NetBurner Community Forum

Hi Everyone,

After 5 months of doing something completely different I have returned to my MOD54415 project.
In the mean time my new hardware is ready.
The hardware is specific to the application which uses the PWM outputs of the MCPWM a lot but in general it very much resembles the DEV-70 development board.
With it I am experiencing something odd which I did not have with the Dev-70 board.

The software I use simply reads the dip switches and drives a few LEDS on the base of it just like you can do on the Dev-70 board. Oddly enough though when I power up the board it does not start up (+/- 80% of the power ups).
Just like the Dev-70 board there is a reset switch and when I press that one it always boots perfectly, it never misses a beat.

I will investigate it in more detail to see what may be different between the Dev-70 and the new board but I was wandering if anybody has any suggestions ?

If I find out myself what is up I will post it because somebody else could do something with the info.

All for now.

First Result: it definitely has something to do with power up behaviour:

If I apply the power by switching on the lab supply most often it does not boot.
If I leave the lab supply switched on and apply the power with a switch it never misses and always boots.
If I use the power supply ON/OFF switch to apply power but keep the reset button pressed when the power comes on it always boots as soon as I release it.

I used the DEV-70 board with a 24V DC supply and converted that down to +5V for the board. The startup behaviour was probably OK for the processor(board) so I never had a problem.

Next: do some experimenting with a resistor/capacitor combination to keep the reset line down while powering up.

There are minimum rise times for power rails, so that might be a cause. Maybe your board needs a reset circuit that monitors the voltage rail.

Agreed 100% w/ Rnixon. Sounds like your PS comes up slowly, or that your lab supply produces some brownouts or spikes when powered on.

yes, it definitely has something to do with power up behavior.

The solution I found is relatively simple: put a 1 microfarad capacitor between the reset pin and ground and it always works.
On the scope you can see the reset pin comes up +/- 1mS later and everything is fine in that case.

I have been using it that way for a few days and it has never failed to boot.