I have a few USB-to-serial-ttl converters.
They have a CP210x chip with 3.3V signals.
The 3.3V DTR signal won't reset my standalone ATmega8 (5V, 16MHz), because the 3.3V is too low.
My reset circuit is 10k to 5V with diode 1N4148 next to it, and 100nF to DTR.
I would like to automatically update a sketch, without using a reset button.
I don't want to have more components with the ATmega chip, so how can I change the USB-to-serial-ttl converter to make it work ?
That didn't work.
I think a voltage drop of 3.3V is not enough for the old ATmega8.
Perhaps I have to solder a non-inverting 74HCsomething to the converter. The converter has a 5V pin, so that should work.
Everything works if I run the ATmega8 at 3.3V, but I want to run it at 5V.
I though you had to ground the RESET pin to reset the microcontroller. Maybe you problem is 3.3V is not enough to bring the chip out of reset mode or give stable power if it's using >8Mhz clock speed.
Do you have a link to the USB2TTL converters you have and a drawing of how you have them wired up to the microcontroller.
Riva, that is the problem indeed. The 3.3V DTR signal is too low to make the ATmega8 (5V 16MHz) reset.
I have the DTR connected to the reset with a capacitor.
fungus, the DTR from the converter changes from 3.3V to 0V. The reset of the ATmega8 will drop 3.3V lower. It is 5V so the pulse will be 5V - 3.3V = 1.7V. That seems not good enough for the ATmega8.
I have done a quick and dirty solution, by lowering the reset voltage.
The reset has now 10k to 5V and 47k to ground. The reset voltage is 4.1V. That is a safe value (I think) and the 3.3V DTR signal will cause a reset.
ADDED: Major disadvantage: the USBasp programmer doesn't work with the extra pull-down of 47k at the reset. So I will have to write the bootloader without the 47k and use them in a project with the 47k.
