ADS1278: Reading data in SPI fixed mode

Hello,

Can you take some measurements and verify the signals are getting to the IO pins of the aDS1278?  An oscilloscope or a logic analyzer could be used to verify communications.

Also, from your above list:

J4.13 -> MOSI pin on Rapberry Pi

This is incorrect, and should be connected as shown:

J4.13 -> MISO pin on Rapberry Pi

Regards,
Keith Nicholas
Precision ADC Applications

Hi,

I use multimeter to ensure that these I/O pins are triggered by  raspberry pi.

But I still did not capture any signal from the DOUT pin.

In this state, DOUT pin has no voltage value,

but it has the voltage value when connected to EVM-PDK.

I guess i have something unset but i don't know.

Maybe S1 or S2 should be changed or not ?

Can you please help me to solve it ? Thank you.

Hi, Keith:

Thank you so much for helping out !

I successfully capture data from ADC.

Another small question, how can i control the power on or down of CH1~8 just like ADCPro ?

And can i reset ADC by trigger pins on ADS1278 EVM ?

Thanks again !

Hi, Kieth :

Thank you for your reply.

If i don't make any settings, does it mean that 8 channels are continuously power up ?

So that can i restart the ADC by raising the voltage of the SYNC pin ?

Another question about data format from ADS1278s, i can capture data by SPI TDM Fixed mode now.

I capture 3 bytes data as a piece of data, then i combined this data into a 24-bits in binary and performed 2's complement operation.

Therefore, i receive 24 bytes data (192 bits) as a round of 8 channels of data.

For example, i capture 3 bytes data : [250,250,10]

250 => 11111010 ; 10 => 00001010

So i combined this data into [11111010 11111010 00001010] and got the value [-329206] based on the two's complement.

Further calculated the voltage is -329206 * 2.5 / (2^23-1)= -0.098 .

But sometimes i get unreasonable values.

Are there problems with the way i capture or calculate the data ?

Thanks for your help !

Hello,

There are no configuration registers in the ADS1278.  The channels are ON if the /PWDNx pins are pulled high, and they are OFF if the corresponding channel pin is pulled LOW.  If you always want the channels ON, you can simply tie each of the /PWDN pins to IOVDD.

Your calculations are correct.  It is possible you are getting bit errors when reading the values.  This can easily occur based on the picture of your setup using high speed clocks.  As a test, you could slow down the MCLK and SCLK to 1MHz, and see if the readings are more consistent.

Regards,
Keith

Hi, Keith :

Thanks for your solution and suggestion.

But i still couldn't get the correct data after I changed the setting for SCLK.

My settings and python code are provided below.

(red words represent comments)

################################################################################

#!/usr/bin/env python

import RPi.GPIO as GPIO
import spidev

# SPI setting
spi = spidev.SpiDev()
spi.open(0,0)
spi.max_speed_hz = 7800000 #SCLK frequency
spi.mode = 0b00 #C_POL=0, C_PHA=0
spi.no_cs = True

# get bytes
def adc_1ch(): #1channel
    data = spi.xfer([0x00,0x00,0x00]) #24bits = 3bytes
    adc = (data[0] << 16) + (data[1] << 8) + data[2] #shift register
    if adc > (2**23-1): #2's complement
        adc = adc - 2**24
    voltage = (adc*2.5)/(2**23 - 1)
    return voltage

def adc_8ch(): #8channel
    ch = []
    for i in range(8):
        ch.append(adc_1ch())
    return ch

# GPIO setting
GPIO.setmode(GPIO.BOARD)
pins = [12,16,18,36,38,40,32] #SYNC, MODE*2, FORMAT*3, CLK Select pin
volt = [1,1,0,1,0,0,1] #1=high, 0=low
for p in range(7):
    GPIO.setup(pins[p],GPIO.OUT)
    if volt[p]==1:
        GPIO.output(pins[p],GPIO.HIGH)
    else:
        GPIO.output(pins[p],GPIO.LOW)

GPIO.setup(22,GPIO.IN) #DRDY pin

# main
i = 0 #timer
while True: #loop 100 times
    GPIO.wait_for_edge(22, GPIO.FALLING) #start to capture data when DRDY falling down
    print(adc_8ch()) #capture data
    i = i + 1
    if i == 100: #timer stop
        break

GPIO.cleanup()

################################################################################

Are there any errors in my settings or logic ?

Thank you so much again.

Hello,

Your code looks correct, but I cannot exactly follow it.

Use a DMM and check all of the configuration pins to confirm you have the correct logic values.  (SYNC, MODE, FORMAT, CLKDIV).

If possible, find a logic analyzer or oscilloscope and capture /DRDY, SCLK, DOUT so we can verify all timing requirements are met.

Regards,
Keith

Hi, Keith :

Thanks for your help.

I don't know how to set CIKDIV on ADS1278EVM.

Could you help me to set CIKDIV ?

Nonetheless, I used DMM to detect that the CIKDIV on ADS1278s chip has voltage.

And i saw the following statement from the document :

In the SPI format, DRDY goes high as soon as SYNC is taken low; see Figure 73. After SYNC is returned high, DRDY stays high while the digital filter is settling. Once valid data are ready for retrieval, DRDY goes low.

Is that mean i shouldn't keep setting pin SYNC voltage high ?

And is there any specific command to be passed to the ADC to get the correct value?

I just send 24 times 0x00 to ADC now.

I'm sorry I don't have a logic analyzer or oscilloscope to capture the signal graph...

Thank you again.

Hello,

Yes, you can use S1 AND S2 to configure the device when not using the MMB0 board.

S1 -> turn on all

NO, In order to turn all channels ON, all S1 switches must be set to OFF.  (These are negative logic.)

S2 -> M0,CLKDIV,F0 turn on, others still turn off

YES, these switches will configure the mode and format settings.  For the above switch settings:

MODE=01 (High Resolution), CLKDIV=1, Output Data rate fDATA=52734sps when fCLK=27MHz.

FORMAT=001, SPI mode, TDM (all data output on DOUT1), Fixed position

S6 -> turn to SPI

Yes, set S6 jumpers to SPI mode per below:

All other settings look correct to convert readings on Channel 0.

The caution statement only applies when the ADS1278EVM is plugged into the MMB0 board.  Since you are only using the MMB0 for the power connections, you can set the switches S1 and S2 as needed.

Also, you stated that you are using a photodiode.  The inputs of the ADS1278EVM need a low impedance source voltage.  The output of a photodiode is a high impedance current source.  You will need an additional amplifier circuit for the photodiode, before connection to the ADS1278EVM.

Please refer to the following cookbook circuit for an example.

https://www.ti.com/lit/an/sboa220a/sboa220a.pdf

Regards,
Keith

Hi, Keith :

I tried to turn off only channel 5 by setting S1 switch,

and i still used SPI TDM fixed mode and high resolution mode.

However, i got sometimes the error data like this :

I guess the data I read is shifted, but the shift is random.

Do you have any experience or methods to solve this situation ?

Thanks for your reply again.

Hello,

If you do not read all data out of the ADS1278 before the next data conversion results are ready (the next DRDY low), then you could get corrupted data.  It is hard to guess without seeing the timing waveforms for DRDY, CLK, DOUT.

You can slow the conversion rate down by using a much slower CLK frequency, but this will require an external clock generator, or changing the oscillator on the ADS1278EVM board to a different value.

Regards,
Keith

Hi, Keith :

If i want to replace the 27MHz oscillator on-board with an external clock,

i guess i must change some settings on ADS1278EVM.

But i'm not sure if my configuration is correct, the following:

JP4 => Open

JP5 => Short

J4,17 => Connect external oscillator

J4,18 => Connect ground from external oscillator

J4,19 => don't connect anything

Is that right ?

Thanks for your help very much.

Hello,

Yes, you are correct.

However, the inputs for U13 should not be floating.

Please connect J4,19=>Ground

Regards,
Keith