nmiy-0020

[joemccarr]

Do you sell a programmed NoICE eprom. I am still trying to
determine if my board is defective or if its the eprom.
The board seems to work ok with the BUFFALO Eprom.

Thanks
Joe McCarron

[nmitech]

? - Do you sell a programmed NoICE eprom.

>> Sorry we don't sell eprom programmer. You can get it at www.digikey.com, or www.mouser.com


? - I am still trying to determine if my board is defective or if its the eprom. The board seems to work ok with the BUFFALO Eprom.

>> You can use the Buffalo monitor to check out hardware. Use BF(block field) and MD(memory dump) commands to check out the memory onboard. Use MM(memory modify) to check the peripherals interface such as I/O registers, lcd, keypad, etc... Or use debug commands, TRACE, ASM, ... to debug your program.

>> Tell us more about the non-working eprom. What code did you burn? program address (size, socket)? Onboard Memmory Jumper setting?

[joemccarr]

I may have found a problem with my nmiy-0020.
Using Buffalo eprom (E000 to FFFF).
Using image craft compiler .
Using buffalo i download my S19 to the board. While
I am looking at EXTAL I see a correct clock. When
I type GO 2000 from Buffalo prompt to begin running my
code the clock signal changes to 5 VDC (no oscillation) . I dont know what interaction is happening but I am sure this is a problem.

Is this something you've seen before on this board.
I verified the problem several times and its repeatable.

Please advise
Joe McCarron

[RMDumse]

If your board can run BUFFALO, it really must be working.

The problem of the clock stopping is not surprizing. There is a STOP instruction in the instruction set which turns off the clock. If your program crashes, or calls memory that isn't there, etc., you can hit a STOP intruction.

Do you have RAM at 2000? Do you have enough RAM for the program you've compiled?

[joemccarr]

You are correct.

I reflowed the connections on the crystal, resistor
etc.. and i am getting a constant clock.
My Buffalo Eprom is working but not my noice Eprom.
I have plenty of ram for the code.

Thanks for your help.

[RMDumse]

I think a RESET or interrupt can bring you out of sleep. Since you're not set up for interrupts, that looks like a dead end. A reset does lots more things than just wake up the processor, but it would at least give you one more indication. If you come back up in BUFFALO running fine after that "stop crystal" condition, you'll have one more indication it is something about the program that is crashing.

[joemccarr]

So...it turns out not to be a cold solder joint, but the code.
I was putting my jmp tof_handler (timer overflow interrupt)
starting at $D3 instead of $D0 (Duh!) which is a BUFFALO
pseudo vector. After I fixed this the oscillator didnt stop.

I will confirm that the interrupt is working correctly tonight.

My NoICE eprom, I'm pretty sure, is broke.