In my original design for Attiny85, the reset pin is not connected. I use 144 Attiny85s for multi-channel signal processing.
When I boot up my MCUs powered by bench supply, they booted up successfully.
However, when I booted up MCUs with SMPS with smaller output current, sometimes MCUs do weird behavior. It's on, but outputs show randomly up and down. Not programed behaviors.
One of my assumptions of the problem is there are no pull-up resistors for reset pins...
But in the datasheet of Attiny85, there is a pull-up resistor to Vcc, 30k ~ 60k....
Do you think, 10k resistor which is typical value to pull-up reset pin, can solve the problem?
Try a 10k pullup on reset combined with 470nF between reset and GND. This creates a soft start that allows the smps to stabilize before the attiny tries to start. In all likelihood it tries to start but drops into a brownout detection state because the smps hasn't ramped up reliably yet.
Thanks for your reply. Is it ok to implement an RC circuit on the mainboard which contains all 144 MCUs for a soft start?
I mean without a pull-up resistor on the reset pin. Because it will cause a huge amount of men power to solder all 10 k pull-ups and caps.... I build separate chips that contain MCUs and a mainboard to supply data and power...
I think that would work, yes. But you'll have to analyze any possible interactions between the 144 reset pins; e.g. together they will draw a little current that may be significant enough to influence the component choice in the RC circuit.
Thanks for your advice, I manage to solve the reset issue by enabling BOD on MCU!! It can be done by software and easy. no need any hardware modification.