Go Down

Topic: General PCB Check (Read 698 times) previous topic - next topic

450nick

Correct, this was one of the things I wasn't sure of. There are several types of inputs I have here:

- 12v analogue; these I've run through quite high division voltage dividers with the assumption being that the division is enough to counter voltage spikes

- 5v analogue; these will take a 5v reference from the self contained 5v power supply on the PCB and return a 5v voltage - since this is a closed loop, my thought was that this would only need a resistor.

- 12v digital; these are all run through opto-couplers to decouple them fro the arduinos

- 5v digital, I think there is only 1 and it just switches the arduino pin to ground

- power; the 12v feed comes from a motorsport PDU system (this one) which provides reverse voltage protection for all circuits, and you can set clamps on each channel to limit current (not sure about voltage, need to check) so there is an element of protection upstream of this unit. The 5 volt feed uses one of these, and I think 2A should be plenty - it has caps built into it so the datasheet claims no external ones are needed.


So presently this is how I've set it up, but I'm definitely not an expert so if you think I should have some caps in there, please do say and I'd love to hear your comments.

MorganS

Always consider wires entering your box as untrustworthy. Even if the upstream device has current limits, that does not prevent that wire from being dragged across the battery terminals during maintenance.

Usually the voltage divider on 12V inputs is sufficient for most automotive wiring screwups. But static discharges are thousands of volts. ESD protection via TVS diodes is pretty easy these days.

Look up "load dump". That is the most violent electrical event that can occur in a vehicle.
"The problem is in the code you didn't post."

WattsThat

IMO, the 10k series led resistors for the PC817's are marginal. Given the current transfer ratio of those parts, I'd consider lower value series resistors for better noise immunity. They may work on the bench but add some noise and a higher ambient and you may have issues.

I'd also be cautious of relying on the internal pull up resistors. They are relatively high values and not the best for noise immunity. In addition, there is the question of minimum load requirements when you're replacing filament lamps with solid state inputs. Lots of unknowns there, depending upon the age of the vehicle.
Vacuum tube guy in a solid state world

450nick

#18
Nov 18, 2019, 06:11 am Last Edit: Nov 18, 2019, 02:19 pm by 450nick
Always consider wires entering your box as untrustworthy. Even if the upstream device has current limits, that does not prevent that wire from being dragged across the battery terminals during maintenance.

Usually the voltage divider on 12V inputs is sufficient for most automotive wiring screwups. But static discharges are thousands of volts. ESD protection via TVS diodes is pretty easy these days.

Look up "load dump". That is the most violent electrical event that can occur in a vehicle.
Thanks for the input, would you recommend a TVS diode on all inputs then to be sure? I will read up but is there a particularly good PN you'd recommend for this application?


IMO, the 10k series led resistors for the PC817's are marginal. Given the current transfer ratio of those parts, I'd consider lower value series resistors for better noise immunity. They may work on the bench but add some noise and a higher ambient and you may have issues.

I'd also be cautious of relying on the internal pull up resistors. They are relatively high values and not the best for noise immunity. In addition, there is the question of minimum load requirements when you're replacing filament lamps with solid state inputs. Lots of unknowns there, depending upon the age of the vehicle.
Thanks for your input too! Having read the datasheet, I've now changed the values of the input resistor to 470 ohms - the PC817 can handle max forward current of 1A and min of 50mA so 470 should do the job I think...

When you mention not relying on the internal pullup resistors, what other method would you recommend?

With regard to the minimum load piece, the wiring loom is going to be completely replaced so nothing will be existing. The power distribution will come from the units mentioned above which will replace all relays and fuses, and the engine/alternator will be a 2019 Chevrolet LT4 motor so very new design and hopefully modern levels of electrical noise if such a thing exists...

Go Up