Hi,
I've bought a CP2102 USB to TTL adapter that I'm using to program my Arduino. The adapter doesn't have an auto-reset so I need to press reset just before the sketch upload starts but I am able to upload sketches.
It is connected using the following pins.
5V -> 5V Arduino
GND -> GND Arduino
RX -> TX Arduino
TX -> RX Arduino
DTR pin (unused)
CTS pin (unused)
There are a lot of posts on the internet that this auto-reset can be built-in by putting a 100nf cap between the DTR line and the RESET line of the Arduino.
So I added the following connection but that didn't work.
DTR pin -> 100nf cap -> RESET pin
The following did work :
DTR pin -> 1 ?F cap -> RESET pin
I was just wondering why I need to use a 1 ?F cap while the internet seems to be able to do it with a 0,1 ?F cap ? 
I'm using the Arduino IDE to upload the sketches
Interesting. Generic 0.1uF, 50V rated, ceramic cap is all I ever use.
Which "Arduino" and from where did you purchase it?
most arduinos already have that circuit in place so that might be the problem. for stand alone chips 10k pullup to vcc is also normally used but i prefer a diode there instead. there are sometimes issues with voltage mismatch between dtr and reset pin which this solves for every setup ive encountered.
ps type of cap dont matter.
10K pullup and diode in parallel is the norm.
Cathode to Vcc, anode to reset pin.
CrossRoads:
10K pullup and diode in parallel is the norm.
Cathode to Vcc, anode to reset pin.
Leads to an interesting question:
Why is such a diode not part of the Arduino design?
It is. Was not on Duemilanove, made it into Uno & later designs.
CrossRoads:
It is. Was not on Duemilanove, made it into Uno & later designs.
Ah! Overlooked that.
Blame it on Sparkfun/ Gravitech; it is absent on the Pro Mini (with which I am most familiar), Nano, Fio, Lilypad ...
interestingly the resistor can be left out with no problem
Depends on the electrical environment. Atmel recommends one, see Section 3
but the diode does solve issues caused by positive spikes pushing the chip into isp mode on some platforms.
See Section 3 also - High Voltage programming mode is what seems to kick off, leaving the chip in what appears to be a hung state.
i found that connecting the 10k to 3.3v instead of 5v also works for all systems w/o need for a diode.
I'd say that's kind of iffy - Section 29 shows Min Voltage for Reset to be a valid High is 0.9 x Vcc, so 4.5V for a 5V powered part.
VIH2
AtmelAVR042 AVR Design Considerations.pdf (236 KB)
It was on a Funduino Uno (Arduino uno clone). But good call. I'll check with my real Arduino (from the starter kit) tonight.
Paul__B:
Which "Arduino" and from where did you purchase it?
Noticed that on my Arduino Pro Mini it works without the cap because apparently that one has the auto-reset out-of-the-box.
On the Arduino Uno it didn't. As it only worked if I pressed reset right before the sketch was uploaded (or when using the 1?F cap).
john1993:
most arduinos already have that circuit in place so that might be the problem. for stand alone chips 10k pullup to vcc is also normally used but i prefer a diode there instead. there are sometimes issues with voltage mismatch between dtr and reset pin which this solves for every setup ive encountered.
ps type of cap dont matter.
interesting that crossroads response has my signature attached. i wonder if i have a virus or theres a bug in the system? or case of abuse of moderator power? wouldnt be the first time.
john1993:
interestingly the resistor can be left out with no problem
Depends on the electrical environment. Atmel recommends one, see Section 3
but the diode does solve issues caused by positive spikes pushing the chip into isp mode on some platforms.
See Section 3 also - High Voltage programming mode is what seems to kick off, leaving the chip in what appears to be a hung state.
i found that connecting the 10k to 3.3v instead of 5v also works for all systems w/o need for a diode.
I'd say that's kind of iffy - Section 29 shows Min Voltage for Reset to be a valid High is 0.9 x Vcc, so 4.5V for a 5V powered part.
VIH2
anyway its also popular to insist chips cannot be run at 16mhz on 3.3v. i suppose it depends on the application and what your business insurance flyer has to say about potential liabilities. in my experience the threshold was closer to 2v for every family i tested.
anyway its also popular to insist chips cannot be run at 16mhz on 3.3v. i suppose it depends on the application and what your business insurance flyer has to say about potential liabilities. in my experience the threshold was closer to 2v for every family i tested.
Well most of us know that one can run our AVR chips at 3.3v @ 16 Mhz, but most of us also know that it is not operating at the AVR datasheet recommendations. For us hobbyist that is fine, but a professional EE designing something commercially for a company would be foolish to operate any component outside it's datasheet ratings.
exactly, pretty much as ive always said. i do remember a time when even suggesting it here was cause for tar and feathers (many may not know actual tar and feathering was fatal in virtually every case). much worse on freaks.
Both on the Funduino and the original Arduino uno it only works with the 1 ?F cap.
ddewaele:
It was on a Funduino Uno (Arduino uno clone). But good call. I'll check with my real Arduino (from the starter kit) tonight.
Paul__B:
Which "Arduino" and from where did you purchase it?