Thx, I know the jitter issue caused by the compiler delays, in the past I was using my own ISR based on Bresenham code. It creates a negligible jitter down to 10uS, I think at 48MHz I can go below 1uS without huge issues. Fortunately data has to be sent rarely somewhere between 1s and 10s.For that I want to use the usb_serial library. The major problem I have is that usb_serial does not work reliably for me with PIC18F25k50, with all good help Rob sustained. In my tests the library works reliably on PIC18F2550 or PIC18F26j50 but not with PIC18F25k50. I've so far tested SO and SS packages and two manufacturing lots thinking it is a microcontroller issue. It seems it is not.
On Mon, Aug 2, 2021 at 8:38 AM 'Oliver Seitz' via jallib < [email protected]> wrote: > You're very welcome :-) > > I want to correct myself about interrupts and jitter: I wrote about > +/-2Tcy. That's the hardware part. But then there's JALs interrupt > handling: If, say, serial_hw_int_cts is transmitting a string and has just > finished one byte shortly before the timer overflows, servicing the timer > interrupt will be delayed until serial_hw_int_cts has finished preparing > the next byte, which can take quite some time. Then, depending on the order > of the service routines which you don't have control over, the timer > interrupt can be served immediately afterwards, or, some more service > routiunes are called, control may even be given back to the main program to > execute one machine instruction just to then go back to the list of > interrupt service routines until finally servicing the timer interrupt. > > That means, if there's more than one interrupt in use, servicing can have > a jitter of indefinite amount, even to the point when interrupt events are > skipped. > > Greets, > Kiste > > > > Am Montag, 2. August 2021, 07:18:09 MESZ hat vsurducan < > [email protected]> Folgendes geschrieben: > > > > > > Kiste, thanks for opening my eyes, PIC18F25k50 will do my job, it has > indeed the ECCP, I didn't read correctly the notes 4 and 5 at pin > descriptions. Thank you again. > > On Mon, Aug 2, 2021 at 8:36 AM vsurducan <[email protected]> wrote: > > Hi Kiste, as far as I read my version of k50 (PIC18F25k50) does not have > ECCP it in 28 pin, I have to stick on my available stock of > microcontrollers during the inexistence at this point (Farnell, TME) of too > many options to buy immediately PIC microcontrollers. > > Thanks for your solution, I think I will do it in interrupts using TMR0 > for a short timebase, count it for the needed frequency and toggle the > output pins. > > Yes, indeed a CLC would be helpful with the standard CCP... > > best wishes, > > > > On Sun, Aug 1, 2021 at 5:48 PM 'Oliver Seitz' via jallib < > [email protected]> wrote: > >> Hi! > >> > >> Well... almost all 28pin? Chips with suffixes like k20, k22, k50, k80, > j11, j13, j50 and j80 do have ECCP, chips with suffixes like q40, q41, q43, > q83, q84, k42, k83 and numerous more have CWG and/or CLC modules, which > both allow the generation of those signals using a simple PWM. > >> > >> > >> With the PIC18F2550, I only have one idea... Externally feed the pmw > signal back to pins A0 and A1, set the CVRef module to somewhere in the > middle, set comparator mode 0b110, and set one of the CxINV bits. Then you > have the symmetric signals on C1out and C2out. It's not the most elegant > solution, I admit... > >> > >> Greets, > >> Kiste > >> > >> > >> > >> Am Sonntag, 1. August 2021, 14:58:43 MESZ hat vsurducan < > [email protected]> Folgendes geschrieben: > >> > >> > >> > >> > >> > >> Thx Rob, the bridge mode (enhanced CCP) works only for a 40 pin device, > I have a 28 pin device on my board. Enhanced CCP is available only on large > size microcontroller. :( Almost all 28 pin microcontrollers suffer from > this issue (PIC18F25k50, etc.) > >> Since I need two signals with opposite phases, 50% duty cycle, whose > frequency should be programmable between 1KHz and 2KHz I think they can be > generated in compare mode...or perhaps in software. I also need to trigger > an ADC read during the high level of those pulses from time to time. > >> > >> On Sun, Aug 1, 2021 at 2:17 PM Rob CJ <[email protected]> wrote: > >>> > >>> > >>> Hi Vasile, > >>> > >>> > >>> > >>> > >>> Not sure if I understood you correctly. If you want the PWM signals > to be exactly the same but only inverted you can use the bridge mode for > one of the PWM's that is when one output goes high the other goes low. > >>> > >>> > >>> > >>> > >>> > >>> > >>> > >>> > >>> > >>> > >>> > >>> > >>> > >>> Kind regards, > >>> > >>> > >>> > >>> > >>> Rob > >>> > >>> > >>> > >>> > >>> > >>> ________________________________ > >>> Van: [email protected] <[email protected]> namens > vsurducan <[email protected]> > >>> Verzonden: zondag 1 augustus 2021 10:14 > >>> Aan: [email protected] <[email protected]> > >>> Onderwerp: [jallib] PWM versus enhanced PWM, oposite polarity output? > >>> > >>> > >>> > >>> > >>> > >>> Hi all, > >>> > >>> PIC18F2550, 28pin PIC device has only the PWM implemented, the > enhanced PWM is available on 40/44pin devices only. However, If I want on a > 28pin PIC two PWMs running at the same frequency, with the same > duty-cycle, how can I generate one active high and one active low PWM > signals without using any external inverter? No deadband between the output > signals is required. > >>> > >>> Any experience is appreciated. > >>> > >>> thank you! > >>> > >>> > >>> > >>> -- > >>> You received this message because you are subscribed to the Google > Groups "jallib" group. > >>> To unsubscribe from this group and stop receiving emails from it, send > an email to [email protected]. > >>> To view this discussion on the web visit > https://groups.google.com/d/msgid/jallib/CAM%2Bj4qvHSZrsUtoF_1GcDrNS2_mup_hc5ZbYOuQu4uGw5TYuAQ%40mail.gmail.com > . > >>> > >>> > >>> > >>> > >>> -- > >>> You received this message because you are subscribed to the Google > Groups "jallib" group. > >>> To unsubscribe from this group and stop receiving emails from it, send > an email to [email protected]. > >>> To view this discussion on the web visit > https://groups.google.com/d/msgid/jallib/AM0PR07MB624142E76C9FF412EBE22FC5E6EE9%40AM0PR07MB6241.eurprd07.prod.outlook.com > . > >>> > >> > >> > >> -- > >> You received this message because you are subscribed to the Google > Groups "jallib" group. > >> To unsubscribe from this group and stop receiving emails from it, send > an email to [email protected]. > >> To view this discussion on the web visit > https://groups.google.com/d/msgid/jallib/CAM%2Bj4qstLOMHMLtJcLLEfywQ_%2BTYOWzEUCkLtghwrskyk%3DO%3DAg%40mail.gmail.com > . > >> > >> -- > >> You received this message because you are subscribed to the Google > Groups "jallib" group. > >> To unsubscribe from this group and stop receiving emails from it, send > an email to [email protected]. > >> To view this discussion on the web visit > https://groups.google.com/d/msgid/jallib/592117056.1109097.1627829331026%40mail.yahoo.com > . > >> > > > > > -- > You received this message because you are subscribed to the Google Groups > "jallib" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To view this discussion on the web visit > > https://groups.google.com/d/msgid/jallib/CAM%2Bj4qvV6WzGGmX3qBpyPbGszEk_n7Mc5onbMrtePRbxf2WJ0Q%40mail.gmail.com > . > > -- > You received this message because you are subscribed to the Google Groups > "jallib" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To view this discussion on the web visit > https://groups.google.com/d/msgid/jallib/156178122.1042446.1627882658700%40mail.yahoo.com > . > -- You received this message because you are subscribed to the Google Groups "jallib" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/jallib/CAM%2Bj4qu%2BMue5z%3DB8nrvbdYmV_59N6RPMQVi85WafLs%2BZUNPR3A%40mail.gmail.com.
