Decoupling capacitors behavior on power-off

Hello, I think that once the power has been disconnected the capacitors would drain quite quickly against the load of the circuit (and this could keep the circuit operational for a few seconds after switch off whilst the capacitor drains). So unless your circuit has very small current drain then I think it should be ok? might need tested tho!

This was more of a concern in the tube era, 'cause lethal potentials.  Big bypass caps had bleeder resistors in shunt so they would drain before the chassis could be opened after powerdown.  Big stereo PAs initially fill their bp caps through a resistor then use a timed relay to shunt down the resistor.  Nowadays they have a thermistor type that can be used without control, it is called an inrush surge limiter.

PIO resistors can have so much dielectric soakage that they can charge themselves back up!

Sometimes, especially when on-card linear regulators were common, a 'backwards' diode was placed the I and O terminals to protect it from the bypass caps.

What should be considered regarding the design of the system or the power supply

This is an excellent example of good engineering.

The turn off shouldn't be much of an issue, as each board will draw x current.

You could check how long before the voltage drops to a level where the IC's stop working, and this would give you a time.

One issue that can cause a problem is where a pin voltage exceeds the Supply voltage on the IC.

(This can occur if the voltage is sourced from another part of the circuitry or interconnected boards.)

With 60 boards using diodes in series with the supply becomes a logistical issue, but would ensure the spare energy in board X doesn't power board Y.

I'd be more concerned with the Power On.

6000uF is likely to represent a very high inrush of current, which could trip the supply or blow a fast acting fuse.

You may wish to test the supply can handle the current peak when they all get powered on at once.

Mark

Thank you very much for your answers.

Yes, as you told, I believe the current will drawn through the boards themselves, specially each of them has a "Power On" LED with the limiting resistor, so I think I will last short.

My concerns were more orientated to what has been told regarding one board supplying another board and exceeding the Vcc of the ICs, because at the end all the boards Vcc/Gnd lines are paralleled. I thought in puting the diodes in series next to the board supply in each board, but then obviously I will get a voltage drop of 0,6-0,7 V in all the Vcc, so if my power supply is +5V (which I intended, from a PC PSU), that means Vcc will be at 4,3 V, which can cause the circuit will not behave as expected. Any alternative in this direction?

Regarding the Power On phase, I was thinking in power on the system progressively by separating the Vcc lines in groups of several boards, let's say 3 or 4 groups (max 20 boards per group), dividing the total current needed. Anyway, I will carry some test for the current peak the supply can handle as suggested here.

I guess you can use Zener diode for each circuitry to limit voltage?
And don't forget to use the proper rating(voltage) of capacitor. Once, I was building a transformerless power supply, took a wrong value capacitor and Bam. It blew so hard I felt air pressure on my face. Luckily no part hit my face/eyes. But it was fun, I'm looking forward to do that once again, may be twice.. thrice....Hey, It was fun! By the way, I am looking forward to this project of yours, You might wanna video record it if you do something mischevious like blown your roof and hit a bird with cap.

mudz

Especially if you use a (fast blow) fuse to protect the Zener's from being over-powered.  You can use backward Shottkys in shunt with the Zeners if you want to improve protection from reversals.