02-15-2019, 04:56 PM
That says it's hardware related not code and because you see it work a little while shows the code does what it should. The I2C bus is definitely dropping out which is why the clock goes to 2009. I'm curious, what is the length of cable between controller and Atlas circuits? Are your connections soldered or header pin cables?
I know it worked with v4.2 but the problem you have is expected with gen2 hardware but shouldn't be a problem with gen3 because gen3 has an I2C buffer which increases the capacitive load of the I2C bus allowing for longer cable and more devices. When I was using gen2 without the buffer I could only use 1ft of cable between an I2C device and the controller reliably. If I used 6ft of cable it never worked. Overtime I've also noticed different results with gen2 and I2C, for example I've never been able to connect more than 2 devices using 1 cable at 1ft reliably. So the fact some of you could run more is honestly a surprise. Gen2 has nothing special for I2C so it's what the Arduino delivers, I believe the max capactive load for an Arduino is 400pf, any traces, connectors, wire, devices and other factors will add up to a load and it must not exceed this limit. By default there are already 3 devices on the bus, clock, LED driver and eeprom. If the load exceeds the limit the I2C will lock up and at some point it becomes unreliable because it's nearing it's limit and possible other things connected could add noise to the I2C bus causing a surge which knocks it out until the Arduino is reset.
Once I decided to go with I2C power bars this all became very clear to me and I really realized how touchy the I2C bus can be. I search and tried but had lots of problems so I actually hired someone on a website to help me get it worked out. In the end I used the PCA9507 I2C buffer which increases the capactive load to 1800pf from 400pf and has made a world of difference.
Now with all that said I'm not trying to sell you gen3 as it should work again but you might have to change cable length, I'm just trying to explain. Whatever the setup for just the ph circuit is on the edge and why another device knocks it out completely.
I know it worked with v4.2 but the problem you have is expected with gen2 hardware but shouldn't be a problem with gen3 because gen3 has an I2C buffer which increases the capacitive load of the I2C bus allowing for longer cable and more devices. When I was using gen2 without the buffer I could only use 1ft of cable between an I2C device and the controller reliably. If I used 6ft of cable it never worked. Overtime I've also noticed different results with gen2 and I2C, for example I've never been able to connect more than 2 devices using 1 cable at 1ft reliably. So the fact some of you could run more is honestly a surprise. Gen2 has nothing special for I2C so it's what the Arduino delivers, I believe the max capactive load for an Arduino is 400pf, any traces, connectors, wire, devices and other factors will add up to a load and it must not exceed this limit. By default there are already 3 devices on the bus, clock, LED driver and eeprom. If the load exceeds the limit the I2C will lock up and at some point it becomes unreliable because it's nearing it's limit and possible other things connected could add noise to the I2C bus causing a surge which knocks it out until the Arduino is reset.
Once I decided to go with I2C power bars this all became very clear to me and I really realized how touchy the I2C bus can be. I search and tried but had lots of problems so I actually hired someone on a website to help me get it worked out. In the end I used the PCA9507 I2C buffer which increases the capactive load to 1800pf from 400pf and has made a world of difference.
Now with all that said I'm not trying to sell you gen3 as it should work again but you might have to change cable length, I'm just trying to explain. Whatever the setup for just the ph circuit is on the edge and why another device knocks it out completely.
