SPI using 18LF4550 with Vdd=3.3V?

[arrow]

Hi

I have implemented some code using SPI, and operating at Vdd=5V. This works well.
I now swapped out the PIC to a PIC18LF4550 and am trying to
make the system work with Vdd=3.3V, 4MHz Xstal.

What I find is that things do not work at this lower voltage.
I was wondering if there is anything that I should set when Vdd is not 5V but 3.3V? (e.g. something in SPI?)

Thank you in advance for your help.
a.

[ckielstra]

Could be something in the startup code and/or fuses settings. What is your compiler version? What does your #fuses line look like?

[arrow]

Hi Ckielastra

My compiler version is: 3.242

My fuses look like this:

[ckielstra]

I compared the startup code for v3.236 and v3.249 but didn't notice any important changes.

I don't know what your previous chip was but most likely the SPI pins have moved to different pins, double check your hardware connections. Also disable all other internal devices that share the same pins as the SPI module, especially disable the analog function for these pins.

Also add the NOPBADEN fuse to set the pins B0 and B1 for digital I/O on power up instead of analog inputs.

[arrow]

Hi

If the code with SPI works for Vdd=5V, should it work without modifications for Vdd=3.3V?

Thank you for your help.
a.

[arrow]

Hi

Another weird development.
If Vdd =3.3V the SPI does not seem to initialize the MMC card.
If Vdd =5V then everything works well. (I have a V divider with resistors out of the PIC and into the card to give the card only 3.2V).

BUT if I measure the SPI Sclock (Pin B1 on the PIC18LF4550) using an Oscope probe then the system works, and data is written to the MMC.

Can anyone please tell me why the Oscope effects the results? and how can I make it work with no Oscope probe?

Thank you
a.

[Ttelmah]

The obvious change, will be the addition of about 30pF of capacitance. The other obvious change, will be a ground connection from the scope probe. Are you sure you have a good ground between the two units?. If so, try adding a little capacitance to the line.

Best Wshes

[arrow]

Hi Ttelmah

Thank you for your suggestion.
I have checked and the two units are definatley on the same ground (using an ohmeter).

I am also guessing that the problem is to do with capacitance.
The MMC specifications says that the maximum bus signal line capacitance should be 250pF.
I am not sure what I have now, nor do I know how to measure it.
The specification sheet says that in the I2C mode the SCL pin has a capacitance of 400pF. I am using the same pin (PIC18LF4550 pin RB1)
for SPI SCLK. Is it possible that my capacitance on that pin is that high?
How can I reduce the capacitance on this pin?

The 30pF cap should be placed from each of the SPI lines to ground?

Once again thank you for all your help.
Regards
a.

[Ttelmah]

The SCL pin, does not 'have' 400pF of capacitance. It supports _driving_ up to 400pF, while meeting the timing specifications given. Your scope (if it has a normal probe), will be roughly equivalent to a 1MR resistor to ground, paralleled by a 30pF capacitance. One possibility, is that you are getting some ringing on the lines, causing problems, and this is serving to attenuate this.

Best Wishes