when running calibration on the ecpix5, what is the uButton?

[matpalm]

more questions sorry seb :)

i can run the calibration process up until the last step OUTPUT -5V...

at this point i'm meant to hold the ubutton to switch to -5V but not sure which it is?

there are three buttons on the ecpix5...

SW1 RESET -> resets the board ( and the cal.py aborts with termios.error: (5, 'Input/output error'))
SW3 GSRN -> does nothing. if i hold it down and hit 'n' i just see the raw ADC values for the +ve case ( see screenshot )
SW2 PROGRAM -> makes the cal.py process restart

i traced the code for top.sv and can see how the button corresponds to RESET_BUTTON and then in pinmap i see how RESET_BUTTON corresponds to site "N5" but then i'm a bit and can't work out how to confirm how N5 related to one of these buttons?

[matpalm]

for reference, in case i'm naming things wrong...

[matpalm]

hmmm. i need to get my head around the calibration, maybe i need something more specific anyways...

i don't want to hijack my own calibration thread but i don't want COUNT_PER_VOLT = 4000

in my project i've been working with FP config of 4 bits for the integer portion and 12 bits for the fractional part with the module working with inputs/outputs mapped for 8V to the range (-1, 1)

for example, 5V => 0.61751866 => 0000.100111100001
and -5V => -0.61044693 => 1111.011000111100

so i need to work out how to do this mapping, which is more than the calibration might provide... the exact way to do this hasn't quite clicked in my brain yet :)

[vk2seb]

Hey there @matpalm, sorry for the confusion on the uButton! That is just the name of the extra pushbutton on the iCEbreaker. The logic in top.sv is a bit convoluted as when we're in OUTPUT_CALIBRATION mode (define enabled), we disconnect RESET_BUTTON from sysmgr_instance (top-level system PLL reset that resets all state machines) and instead connect it here such that it modifies the output DAC values. Unfortunately the button has to share both functions as on some boards (i.e ICEbreaker) there is only 1 button. So feel free to reassign the button in the top-level lpf to whichever input works, or even to a pin on a different PMOD. If you don't have a button you could even hard-code the output to +5V or -5V and reprogram the board while the calibration tool is running (I haven't tried doing things this way but it should work). It seems you are correct that the button I assigned shares some other reset functions.

The calibration tool is mostly designed to zero out the DC biases present in the CODEC from the factory. If you wish to use a different fixed-point configuration than the standard 4000 counts/volt (i.e 14.2 fixed point mV), then the relevant places to start hacking are probably:

• The ZERO and MULTIPLY stages of cal.sv -> you can see we multiply by a constant from the calibration memory and shift right 10 bits. You can change this if you want to change the meaning and scaling of the raw ADC -> fixed point counts out of cal.sv
• The relevant constants in the calibration script
• The test harness for the calibration module might be useful to test your changes

At some point I'd like to clean this up and make it easier to change the fixed-point format globally, but hopefully the above pointers are enough for you to start hacking :)

[matpalm]

Awesome! Thanks for the verbose info, am new to a bunch of this so it really helps!