FT231XQ to dsPIC33EV256GM006-I_PT: PIC Deaf?

[NivagSwerdna]

I'm using parts that are new to me and a bit confused...

I have a FT231XQ which is USB powered and connected via 1k resistors to a dsPIC33EV256GM006 also USB i.e. 5V

Transmitting from the PIC to the FT231XQ to my PC works fine...

Receiving is very weird... On the PIC I am getting RX interrupts by the data is always 0x00; it looks like it is correctly decoding the frame but always reading zero.

Now... a close inspection of the data sheets suggests the FT231XQ will output to VCCIO which in turn is supplied from 3V3OUT and scoping shows that what is going into the PIC is of similar levels.

On the PIC I am receiving on pin 31 which is RPI24/RA8

Could it be that my level is high enough for some function but not high enough for others?

Reading the spec says Data I/O HI should be 0.75VDD = 5*0.75 = 3.75V so I am out of spec.

Any quick bodge I could try to pull it up high enough?

I'm not sure if this is a hardware issue or that I am fighting against MCC.

Thanks in advance
 

[SiliconWizard]

Had to look at the product to figure out the dsPIC33EV256GM006 was a 5V part only.
So yeah you're definitely out of spec. The easiest would be to add a level shifter. You could do that with a couple transistors if you don't want to use a full-blown level shifter.
Don't bother trying to "understand" the behavior - it's likely a bit random as you're probably close to the actual threshold for the HI level (0.75*VDD is a guaranteed treshold - a safe figure - not the actual threshold for a particular part.) Design to specs.

[voltsandjolts]

You could try a 680 Ohm pullup to 5V as a quick and nasty hack.

[NivagSwerdna]

Thankfully CNPUA = 0x0100 was enough!

Getting data now.

Maybe URXINV and a BS138 inverting shift with and a 10k pull up for Rev 3!