N-Bit_MATH

[keymuu]

There's no way for me to tell what's happening, since your snippets all compile.

If I had your whole program, maybe I could see what's going on.

Thanks Darrel,

the hole program is big, but here it is attached:

Maybe you can see with one small glance what is the cause of the error, the origin (that is me, but...). I sure wish to find the problem, what is causing this mysterious behavior.

Nevertheless, maybe you can see the problem with one glance...

[Darrel Taylor]

I don't think the file was accepted.
Try adding a .txt extension.

[keymuu]

I don't think the file was accepted.
Try adding a .txt extension.

Yep...

Did not check it until now. Let's hope it comes through this time.

[Darrel Taylor]

DEFINE RESET_ORG 1000h ' Room for Microchip USB Bootloader

That's 4K, +28K = 32K
The chip is FULL.

[keymuu]

DEFINE RESET_ORG 1000h ' Room for Microchip USB Bootloader

That's 4K, +28K = 32K

Oops
TRUE ! and thank you for pointing out that.

Didn't know/understand that when the number shown in the lower left corner (Success 28676 : bytes used) of MicroCode Studio did not include those 4k! That explains now really a lot.
Obviously the program is too big

There is not a big brother to 18F4550, that would be otherwise like but with 48 or 64 k of program memory? Or is there?

USB is the reason in the first place to choose 18F4550...

Then of course ...
when it is not possible to continue with a processor with only 32k of program mem then..
How about... PIC18F4620 with 64k of program memory and almost pin equivalent (with 4550) but unfortunately without USB but then again, I wonder how much code it would take if one would integrate USB/UART circuit FT232 unit along with 18F4620.

I suppose that it (FT232) has been used all around the world,
so I will try to look around and see what kind of experience people has...

Any experience with FT232? My only fear is that the "driver" for it in PIC is too big, or maybe the algorithm will for a working system consumes most time, I don't know.
What other troubles would one expect with such a plan?

Let's play with the idea that it already done and working (18F4620&FT232) so would it be possible to use that system with some kind of a bootloader?

[Darrel Taylor]

I don't know what the Microchip bootloader is compatible with.
But the MPS shows these USB parts with 64K or more.



The chips listed that are not supported in PBP 2.60A, will be added in PBP 2.60B.

Note: They are all 3V parts.

Otherwise, you can make your own FT232 equivalent with any USB PIC chip using CDC.
The 13K50 is probably the smallest @ 20 pins, and is 5V.

[keymuu]

I don't know what the Microchip bootloader is compatible with.
But the MPS shows these USB parts with 64K or more.



The chips listed that are not supported in PBP 2.60A, will be added in PBP 2.60B.

Note: They are all 3V parts.

Otherwise, you can make your own FT232 equivalent with any USB PIC chip using CDC.
The 13K50 is probably the smallest @ 20 pins, and is 5V.

Thank you Darrel,

thank you indeed, this looks really promising.
It is really nice from you to present such a table.
Where on earth did you find that? (rhetorical question)

I looked for something like that but did not find anything
besides what I already told (18F4620/FT232).

Do you have an Crystal ball?
Can you tell when PBP 2.60B will be released?
Within 3, 6 or 9 months?

There is of course some problems when changing from a 5V processor to a 3V processor when the environment is 5V. However, that is not a big problem when there are enough pins that are 5V tolerant that can work directly with other 5V circuits.

But on the other hand, I have not done that before,
could you please give some good advice how to do it?
Using two regulators one 3V and one 5V? The 3V reg could be small, something 500mA or even less, maybe just 100mA.
Using only 5V and a Zener diode (very old way of thinking)?
Or maybe some other approach?

Everyone in that list has enough program memory (64k or more) and a RTCC. Meaning that I can discard the I2C RTC and freeing such area of the board, nice. And changing the LCD to an I2C device (Melabs Serial LCD. Is it goog?) there is a possibility to choose 18F26J50 that is a only 28 pin device, so the board is not so populated any more as it is at the moment. Very promising indeed