USB CDC Communications for Dummies!

[mackrackit]

The default config in the PBP directory looks correct for a 20MHz external.
Can you confirm that the chip is running at 48MHz?

[cncmachineguy]

I can't confirm it right now, but I feel it is. I have been playing with receiving serial at 115K without any problems. But I was using configs set in code. I also tried those same configs, still nothing. I will double check all configs tonight.

Thank you for verifing the defaults should work. I know Darryl said not to mess with the code until it connected, but I think I can at least toggle a pin to make sure there is life, and at what speed.

[mackrackit]

Post your code. Maybe there is something else.

[cncmachineguy]

Here it is:

' USB sample program for PIC18F4550 CDC serial port emulation
'  Requires PBP 2.60 or later
'  Compilation of this program requires that specific support files be
'  available in the source directory.  For detailed information, see
'  the file PBP\USB18\USB.TXT.
 Include "cdc_desc.bas" ' Include the HID descriptors
Define  OSC     48
buffer Var Byte[16]
cnt Var Byte

 OSCTUNE.6 = 1  ' Enable PLL for 18F87J50 family
 Pause 10
 USBInit   ' Get USB going
' Wait for USB input
idleloop:
 USBService  ' Must service USB regularly
 cnt = 16  ' Specify input buffer size
 USBIn 3, buffer, cnt, idleloop
' Message received
 buffer[0] = "H"
 buffer[1] = "e"
 buffer[2] = "l"
 buffer[3] = "l"
 buffer[4] = "o"
 buffer[5] = " "
 buffer[6] = "W"
 buffer[7] = "o"
 buffer[8] = "r"
 buffer[9] = "l"
 buffer[10] = "d"
 buffer[11] = 13
 buffer[12] = 10
 buffer[13] = 0
outloop:
 USBService  ' Must service USB regularly
 USBOut 3, buffer, 14, outloop
 Goto idleloop  ' Wait for next buffer

[cncmachineguy]

revisied code with a speed test added

' USB sample program for PIC18F4550 CDC serial port emulation
'  Requires PBP 2.60 or later
'  Compilation of this program requires that specific support files be
'  available in the source directory.  For detailed information, see
'  the file PBP\USB18\USB.TXT.
 Include "cdc_desc.bas" ' Include the HID descriptors

Define  OSC     48
trise.3 = 0
buffer Var Byte[16]
cnt Var Byte
    
 OSCTUNE.6 = 1  ' Enable PLL for 18F87J50 family
 Pause 10
'*****************************************
'--------------Speed test 
 
speedtest:
    toggle porte.3
    goto speedtest
 
'--------------End test
'******************************************    
    
    USBInit   ' Get USB going
' Wait for USB input
idleloop:
    
 USBService  ' Must service USB regularly
 cnt = 16  ' Specify input buffer size
 USBIn 3, buffer, cnt, idleloop
' Message received
 buffer[0] = "H"
 buffer[1] = "e"
 buffer[2] = "l"
 buffer[3] = "l"
 buffer[4] = "o"
 buffer[5] = " "
 buffer[6] = "W"
 buffer[7] = "o"
 buffer[8] = "r"
 buffer[9] = "l"
 buffer[10] = "d"
 buffer[11] = 13
 buffer[12] = 10
 buffer[13] = 0
outloop:
 USBService  ' Must service USB regularly
 USBOut 3, buffer, 14, outloop
 Goto idleloop  ' Wait for next buffer

the results of the speed test: pin toggles at 3Mhz. To be clear, there is .3333uS between toggles. Now I don't feel like that is correct. lets see: 48Mhz=12Mips. So thats .08uS per instruction? So we are executing 4 instructions? Hmm, that doesn't sound right.

Ok, I will see if someone else can think better then I this evening

[mackrackit]

Might be a problem with the buffer size compared to the data being sent. They are different sizes.

Here is an example:

Buffer    VAR BYTE[16]
Cnt       VAR BYTE
B0        VAR BYTE
B1        VAR BYTE

ADCON1 = 15               ' Set all I/Os to Digital      
CMCON = 7                 ' Disable Comparators
Cnt = 16 
  
for b0 = 0 to 15
    lookup b0,["USB CONNECTED!",10,13],B1 
    BUFFER(B0) = B1
    NEXT B0
 USBINIT 
' Main Program Loop
Loop:
    USBService        ' Must service USB regularly
    USBOut 3, Buffer, Cnt, loop
    goto loop

As for your speed test... my brain is ready to fuss out for the day also...

[cncmachineguy]

Hmmm.. I would think OS should still see it as a new device even if thats the case.

Well I am studing the schematic now, I am not sure I have the "CAP" on Vusb. But I can say this thing works with a USB bootloader no problem. Now does that mean winblows will see it? IDK

I see there are pullups on the data lines. And according to the datasheet, The internal pullups must be disabled. I have been looking at the include file, and I think there are being set.

I am just about to give up on this "pre-configured" answer. After all, it may NOT be the answer for this chip anyway. I will start looking at your solution to see if I can make that work.