[USB] HID does not work for me

octal · Post by **octal** » Fri Jan 26, 2007 8:09 pm

I tested the HID example, it does not worked for me. I tested it on my EasyPic3 with 18LF4550 and a 20Mhz cristal. I tested an old program I had written in C and it worked, so my hardware is out of problem. I reloaded the example of SFBasic, and checked the config and it's the same.
It does not work! In the hardware manager, it detected and uncknown USB hardware, and in a software I had writen that let me manage HID devices, send reports to them and read reports from them (an USB terminal like) I do not ever detect my hardware.

I'll try to check the example, and may be write a simpler example to see what happens.

Best regards

David Barker · Post by **David Barker** » Fri Jan 26, 2007 8:24 pm

Try this

// device and clock...
device = 18F4550
clock = 48

// 20Mhz crystal, 48Mhz internal (FS USB)
config
PLLDIV = 5,
CPUDIV = OSC1_PLL2,
USBDIV = 2,
FOSC = HSPLL_HS,
VREGEN = ON

// import modules...
#option USB_SERVICE = false
include "usbhid.bas"

while true
Service
wend

This disables interrupts and polls manually. All it will do is enumerate and maintain the connection. Let me know what happens...

octal · Post by **octal** » Fri Jan 26, 2007 8:50 pm

does not work David. Sorry.

Is the colck calculated correctly by the compiler ?
I have found in the help file

This can be extremely useful when used in conjunction with the pre-processor to generate different timing code. However, it is important to note that the built in routines will only currently generate the correct timings for a certain range of clock frequencies. These include: 3.58, 4 to 13, 14.32, 15 to 40 and 64MHz. Any frequencies falling outside of this range will be matched to the nearest supported value.

in the precalculated values we can not see 48Mhz... so I think that the compiler takes 40 instead. What do you think ?

I'm going to analyse the code of the HID lib and compare it to Microchip HID firmware ... and I'll have a deep look in USB Complete of J.Axelson ...
I'll take you informed.

Regards

Alfch · Post by **Alfch** » Fri Jan 26, 2007 9:49 pm

Hi Octal

I just discover and test CDC samples.
When I compare two ASM files from same config but different 'clock' values (48 and 40) I find different values for Timer1 reloading (209 vs 216).
So we can assume that 48 is a new value not yet documented.

Regards

David Barker · Post by **David Barker** » Fri Jan 26, 2007 9:50 pm

I'm afraid it appears to be a problem with the SE version, which doesn't have bank switching enabled. I thought I had managed to get around this problem for the USB libraries (which use dual port RAM located at $400) but it's not working correctly. My sincere apologies and shame on me.

It looks like I will have to pull the libraries from the SE version for an indefinite time.

BETA testing was performed on a number of boards (including EasyPIC3) with great success. Of course, myself and all other BETA testers are using the full version of Swordfish which does not have this limitation - so no problems running the USB modules.

Apologies to all SE users.

octal · Post by **octal** » Fri Jan 26, 2007 10:04 pm

Ok David ... I understand ...

I'm using Microchip C18 for USB development and apparentelt I have to continue do that for some time.
I'm going to wait until I collect sufficient money and definitely buy a licence of SWBasic.

Best regards.

Octal.

Alfch · Post by **Alfch** » Fri Jan 26, 2007 10:04 pm

No problem David !

The good news is : SE users will be allowed to do some personal tests on HID and CDC libraries !!!

Some other compilers don't offer this chance !!
Waiting for your USB libraries I was not certain to make my own small tests.

Regards