+5 volt Reference Turn-On

     Another problem with the +5 volt reference is that it is coming up before the +5VF supply.  When this happens the protection diodes inside the MPC555 turn on and excessive current starts to flow.  All the QADC inputs, reference, and supply always need to be less than one diode drop from the +5 volts supply.  At first I figured this wouldn't be a problem because surely the +8VF supply which sources the +5 volt reference would come on later than the +5 volt supply.  But what looks to be occurring is because the +5VF supply is so heavily loaded and the +8VF supply is lightly loaded the +8VF LM2940 regulator is coming up much fast than the +5 volt regulator is.  The precision reference follows the +8VF supply so it is coming up before the +5 volt supply.
     There really isn't any easy way to delay the +5 volt reference coming on.  About the only thing that can be done to slow it down is to actually short the output (current limiting works good here) for a brief period of time releasing it after the main +5 volts supply is up.  Kind of a crowbar approach but does seem to work.  The big problem now is finding a location for all these extra components.

This is what I saw first between the +5 volt reference and the main supply.  Besides the fact that the reference is leading the +5VF supply it bothered me that it was by one diode drop.

Here is the relationship between the +5VF and +8VF supplies.  Really clear that the +8 volt supply is coming up much faster.

Went back to the two 5 volt supplies and added a schottky diode between then.  Virtually no difference from the first plot I took.

Next I unplugged the Axiom PCB completely.  Now both the +5 and +8 supplies come up at the same time to so the +5 reverence doesn't get much ahead of the now lightly loaded +5VF supply.

Re-attached the +5 volt supply to the Axiom PCB.  This loads the +5VF supply heavily again.  This plot shows exactly how far the +5 volt reference is ahead of the main supply because without all the other pins attached to the Axiom there is nothing to clamp it.  Almost 1.8 volts!

This was to 100% verify that the MPC55 was clamping the two supplies together.  I added my external schottky diode back across the two supplies and we are exactly back where we started from in the very first plot.


Modified with turn on delay shunt.

This is what it looks like with the delay.  The +5VF supply is clearly stabilized before the reference comes up.

This is the initial charging current through the 1uf cap.  Channel #2 is the at the junction of the 1uf cap and the 100 ohm resistor.