[SOLVED] SPI bit banging help

[pravin b]

Hello Friends,

I am trying SPI bit banging approach to interface ADC 0832 with AT89C51RD2. I have written following code for same purpose; however I am not able to develop logic for receiving a serial data from ADC 0832 and convert it into readable/usable byte to transmit received byte to PORT 2. Please find the attached code and my hardware simulation in proteus. All type of inputs/help are/is welcome.

Also, when I observe waveforms on oscilloscope at pin dout (P3.2) or MISO, I expect my immediate first bit is to be low (signifies mux settling time as per datasheet) and then 8 bits of data. But in my case I am receiving first bit low (mux settling time), 7 bits of data (I think so) and my 8th bit always remains high (I don’t know why!). Please help me to clear this doubt.

I am here by attaching my code and proteus simulation file & uvision project file for reference.
Thanks & regards,

Pravin B

#include<reg51.h>
sbit clk=P3^0;
sbit din=P3^1;
sbit dout=P3^2;
sbit cs=P3^3;

#define outport P2

void delay(unsigned int);
void adc_init();
void adc_read();

void main()
{
clk=0;
din=0;
dout=1;
outport=0x00;
while(1)
	{
	adc_init();
	adc_read();
	}
}

void delay(unsigned int count)
{
unsigned int i,j;
for(i=0;i<=count;i++)
	for(j=0;j<=1300;j++);
}

void adc_init()
{
clk=0;
cs=1;
din=0;
delay(1);

clk=0;
cs=0;
din=1;
delay(1);

clk=1;
cs=0;
din=1;
delay(1);

clk=0;
cs=0;
din=1;
delay(1);

clk=1;
cs=0;
din=1;
delay(1);

clk=0;
cs=0;
din=0;
delay(1);

clk=1;
cs=0;
din=0;
delay(1);

clk=0;
cs=0;
din=0;
delay(1);



}

void adc_read()
{
int m, byte_in=0x01;
clk=1;
delay(1);
clk=0;
delay(1);

for(m=0;m<=7;m++)
	{
	clk=1;
	cs=0;
	delay(1);
	clk=0;
	delay(1);
	byte_in=byte_in&dout;
	byte_in=byte_in<<1;
	delay(1);
	}
	outport=byte_in;
	cs=1;

}

 

[KlausST]

Hi,

please adjust your scope inputs to "DC" instead of AC.

7 bits low and the LSB high may be a valid 8 bit output. You give 0V into the DAC Ch0. The single LSB high may be because of DAC offset voltage.

Where exactely is your doubt? What did you expect?

Klaus

 

[pravin b]

Thank you Klaus!

My doubts are;
1. My first bit on MISO line has to be '0' always? (mux settling time, as mentioned in datasheet).
2.when I apply 0 Volts to ch0 why did I get my LSB "1"; instead of "0"? (which might due to offset voltage as you said, but since I am using proteus simulator I should get '0' ideally; isnt it?)
3.I have written function adc_read() to read the data received on MISO line; but it seems I don't get any value in variable byte_in. thats why my "outport" is always active low i.e. 0x00. Am I going wrong somewhere in code?

Regards,
Pravin

 

[KlausST]

Hi,

Looking into the datasheet ADC0832-N Timing shows:


MISO at the first three rising clock edge is TRI-STATE. I recommend to use a pullup on MISO to get a vaild HIGH on TRI STATE.
The next bit is LOW
Then 8 bits MSB first.


But review your ssoftware:

* clock low
* CS low
* wait t-set-up
* loop at least 12 bits: wirting three valid bits folowed by don´t cares, read in the same time 3 bits tristate, 1 bit low, 8 bits data
(to be 8 bit SPI compatible i recommend to loop two times 8 bits simulataneously reading and writing data, then code is reusable and a change to HW SPI is easy)

.. loop:
* set output data
* Wait t-setup
* clock high
* wait thold
* read data input
* clock low

Klaus