Thank you @megari for the info and the hints, but I’m not sure how to proceeded.. Isp pin 5 (reset) seems connected to pin 5 of LSF 204, and its pin 10 goes to atmega’s reset pin: programmer pulse low right the isp pin, but ic’s pins remain high (2.89v). So, where could I add the resistors?
I’m pretty sure my cable connection it’s right, I checked the continuity on isp connection, and my programmers work fine with all my uno and mega arduino boards, so I’m afraid my koryuu has something wrong. I ordered a couple of.atmega328p, but i’d leave the ic replacement as last chance.
I see. I can confirm that the following is correct:
- ICSP pin 5 (/RESET) is connected to LSF0204 pin 10
- LSF0204 pin 5 is connected to atmega328p pin 29 (/RESET)
So you definitely do have the RESET line set up correctly. What is a bit odd is that 2.89V is a bit low for high logic level (nominal 3.3V). The pull-ups on the MCU should not be able to keep the voltage that high if something is actively pulling the /RESET line low. I wonder if a longer pulse width might help. Anyway, something is definitely wrong here.
As for trying to pull the /RESET line low by grounding it through a resistor, you can do that by connecting the /RESET and GND pins in the ICSP header together. Using a resistor in the single-kOhms range is just a precaution. Due to the pull-up resistor in the MCU being expected to have a high Ohm value, the current should be negligible even when shorting /RESET and GND. However, we’re investigating something suggesting a potential malfunction in the /RESET line, so maybe best be careful…
UPDATE: i desoldered Atmega328P pin 29 (Reset) and… Koryuu still works fine, so the reset pin was disabled by factory (or, i think, the Koryuu had to stay always in reset state and never start)
That is not quite accurate, I’m afraid. When the the /RESET pin is not disabled, it has an internal pull-up resistor active due to an override within the chip. Please see atmega328p datasheet section 13.3, “Alternate Port Functions”. The tables are a bit technical and not the easiest to interpret, but it will help to remember that RSTDISBL is 1 when unprogrammed (the default), 0 when programmed.
Anyway, at this point it does seem that something is wrong with your unit. It would be interesting to know if replacing the MCU fixes the issue, if you’re willing to try it.