 voltage conversion with ads1115

Projects Programming
1

I want to convert ads1115 .16 bit input to mcp4725 12bit voltage with ardunio but the codes are working.
but it gives 2.5v output with max 0.
I'm not sure if I'm wrong with 0-5v output codes?
can you help me . or is this possible.

At the same time, I tried to understand the voltage responses using 16bit and 10bit inputs.. I couldn't see much difference.

I uploaded these codes to arduino pro mini and ran

#include "ADS1X15.h"
ADS1115 ADS(0x48);
#include<Wire.h>                   //Include kutuphane ı2 data Wire
#define MCP4725 0X60
 unsigned long  buffer[3]; 

 
byte outc= 7 ;
unsigned long Voltage = 0;
unsigned long Voltage2 = 0;
unsigned int compr1 = A1; // karsılastırma :
unsigned int compr2=A0; //  karsılastırma :
byte outM=8; // cıkış  a0 ....a3 devre :


void setup(void)
{

Wire.begin(0x60);
Wire.setClock (400000);  
  
ADS.setDataRate(5);  
ADS.setMode(1); 
ADS.setGain(7);
ADS.begin();
pinMode(outc, OUTPUT);
pinMode  ( outM, OUTPUT );  // s :
pinMode( compr1 , INPUT);   //  modlar :
pinMode( compr2 , INPUT);   //  modlar :

}

void loop(void)
{

unsigned long adc_A0, adc_A1;

adc_A0 = ADS.readADC(0);
adc_A1 = ADS.readADC(1);
Voltage = (adc_A0 * 0.0078125)/1000;
Voltage2 = (adc_A1 * 0.0078125)/1000;
  if ( Voltage <= Voltage2 )
   {digitalWrite (outc ,1); } 
      else 
  {digitalWrite (outc ,0);}
delay(10);
{
   unsigned long compr1 ,compr2 ; compr1 = analogRead ( A0) ;  compr2 = analogRead ( A1) ;
  if ((compr1 <= compr2 ))
               {digitalWrite (outM ,1); } 
                   else 
               {digitalWrite (outM ,0);} 

}

{  adc_A0 = ADS.readADC(0); adc_A0  = map(  adc_A0, 0 ,65536 ,0,4095 )  ;
  buffer[0] = 0b01000000; 
  buffer[1] =  adc_A0 >>4;
  buffer[2]= adc_A0 << 4;             
  Wire.beginTransmission (0x60);        
  Wire.write(buffer[0]);            
  Wire.write(buffer[1]);           
  Wire.write(buffer[2]);           
  Wire.endTransmission(); 
}

}
2

I wanted to read the max output 0-5v. but I am reading 0-2.5v with the map function, is the map suitable for the correct calculation?

/thousand; I don't know exactly what it means. I think he's talking about the reference Voltage. is this true ?

3

Thats one problem. The ADS1115 is a 16bit ADC but one bit is reserved for the sign, so you need to map it to 2^15 instead of 2^16, i.e 0, 32767.

Also the 0.0078125 factor is based on setting the gain range to 256mV. You need to set the gain to the 6.144 volt range to read 0 to 5V. As well as factor in that 2^15 bits represents 0V to 6.144V not 5V.

1 Like
4

thanks robert I noticed that the gain factor gets bigger when upscaling analog signals. that is, I made the gain 7 and increased it 16 times and realized that the effect was not straight logic. We use the voltage as a buffer for its input. I think it has an effect on resolution. but 10bit 4.5mv still reads range sensitive
Using ads1115 does not seem advantageous to me. Is it possible to read ads1115 at 0.010mv intervals?
(adc_A0 * 0.0078125)/1000; What is the last thousand?

Can you finally suggest a Gain tuned IC that will meet my expectations?
The popular one may even have 1000 wins. I was looking for an I2C controlled instrumentation amplifier but found the ltc6915 expensive. I'm in the testing phase, I haven't been able to decipher the serial communication codes yet.

5

This makes no sense to me. There is no gain of 7 for an ads1115. There are gains of 0, 1,2,4,8 and 16 covering the range from 6.144V full scale range at 0 down to 256mV full scale range at gain 16. I would guess but don't know depending on the library you used since there is no logical option for a gain of "7" that you are probably just defaulting to a gain of 0.

By "upscaling" and "increasing 16 times" do you mean in the mapping function? That is just a bit operation, it does not matter where the number comes from, it's purely mathematical. If you do it correctly from 2^15 to 2^12 that is going reduce your resolution by a factor of 8, but it will not result in non-linearity in terms of gain.

Reading the ADS1115 at intervals of 0.010mV? That's 10 uV. The advertised resolution at 16 gain is at best 7.8125 uV/bit. As soon as you map it to 12 bit your best resolution is now only 62.5 uV/bit, so does it matter? That's not even addressing accuracy and stability at that level.

Voltage(V) adc_A0) =( 256mV/2^15) /(1000mV/V) is where those numbers come from. Your code is assuming you are at 16 gain i.e a full scale range of 256 mV, but I don't think you set the gain to 16 successfully, if that is even what you want.

I don't have any other recommendations without any knowledge of what you are trying to do or what your design looks like.

6

Posting an annotated schematic would help us help you. I and many others do not read frizzy.

8

I want to prepare a project where I use the ingaas sensor for preliminary work.
this sensor detects heat from 150 celsius to 1000 celsius and when i input 5v and read output from 0.10mv to max 200mv. I have tested
I have used various instrumentation amplifiers
I got high gains in Ina188,ina118,ina129,ad820 opamap, but for some reason I couldn't run it very stable due to voltage deviations. I may be getting a fake product from Aliexpres. I tried it with original products. more or less the same
I set the ref inputs to zero and tested, there was a constantly fluctuating voltage. I keep trying different boosters to remove these effects from there.

9

20230222_194712
20230222_194712420×700 131 KB

10

20230222_194605
20230222_194605700×420 98.6 KB

11

20230222_211731
20230222_211731800×480 76.5 KB

12

A0 and a1 work as comparators. You can suggest a more precise comparison. a warning led and npn transistor sending pwm signal
A0 and a1 work as comparators. You can suggest a more precise comparison. warning led and npn transistor sending pwm signal..
I compiled ads1115 and promini Voltage test code

#include "ADS1X15.h"
ADS1115 ADS(0x48);
 byte outc= 7 ;
float Voltage = 0;
float  Voltage2 = 0;
unsigned int compr1 = A1; // karsılastırma :
unsigned int compr2=A0; //  karsılastırma :
byte outM=8; // cıkış  a0 ....a3 devre :


void setup(void)
{
ADS.setDataRate(7);  
ADS.setMode(0); 
ADS.setGain(0);
ADS.begin();
pinMode(outc, OUTPUT);
pinMode  ( outM, OUTPUT );  // s :
pinMode( compr1 , INPUT);   //  modlar :
pinMode( compr2 , INPUT);   //  modlar :
 ADS.readADC(1);
ADS.readADC(0);
 
}

void loop(void)
{

int16_t adc_A0, adc_A1;

adc_A0 = ADS.readADC(0);
adc_A1 = ADS.readADC(1);
Voltage = (adc_A0 *   0.1875)/1000;
Voltage2 = (adc_A1 *   0.1875)/1000;
  if ((Voltage  <= Voltage2 ))
   {digitalWrite (outc ,0); } 
      else
  {digitalWrite (outc ,1);}          
delay(10);
{
   unsigned int compr1 ,compr2 ; compr1 = analogRead ( A0) ;  compr2 = analogRead ( A1) ;
  if ((compr1 <= compr2 ))
               {digitalWrite (outM ,0); } 
                   else 
               {digitalWrite (outM ,1);} 
}

}
13

Since your comparators use the same reference as the A/D have you considered using the A/D converters to read the . You could do it with software and a transistor as is. The diode would go directly to A0 and A1 can stay the same as your trip point setting.

14

In one place you say you read output 0f 0.1 mV to 200 mV. In your diagram it says 0.01 mV to 100 mV. Which is it?

The LM2662 is going to be noisy for low level signals. Try disconnecting that and seeing if that is a source of noise. You may need a bipolar supply instead of a charge pump. Or use Vref to create a virtual ground instead of trying to swing to a negative rail. Also if you are working in the low end of the sensor range a breadboard is going to be noisy in general. Also you are sampling the ADS1115 at maximum rate. Try a slower rate. Ground plane, PCB design, sensor housing will all be critical factors.

15

thanks . I used a Schottky diode. Since the internal resistance is small, it does not affect the voltage much. my intended use ( -5v) was successful in preventing negative voltage leakage. I went to the negative and saw that the promin was not affected.
I prevented the promin from being damaged.
I'll try without using the -5v lm2662 and with an additional reference source. Diode goes directly to A0 without using a diode
. I don't understand where to add the transistor
I'm new to software...

16

robert thanks
yes, i'm aware of the noise. I should see my pre-tests and bugs here to prepare the pcb.
I don't have very advanced tools, but I do have good measuring tools.
the sensor says 0.01 mV to 100 mV. Which one? according to temperature
Varies continuously from 150 degrees Celsius to 1000 degrees Celsius
For example, I am reading 90mv output at 800 satigrade, of course, I am reading from 50cm distance, this changes at different temperatures.
I read 0.01 mv output at 150 degrees, I showed the product at 1000 degrees, I read max 100mv, 50cm distance will be read
A hot product passes in front of the sensor, for example, it has exceeded 300 degrees Celsius. Suppose there will be a reverse voltage. By increasing the value, it will control the led or relay as an output, check whether the product is on or command a robot to control hot products. various hot products. There will be different voltage conversions. For example. the same lm35-like software is also adapted for low temperatures.

I will try to create a reference voltage. ads1115 differential reading can be done eg sensor voltage.
ads A0. DIFFERENCE . A1 . If it can read the REF voltage, I would try that. i can get rid of lm2662
I set another additional reference voltage129 the voltage did not start from 0.
I read output 1.5v reference 1.6 starting from reference voltage
How should I try for single source reads? Can you help

17

Let's step back a little because we clearly have a language barrier.

Do some basic gut level engineering. Call your temp range 0-1000C for easy do in your head numbers. One of the biggest mistakes people make is doing math on paper without considering the realities, so keep it gut level for a proof of concept design...

You are asking your sensor to read 0.01 to 100mV. That's a range of 10000 uV.

That's already really challenging.

Next assume the consumer of your system will not settle for less than 1 Degree C resolution. Seriously, in terms of a robot kicking out a part that is at 850C vs 860C, does that matter? Doubtful. Clients are not rational so assume 1 degree resolution is something they can hold on to.

That means your entire system has to measure and operate upon 10000uV/0-1000C = 10uV/1C. Wow.

You will maybe never engineer that. You certainly will not protoype it on breadboards. Your objective is to build a temp based QC system, not a high resolution magnetic sector mass spectrometer. Don't try.

Instead consider multiple sensors at appropriate resolutions to cover that range. You are using a microcontroller so you haver very fine grained control over your system logic. Exercise coarser grain on your analog components. Use a bank of sensors, keep them in the 1+ mV output range and you may not even need to deal with the complexities of adding in op-amps.

An ADS1115, will give you solid results if you are trying to measure 1-2 mV absolute voltage from stable ground.

19

super explanation really, although there is a translation problem in my articles, I noticed it.
I don't want to read 0-150 degrees Celsius. Even 150 and max 900 degrees Celsius are sufficient.
I need to test the output of the sensor as 5mv or 1mv max by changing the resistor.
maybe the system will need 24bit adc. What is required of me is not a very precise temperature measurement! .
Roughly enough. I am trying to read as seen in the example image above.
The actual working press is pictured below,
Smaller materials arrive at the front of the machine.
According to the temperature measurement, it transfers the material to the press with the robot arm. It gives output in a certain temperature range and sends it to 1-0 plc.

 I got good results with ads1115. Can you share the 0-5v example conversion code with pwm and mcp4725. similar to. sufficient !

example adc_A0 = map( adc_A0, 0 ,65536 ,0.4095 )
and pwm flip . adc_A0 = map( adc_A0, 0 ,65536 ,0,255 )

thanks

20

Ok a difficult assignment.

4 Wire Pressure Transducer with max output voltage 45mV
21

I found two important resources .ads1115 now I understand ! . It's really complicated to write code, I want to add a lot of features and functions.
For example rotary encoder instead of potentiometer. Hence the voltage noises. I want to get rid of such effects.
It will take a lot of time to prepare a fully equipped project.

22

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.