The diodes on the relays are a problem and in your case you can choose the 1N4007 or those inside the ULN2803A, the 1n4148 may also be fine but it is necessary to check the current associated with the relay coil.
[EDIT] I saw that the current in the relay is quite high about 70mA. It is better to use IN4007 or ULN2803A .
Warning the contacts of the relays used in DC can tolerate a current of 5A Max.
Instead those on the power supply if you want suppressor diodes you have to choose the TVS suitable for the supply voltages.
And if I just select the same as the Relay card I should be fine? Right?
Because I use the same relay.
There should also be no problem using the same diode for 12 and 28V.
The diode is not a problem. Let’s not snowball into it too much. The only problem is getting them in the right direction on the PCB.
There are much bigger problems with PCBt, in my opinion, that I don’t really know how to solve.
I see this as causing problems.
Which path will return current try to take? How can I improve this?
Are there any sizing issues since I’m going to PWM? With tracks? Soldering surfaces? Mosfet choice?
For example, someone on an electronics forum says that there may be problems with PWMing the mosfet circuit because I only have 3.3V from the GIPO. But when I’ve tested it’s worked fine, so I don’t know what to believe?
Will I get interference between tracks?
Will analog signal have interference thanks to my design?
What else have I missed?
Will the diodes from the D-Sub (RELAY_CONTROL1) be able to protect an arduino?
Will the diodes let the signal through, is there anything special to think about regarding these diodes? Which ones should you use?
How should I connect pin 10 on ULN chip?
Should I really connect pin 10 on ULN chip?
Absolutely most important should I can I improve the return current path?
The relay coil is powered at 5V, there is no correlation with the other powers. The diodes indicated for relays are fine.
For all these problems you must create at least one driver for dimmers on 2-sided test pcb and take tests.
With mosfets with 3V3 of command signal, RDSOn can become problematic.
You also have high values of the gate piloting resistance (R30, 1K) which reduces switching times, but this is dependent on the frequency of the pwm.
You must also take into account the heat dissipation part for mosfets.
You need to clarify the current limits for the project yourself.
For diodes to be placed on the D-Sub connector, either bat54 or 1n4148 are fine.
They avoid the cotocircuits between the control groups.
For pin10 of ULN2803A has already answered @straubm, connect it to 5V which is the supply voltage of the coil of the relay.
You have to improve the part of the PCB with high currents.
If you mean D13 diodes, it is to be understood what type of circuit is connected to the DIM-Channel connector.
If it is resistive it is not necessary, but if it is inductive it could serve a high current diode and fast switching if the frequency of the pwm is large.(power diode given the currents involved).
You have to build parts of the board to test.
You can view what the circuit that connects to the dim-channel connector is like?
If in some situations you need that some dim-channel control circuits require more current, you realize some with higher current specifications instead the others the dimensions for a lower current.
So you can use smaller components for both relays and semiconductors.
