A factory-fresh ATMega328p does not have a program; nor is it programmed for
an external oscillator.
A "new" Arduino Nano does not have a program installed.
It has a bootloader and may or may not have a blink program installed (as Jack suggests).
More to the point, it may not be running on the external oscillator -- it depends on
the fuse settings. There are variations on the bootloader; some may include a blink
program to let the user know it is there. But that tells us little about which
oscillator is running.
The default oscillator is an internal 125 KHz, not the external 16 MHz oscillator.
If you upload a program to that without changing the fuses to reflect the external
oscillator, all your timings will be off, including those related to communication with other
chips, etc.
If one uses the Arduino IDE to compile and upload, the fuse settings are automatically included in the
hex file. However, note that the Arduino IDE expects a bootloader to be already installed. Presumably,
Some programmers and IDE's may not do this. So the mcu may be left in a default state with a 125 KHz oscillator.
This may or may not be the case with the 1 MHz output, but it is worth investigating. Use something like
AVRdudess to read the fuse settings (You can do this without programming the chip), and compare the
settings with those from the fuse calculator available online. If necessary, the fuses can be programmed
to use the external oscillator of the Nano. It is also possible that the 16 MHz oscillator of the Nano is bad.
Then the device will not work or it will revert to internal oscillator. Both these conditions are not good, of course.
In other words, the Nano itself may be the source of the problem. It would seem to be at least likely, since the
25 MHz oscillator for the Si5351 chip is working. If so, then the chip itself is not getting properly initialized or
is not being programmed correctly. This seems to me to be much more likely than having a bad Si5351.
What to do? First, try a load or reload of the Raduino program. That will probably fix things, unless the Nano
oscillator is bad. Merely detach the Raduino and power it from USB to do the loading. If that doesn't work, get
another Nano and load it on a breadboard; attach a lcd and tuning pot, and see if the display changes.
If it does, it is probably ok and points to the real problem. Do not exceed 5v on the breadboard, no matter what
the source of power.
Another problem might be a short. Or low voltage supply to the Nano. The LM7805 regulator should run a little warm, but not hot.
If it does, it is drawing something approaching 1 amp, and that is double or more for a normal Nano. I would suspect first
the backlight of the lcd. These are notorious current hogs. The solution if the draw is too much is to install a 39 to100 ohm
1 or 2W resistor inline at pin 15 of the lcd. That is the backlight input line (pin 16 is the ground). Pushing the 5v regulator almost
to the point of thermal shutdown would probably result in a voltage starvation problem for the Nano, or too little current if the
backlight is drawing enough; either will disrupt the proper operation of the mcu.
Another problem might be that the 3.3v regulator which supplies the Si5351 is not performing right. Especially if there is poor or
inadequate 5v/3v interface. A little smaller 3v level would put it very near to the 0/1 level of the 5v. This is not too good for determining
IO through the pins. It is supplied by the LM7805 and so may be suspect as well. Either way it points to the Nano itself.
john
AD5YE
