Heart rate sensor MAX30102

[astanapane]

i've done many tries. I just want to see if the red LED inside the sensor is alive. At least by sending a command to activate it.

This is as from the pseudo code the first step. I really do not know if that is the right way to send Device ID + write mode
Then send the FIFO_WR_PTR (i still dont know what are the bits needs to be configured)
REPEAT START

addr = $04   ; FIFO WRITE POINTER REG ADDR IS 0X04
i2cwrite  SDA,scl,contwr
i2cwrite  SDA,scl,contwr,addr
i2cwrite  SDA,scl,contwr
i2cwrite  SDA,scl,contwr,addr

Apart from pseudo code in the datasheet, Next i tried to send mode for the heart rate mode only

addr = $09   ' mode configuration 
i2cwrite  SDA,scl,contwr,addr,[modled],error ' ModLED = $02 as per data sheet heart rate mode only

In the end i tried to send the led pulse amplitute. 

addr = $0C   ' led pulse amplitute
i2cwrite  SDA,scl,contwr,addr,[led1_pa],error' LED1_PA = $3F  , 12.6 mA

no luck......keep walking.

[richard]

given that you can buy these for < $20us delivered to your door i doubt its worth while making them other than as an intellectual exercise
https://au.banggood.com/BOXYM-oFit-2...r_warehouse=CN
i have tried the sparkfun arduino demos and was not particularly impressed with accuracy or stability of readings obtained
{maybe my heart is inaccurate and unstable}
the raspberrypi code goes about it quite differently&nbsp; with some dsp done on data collected in a fixed sampling period [not tried it].
i have no experience with these things , i don't know if reverse engineering the arduino code is even worthwhile
but no matter how you go about it the sort of digital signal processing needed will be lots and lots of hard work with pbp
do you have a plan ? have you established optimal device settings to get the best readings ? how much data is needed and from where ?
reading and writing to the device is child's play, setting it up to get meaningful data and being able to extract information from that data is another story

[Ioannis]

I double what Richard pointed.

Ioannis

[astanapane]

Thanks both for your kind reply.

I will find the way. I know it is quite difficult and options are limited to use Arduino.

[richard]

i tried the arduino code again with the sensor attached to my finger with a rubber band , it works a bit better that way
do you have a stl model for an enclosure ? . its difficult to make any sense of wildly unstable readings

i'm thinking a ESP8285 M3 board would make a better platform and allow use of arduino libs if they can be proven to be adequate
ESP8285's these cut down esp8266's are very cheap and still have more processing power than a bucket full of pic chips they are quite happy to run on two aaa batteries and consume little energy if wifi not engaged

[astanapane]

Hi Richard,

once i have all the components (including the arduino board or ESP8285) then i will be ready to design the 3D cad model and printed.

I will order some arduino and ESP and check them out.

[Ioannis]

Use of such finger sensors need to have your finger cleaned from oil with alcohol (isoprop. alcohol or similar).

It is necessary for the sensors to keep working with consistency.

Ioannis