Writing in EEPROM and reading by 6502

[Chromatix]

I think your code is probably correct, given the very simple hardware configuration. Just tie off those input lines and try again.

[hitlp]

Chromatix wrote:

I think your code is probably correct, given the very simple hardware configuration. Just tie off those input lines and try again.

I tried now and got a similar garbage.

[Chromatix]

Okay - so we should take another look at your physical circuit layout. You now have a second oscillator in use, and we know that clock edges are a sensitive subject.

[Dr Jefyll]

hitlp wrote:

Chromatix wrote:

Just tie off those input lines and try again.

I tried now and got a similar garbage.

Did you tie them high or low? Either way is OK, but it's info that needs to be taken into account.

[hitlp]

Dr Jefyll wrote:

hitlp wrote:

Chromatix wrote:

Just tie off those input lines and try again.

I tried now and got a similar garbage.

Did you tie them high or low? Either way is OK, but it's info that needs to be taken into account.

I tied A13, A14 and A15 at low (GND).

[Dr Jefyll]

More questions. Sorry!

hitlp wrote:

I'm using the AE29f2008-12 which is a memory flash removed from a PC BIOS. The pinning is very similar to the AT28C256.

Except the AE29f2008-12 has one more address input than the AT28C256, is that right? Just checking.

Quote:

since the 6507 only has addressing up to A12, I connected this port to 6551 as if it were the A15 of 6502.

Alright, we use A12 as the substitute for A15, and you say you attached A12 to the 6551. Did you also attach A12 to pin 3 of the inverter? (both changes as shown in the diagram upthread)

And can you also tell us please what assembler you're using. Also, what do you use to load the AE29f2008-12 with the file produced by the assembler? And what sort of file is it?

[hitlp]

Dr Jefyll wrote:

More questions. Sorry!

You can ask all questions that you need. lol

Quote:

Except the AE29f2008-12 has one more address input than the AT28C256, is that right? Just checking.

Actually the AE29f2008-12 has 3 more address input that the AT28C256. AE29f2008-12 have A15, A16 and A17. All those pins I've attached on GND.

Quote:

Alright, we use A12 as the substitute for A15, and you say you attached A12 to the 6551. Did you also attach A12 to pin 3 of the inverter? (both changes as shown in the diagram upthread)

I attached as below:

6507 Pin A12 -> 6551 Pin 3 CS1;
6507 Pin A12 -> SN74ls04n Pin 3;
SN74ls04n Pin 4 -> AE29f2008-12 Pin 22 CE#
SN74ls04n Pin 4 -> AE29f2008-12 Pin 24 OE#

Quote:

And can you also tell us please what assembler you're using. Also, what do you use to load the AE29f2008-12 with the file produced by the assembler? And what sort of file is it?

I'm using ca65. This is my Makefile

Code: Select all

SOURCES = kernel.s

%.o: %.s
        ca65 --cpu 6502 -o $@ -l $(@:.o=.lst) $<

all: kernel

kernel: $(SOURCES:.s=.o)
        cl65 --cpu 6502 -C kernel.cfg -t none -o $@ $^

clean:
        rm -f kernel *.o *.lst

The result is a binary named just as "kernel". I opened this file on HxD program and I'vo got this result. I thinking that is ok.

captura02.PNG (9.73 KiB) Viewed 1027 times

captura03.PNG (3.79 KiB) Viewed 1027 times

[Chromatix]

The 6551 has /CS1 (active low) and CS0 (active high). Presumably you've tied CS0 high?

So you have an assembled object file. How is that getting into the ROM? And given that you have a 2KB address range mapped to the ROM, why is your image 8KB long?

[hitlp]

Chromatix wrote:

The 6551 has /CS1 (active low) and CS0 (active high). Presumably you've tied CS0 high?

Yes .. I put CS0 on VCC.

Chromatix wrote:

So you have an assembled object file. How is that getting into the ROM? And given that you have a 2KB address range mapped to the ROM, why is your image 8KB long?

It is a good question .. But in this case, the 6507 doesn't ignore the rest oh unused ROM?

[Chromatix]

It does. So think about exactly what vector addresses it's reading, and what happens next.

[Dr Jefyll]

You haven't answered all the questions, hitlp. What do you use to load the AE29f2008-12 flash ROM with the file produced by the assembler? I'd like to ensure that that tool is being directed properly -- that you're telling it to do the right thing.

[hitlp]

Dr Jefyll wrote:

You haven't answered all the questions, hitlp. What do you use to load the AE29f2008-12 flash ROM with the file produced by the assembler? I'd like to ensure that that tool is being directed properly -- that you're telling it to do the right thing.

Hi Dr. Jefyll. I forgot to answer. I use the XGecu TL866II Plus USB recorder with Xgpro software version 8.30.

[hitlp]

Hi guys,

I was thinking here, the 6502 processor has as the RESET VECTOR the address FFFC (1111 1111 1111 1100) and FFFD (1111 1111 1111 1101), both needing 16 bits for correct addressing?

Since 6507 has only 12 bits of address, it is correct to state that in this case, the RESET VECTOR would be the FFC addresses (1111 1111 1100) and FFD (1111 1111 1101)?

[Chromatix]

Right, the highest 4 bits are simply discarded before they reach the outside world, and the ROM will see them as whatever you've strapped them to. So again, think about where in the ROM the CPU will read its vectors from, and whether your image matches up with that. (Hint: it doesn't.)

[hitlp]

Hello guys. It's me again.

I came back with a new 6502 processor and by the time I communicated with the ACIA (6551) it worked correctly. Now I'm trying to do a serial output with "Hello World!". I'm trying to use Putty but I do not see anything in the text output. Using the logic analyzer, every time I press the 6502 reset, it "spits" the same output. But I can not analyze it.

What can it be now?