MCP4725 Odd Behavior - does not update with number directly

[xodballxod]

Seams like a potential safety issue with the i2c communication. The clocks can’t match, or one dominates the i2c bus. In my case, the ADS1115 appears to dominate. If you choose different clock speeds, then the speeds have to be out of phase… and never match on moment to UPD, I’m not sure the best way to get that…

One option is to use prime numbers, but I’m not sure if they could overlap ever…

A Flip-Flop could be used for just 2 i2c devices, assuming you want them sampling at the same rate

But, what happens if you have a lot of i2c devices? Then, you need a new node that XOD doesn’t have in the the CORE, the Select-Pulse Node (but doesn’t work for this)

Select Variadic Out Pulse

For multi function projects that are menu driven could use a node with variadic output of the select, can make one with nth-input style, or one with select style, I’ve got a non-variadic version here, where VAL matches whatever IN (which could just be an integer), then the corresponding pulse pulses out. Would there be a way to make this variadic? [image] Another way to think about it. We would want the NONE to be to the left of the variadic output. [image]

But, that node method doesn’t work! Likely has something to do with the time to write and read-back from the I2C bus… I’m not sure, but not all methods work…

Now, if all the I2C devices want to be read as fast as possible, then you can make a daisy chain, but have to add the DONE pulse node out of the MCP4725, and place a defer. Using BUS instead of wires, naming convention can be “DEV#-out”, and the last I2C DEVICE wired back to the start.

I’m not sure if it would be possible for XOD to handle the I2C networking managing, automatically?