simplifications

[picpic020960]

Hello

I want this assembly.

http://www.fischl.de/usbsticklogger/

I have the component.

the diagram programmer 'openprogrammer' can it be easier?

(no 'Switching voltage regulator')

thank you for your reply.

greetings

(Google translation)

[Alberto Maccioni]

Sorry but I really don't understand what you are asking.
The link points to a serial logger project.
What do you need?

[picpic020960]

Bonjour

i want to load firmware in pic24fjxxgb002 with Openprogrammer and easier hardware (PIC18F2550 and some R and C ).

if the Pic24fj is powered with 3,3v is it possible ? (like LVP programming)

sorry for being so unclear

[Alberto Maccioni]

Yes, this device requires only a 3.3V regulator and no DCDC converter.
You can remove the unneeded components and concentrate the remaining ones in a single circuit.

[picpic020960]

Bonjour ,

other issue : i have only 4 , 20 or 24 MHZ quartz.
(the bootloader work with 20 MHZ quartz but i will
tried with JDM for first time )

With MPLAB and the simulation can i find the good values of fuses

// Define configuration registers (fuses)

pragma config PLLDIV=3, FOSC=HSPLL_HS //for 12 MHz crystal

pragma config CPUDIV=OSC1_PLL2, USBDIV=2

pragma config IESO=OFF, PWRT=OFF, BOR=ON_ACTIVE, VREGEN=ON, WDT=OFF

pragma config MCLRE=OFF, PBADEN=OFF, STVREN=ON, LVP=OFF, XINST=OFF, DEBUG=OFF

and the RS232 (for debugging)

TRISC &= 0xBF; // Set RC6 as an output
TRISC |= 0x80; // Set RC7 as an input
RCSTA   = 0x90; // Enable serial port, enable receiver
TXSTA   = 0x24; // Asynch, TSR empty, BRGH=1

// Baud rate formula for BRG16=1, BRGH=1: Baud Rate = Fosc/(4 (n + 1)),
// or n = (Fosc / (4 * BaudRate)) - 1
// At 48 MHz, for 115.2K Baud:
//     SPBRGH:SPBRG = n = Fosc / (4 * 115200) - 1 = 103.17
BAUDCON = 0x18; // BRG16=1 txd inverted

// SPBRGH = 0x00; // At 48MHz, SPBRGH=0, SPBRG=103 gives 115.2K Baud
SPBRGH = 0x02; // 0x0270 gives 19200 Baud

if CLK_48MHZ

// SPBRG = 103; // For 48 MHz clock
SPBRG = 0x70; // For 48 MHz clock

else

SPBRG   = 52;   // For 24 MHz clock

endif

printf("USB Test Startup\r\n");

thanks to help me

Best regards

( Tutto Va bene !!!)

[picpic020960]

Bonjour ,

ok for the Quartz and the fuse.

One question : why the Quartz frequency is 48 MHZ in the MPLAB simulation porject ?

by the way Happy New Year

[Vitor_Boss]

Read the datasheet for more information.

For a 4MHz use "pragma config PLLDIV=1"
For a 20MHz use "pragma config PLLDIV=5"
For a 24MHz use "pragma config PLLDIV=6"

If you gonna use the JDM to write PIC you can easy change the configuration bits without recompile the code. Open the Winpic800 with the firmware, go to Config. tab and change the section "Oscillator Selection" to your hardware.

The MPLAB project is configured to run at 48MHz on pic using PLL.

[Alberto Maccioni]

48 MHz is what is given to the CPU.
PLLDIV is chosen to have 4MHz at the input of PLL.

[picpic020960]

Bonjour,
the test with the bootloader = KO
with JDM the led blink but the Device is no good for USB enumeration.
I want to degugging but
the rs232 debug at 19200 bauds is not readable : one bit is < 50 us
U(55 hexa) ==> A5
S(53 hexa) ==> A3
B(42 hexa) ==> A2
is it possible that the trace are to long for the USB enumeration Process ?
have you some solutions for my issues ?

thanks and best rgards

[Alberto Maccioni]

The serial debugger has never worked correctly with USB; it prevents proper enumeration.
I suggest you first have a working USB programmer, then experiment with your idea, so that the serial programmer, which is extremely unreliable, is out of the game.

[picpic020960]

re-Bonjour

thanks for the information
version without debugging Ok but some questions :
Why firmware version 0.0.4?
Who can i see the USB peripheric ?
Who can i see the log ( /option log activity ) ?
(test with 18F2550 only )
OS : Vista

screenshoot

Device detected: vid=0x04D8 pid=0x0100
Path: \?\hid#vid_04d8&pid_0100#6&36f79b7&0&0000#{4d1e55b2-f16f-11cf-88cb-001111000030}
Programmer detected
Communication timeout
Firmware version 0.0.4
Hardware ID: 0.0.0 (?)

[Vitor_Boss]

Try flash the last firmware(0.0.8)

[picpic020960]

Bonsoir ,

after some unsuccesfull tests i can said :

• the usb link with quartz 20 MHZ very fragile , just the connection with usb pin
  only , without run Openprog: blink led ( 4 hz then 1 hz then 4 hz)
• some time Led blink 4 hz then 1 hz then i run Openprog : it found firmware version
  0.8.0 ( one time for ten tests)
• the test is on breadboard but the hardware is very simple
• i will test with quartz 4 MHZ : i hope it will best !
• i never see the USB designation of Open Programmer in the 'device manager' of
  VISTA

questions :

• some body has test with 20 MHZ quartz ?
• what you see in the 'device manager' when the usb connection is correct ?

thanks and best regards

[Alberto Maccioni]

I think your problems come from the breadboard assembly: at high frequency you have to keep wiring very short and also load less the quatz nodes (see datasheet); or you can try to load more and have a more stable oscillation, but I don't know what is the limit.
I cannot check right now but you should definitely see a HID device with proper VID&PID in the device manager.

[picpic020960]

and with a 4 MHZ quartz ?
perhaps with low speed USB but who do it ?
the firmware can work slower ?

for the device manager i see two lines with VID&PID...

and when after Openprog is running how to see the log file

(it is a little harder than I thought)

Thanks

[Alberto Maccioni]

4 MHz may be better in case of breadboard, and it doesn't change the USB speed: just set PLLDIV to 1.
Pay attention to the oscillator section: use proper capacitors and short wires.
The firmware has to run at 48 MHz.
The log file is generated only for reads and writes to a target device.

[picpic020960]

it is strange.
with Microchip USB bootloader (a) , quartz 20 MHZ and breadboard i have no
problem with the USB connection...
(perhaps (a) dialog not need high performance like Openprog)
i will test with 4 MHZ and make feedback

[picpic020960]

it is strange.
with Microchip USB bootloader (a) , quartz 20 MHZ and breadboard i have no
problem with the USB connection...
(perhaps (a) dialog not need high performance like Openprog)
i will test with 4 MHZ and make feedback

[Csanyi Pal]

Sorry, this post has nothing with this thread. It can be deleted.

--
Regards from Pal

[Csanyi Pal]

Sorry, this post has nothing with this thread too. It can be deleted.

--
Regards from Pal

[Alberto Maccioni]

It sure is strange; the USB connection should either work or not regardless of the application, it's not a matter of performance.

[picpic020960]

Bonjour ,

test with 4 MHZ quartz = KO
is it possible that you load in your website .hex generated by your environment for
quartz 4 MHZ and quartz 20 MHZ (exists change when i compile , linker , compiler version) ? to be sure that the problem comes from the 'hard'

i continue search with HIDCOM for learning.

with Op the success rate is better ...but wihout stability I can not go far

have you USB packets trace with Snoopy (XP) ? how much packets for the enumeration ?
when i test , 14 or 16 packets and no more

best Regards

'All begins with Breadboards'

For Pal :

comparison : same hardware (breadboard) firmware bootloader Microchip OK , Open Programmer KO

i can't delete the double post

thanks

[Alberto Maccioni]

Here is the firmware compiled for 4 MHz quartz.
You have not really described the problems you experience: KO is not really telling anything.
Never traced USB, it has always been rock solid in my case.

[picpic020960]

Bonsoir Alberto
thanks for the .hex. I will test tomorrow.
i want to test PIC_HID with loading by microchip bootloader (more comfortable) but nothing run. Perhaps fuse pb.
with loading by JDM in 'debug_print 1' i see rs232 trace.
i insist : all these tests are made with breadboard.
NB : some time with microchip bootloader (USB) i have BSOD :-(
best Regards

[Alberto Maccioni]

I think the problem with the bootloader has nothing to do with the breadboard; it's a functional bug: if a program is designed to run from a certain address you have to make sure it's being loaded correctly from the bootloader and is given control after some time.