Objective Development Forums

another thing, as we want to increase the report size from 5 bytes to 8 byte, what else to do with tweaking ? assign more bytes to the report buffer, seems not feasibl This seems a bit unclear. You don't set the size the report does? If you want to send more then the report needs, yes you need to a...

usbHidReportDescriptor needs to be PROGMEM const char usbHidReportDescriptor or const ROGMEM char usbHidReportDescriptor I forget the order that is required and I'm sure that wont be the only error you encounter. I believe i had 50 some lines to fix. Its actually the IDE that is giving you the grief.

Right I see where I added that up wrong. Though its expecting something there... Glad it works well enough to get your work done.

Note the last 0x00 byte after sending the correct amount of data for all reports. IT WORKS Now... Can anyone explain to me why the hell I need to send that additional byte? Sounds to me like your count is off, I bet your device description has 9 bytes total instead of 8. You could fix the descripto...

You are correct about not need to 0 index it, here is an example I use and its works. No, I also do not believe its order dependent but in my case I do send in order. Every time I send my report I send report 1 then 9. I can not say what is going on there with windows, it should not matter what is s...

My goal is to develop a custom control panel I really hope you figure this out, because I would absolutely love to see that work. I had to give up on it. Are you setting the report id in your out ID report? 0 for Keyboard 1 for joystick? Also when dealing with reading a joystick there really is no ...

I wanted to figure out why there were separate classes for Game Pad and Joystick. Yeah I ran in to this as well. Joystick is mainly for flight application as it has ruder and throttle controls. The game pad is more for everything else. For example the PID layer will not work on the joystick only ga...

No it does not work that way...

The M$ OS gets the description form the registry. If the device is new its pulled from the device. So if you change the name in the device mem like that you need to also delete the device from the registry. usbdview may not work for that, not sure.

make sure you set your check before you init the usb. and make sure you turn on interrupts. PCMSK0 |= (1 << PCINT0); PCICR |= (1 << PCIE0); sei(); // turn on interrupts ISR(PCINT0_vect) { usbConnected = 1; } also don't you want PCINT17 or PCINT16 https://sites.google.com/site/qeewiki/_/rsrc/12975585...

They didn't reply to me, they replied to the above user, I dont retain every post I comment in. Anyways I think I jumped the gun on that one .The longest wait I have seen is 4 ms in testing. So I take that back, its not a time out issue. What's your code doing for 8ms between calls? I just meant tha...

ok found my issue, too long of a wait in-between polls. What it the max with the current v-usb version?

blargg, that was helpful thx. I finally got the init down to the point I'm happy with it but in my poll loop the usb is constantly drops out. AS you suggested I dont want to watch dog it. I'd rather find the error. My build report is well under 8 ms, and my polls occurs every 8 ms on time, each time...

So why do so many people full with timers, watch dogs, status registers, and resets?

Ok I have been using v-usb for 7 years and all I do is this... //init static void usbReset(void) { usbDeviceDisconnect(); /* enforce re-enumeration, do this while interrupts are disabled! */ uchar i = 0; while(--i) { // USB disconnect for >250ms _delay_ms(1); }; usbDeviceConnect(); } cli(); sreg = S...

well UsbPoll() is a bool 1 or 0 so its as simple as if ( UsbPoll() ) but inorder to use that the USB init has to have run, and if you ran that its assumed you know the usb is connected, so I think that is out? Again depends how you do things. Pin change is the best bet and very easy. You should not ...