Instant Interrupts - Revisited

[RYTECH]

Hi,

I have some mostly working code now using your interrupts thanks to your "hint" about the endpoints earlier...

I was reading another thread that was about "If plugged" - The solution works great monitoring SOFIF and IDLEIF registers - the main loop now does run when the cable is unplugged.

The only case that is still at issue for me is that the device doesnt go passed the "DoUSBinit" subroutine if it's powered on while not connected - which for me is a big deal because the device i'm building is going to be powered up at times with just the LCD i have connected. This LCD is managed by a 16F767 in another device which has some fans too and a light. it's all connected via cat5e cable for one way serial transmission from the 18F4550 via serout2 commands, and i give it 12v and GND each via 3 wires respectively which feed a 7805 and a couple transistors for pulsing the fan speed and light intensity, that all appears to work nicely)

So I guess the last thing to do before I glue everything together (lol) is to get the device to go to the main loop if usb cable is not plugged at initialization time (and have the usb still work after if its eventually plugged in while the device is running).

Anyone else have this "self-powered device" / usbinit kind of issues?

[RYTECH]

Darrel the "extra" data is for parsing strings that come in after my oncomm event in visual basic. I get the stuff in between the letters - I haven't successfully used "binarymode" in vb .. always stuck with text mode.

I use whats come in between "A" and "B" to fork the other data over to certain variables... there's probobly a much better way to do it... I've already seen instances where manipulating the strings in higher speed timers cause lockups and weird things happening.

But that's not really the right forum for that.. but i'm interested to resolve the issue with the "cable not in during usbinit on device initialization" issue.

[RYTECH]

Should i wire up an input to the USB V+ and monitor that in addition to the IDLEIF SOFIF registers? USB ---4.7k ----pic input --100k---GND?

[Darrel Taylor]

The only case that is still at issue for me is that the device doesnt go passed the "DoUSBinit" subroutine if it's powered on while not connected

Yeah, when it's not connected, it'll never make it to the CONFIGURED_STATE.

Change the Init routine to this.

DoUSBinit:
    pause 500
    usbinit                  ; initialise USB...
    USBSERVICE               ; service it once
    UIE = $7F                ; enable USB interrupts
    UEIE = $9F               ; enable USB Error interrupts
    @ INT_ENABLE USB_INT
return

[RYTECH]

Thanks Darrel that did it for me!

[RYTECH]

Does the USB operation of the 18F4550 have a direct effect on the A/D channels..? Let me explain why i ask....

Case a) when i power up my circuit the A/D values float - not 200 or 300 millivolts - which for my inexperience, hackjob craftsmanship, and source voltage reference would be acceptable, but sometimes, usually when i power up without the usb cable connected, the values float sometimes as much as 800 millivolts on my 12v line, i've metered the several points, there's no floating, so it's something to do with the IC's operation. I can usually correct the issue by plugging in the device and unplugging it. I'm stumped as to what exactly is going on... i'm wondering if this 800 millivolt swing has something to do with the pic enabling or disabling its internal pullups on the usb data lines. it seems to revolve around the state of the usb connectivity... when the problem is not present my readings are mostly acceptable, the caps and aquisition times seem to work ok.

wonder whats going on... i think i chose the wrong IC for my solar powered charge controller... I really like the usb connectivity, but the IC with only 1 power led and 1 led that flashes every so often (about 1/16 cycles through main) the circuit draws about 120 mA... is that around normal for not having any sleep cycle in main? I tried to activate the WDT to implement a sleep but it seems the USB bootloader (or something) doesnt like it, cause when i reset the chip after the WDT enable change the USB reports the device is not recognized....

[RYTECH]

so i guess i actually have 2 issues still, i cant sleep (neither can my device) and the A/D sometimes acts "wonky" depending on the state of the usb pluggedness.