Digital outputs automated in a MEGA

[Billduck1302]

I need to sample 14 voltages on a Lithium battery pack. I am using a relay board to address individual cells. For example, to get the Cell 1 voltage, I send "0" to a relay board pin, and "0" to another relay board pin. I note that the "output" command icon only allows the selection of a pin number (eg D22) and the value, 0 or 1.
It would be nice if I could have the pin numbers in an array and choose which element of the array to populate the "output" command icon pin #, such as "D22".
Thanks in advance.

[mnfisher]

It's certainly possibly, if maybe cheating slightly...

First note that you can define pins in properties (as s single digital pin) and assign values in a calculation block

Pin = 1 to set pin high for example

Take a look at the code generated - i think it is set_pin('D', 2, 1); for d2 set highfor example.

Then use something like - on my phone so typing for is a pain...

Have an array of 2 character strings ("d2”, ”d3”...}

Then port is str[n][0] and pin is str [n][1] - '0'

The in a c block do set_pin(FCL_PORT, FCL_PIN, 1);

i will post on FC tomorrow...

Martin

[mnfisher]

Hi Bill,

Okay - a simple example - I used an Arduino UNO - but it will work without change on a Mega. You'll need to add more ports to the 'switch' statement in the SetPins macro. It only handles ports B, C and D here.

Here I show toggling a pin set in properties (just once!) and then an array of three members toggled repeatedly.

Note that a 'neater' (ie quicker) way to do this would be to calculate the port address from the name ('B','C' etc) and write directly to the MCU registers. This would be quicker than using the switch statement - I do this here viewtopic.php?t=259&hilit=pullups - it uses the fact that the difference in address for each port is a constant.

Martin

[Billduck1302]

I can not say that I follow your suggestions. It is a little advanced for me.
But I will attach what I have come up with. One program samples 7 cells.
And places the float voltages into an array. I have not tested it yet. The second
program - sorting - sorts the values in the array.

[mnfisher]

Will take a look.....

[Billduck1302]

This is a sorting algorithm. It works for 8 elements in the array, but not 8+.
I need 14 elements.

[mnfisher]

I posted a bubble sort many moons ago (https://www.flowcode.co.uk/mmforums/vie ... ino#p88632) - this sorted 'ints' but would be easy to change to floats
I've also posted a quicksort - though it is probably overkill for 14 numbers (https://www.flowcode.co.uk/mmforums/vie ... rt#p104653)

Your algorithm probably fails because it goes through the loop a set number of times (5 x length + 4) - what it should really do is repeat until there are no swaps made (have a flag (Boolean) 'sorted' - initially true and set it to false if you make a swap and loop until 'sorted'....

In pseudocode...

Code: Select all

loop 
    sorted = true
    for (.i = 0 to  length - 1)
        if swap needed then
            sorted = false
            swap(array[.i], array[.i + 1])
        end if
    end  for
until sorted

The sample data you have - the sort is attempting to reverse the data - which is the worst case for bubble sort. I haven't tested it - but it probably needs more passes (it's not the most efficient sorting algorithm)

Martin