XEGS Eeprom problem and memory

[Crc_Error]

So sorry to the moderators and all....I posted this issue in the programming section when I should of posted it here.   This is my issue with a XEGS I am trying to repair...

 

 

[+Nezgar]

@TGB1718

600XL, 800XL, 65XE, 130XE, 800XE use a 16KB OS ROM and a separate 8KB BASIC ROM. So a 16KB 27C128 type eprom is appropriate.

 

XEGS also adds 8KB BASIC and 8KB Missile command to a single ROM, so 32KB, or 27C256 should be fine....

 

27C256 is normally a drop-in replacement with no jumper or pinout changes necessary in the XEGS so I'm not sure why it's not working yet in this case.

 

A 16KB OS should also work, but BASIC and Missile command won't work since they'll mirror parts of the 16KB OS.

[TGB1718]

@Nezgar thanks, forgot about missile command in there

 

[Crc_Error]

Thanks for info.  Yes I am stumped.  I am curious if there might be some other issue on the motherboard.  Also only 40K shows up on memory test. Is the other 24k hidden? 

[+Nezgar]

7 hours ago, Crc_Error said:

Thanks for info.  Yes I am stumped.  I am curious if there might be some other issue on the motherboard.  Also only 40K shows up on memory test. Is the other 24k hidden? 

When BASIC is enabled, it "replaces" 8K of RAM. If you hold option on powerup (XEGS might be slightly different), it should disable BASIC and show 48 blocks in the memory test. The last 16KB of RAM "under" the OS ROM is not tested.

[1050]

When you are selecting eproms for Atari OS replacement, I've found out the hard way, that you want one with a speed of 200ns even though this seems quite slow, they seem to like it much better. I have zero experience with getting an eeprom to ever work in that or any other purpose.

[SoundGammon]

I had a guy make me a custom XEGS OS, he took out Basic & Missile Command and put in Galaxian & Star Raiders!

[+Nezgar]

49 minutes ago, 1050 said:

you want one with a speed of 200ns

Hmm, that's a possibility I didn't about... The 27256 I sent @Crc_Error would have been 120ns, and an a8diag on a 100ns 27128. I have burned many XEGS ROMs that worked successfully and I'm sure they were the same type. Regardless, maybe I'll dig around and see what I have that's slower to see if that makes a difference.

[Crc_Error]

Interesting do you have any 200ns kicking around Nezgar?  I tired it again last night no luck.....But boots up with faulty chip.  

[1050]

So oddly enough 120ns is perfect for a game cartridge but put it in the OS slot and they become quite flaky and often won't boot at all. I can't claim to know why either, just what I've experienced and I see Steve at AtariMax using pretty fast flashroms which I also can't get working in either spot for my own projects. Would really like to know why it is that way. TBH I've done just enough to know I don't have much luck with fast ones and I ran out of steam at that point, if it were more important to me maybe I'd still be working on it. But I'm stalled and not very active any more, so not interested enough to make any of my own progress on that why question. It's more of a curiosity to me nowdays.

[TGB1718]

Had a look at the circuit diagram and chip datasheet and think this problem is caused by the data

being removed from the data lines too early, look at this timing sheet, you will notice the "VALID OUTPUT"

is still valid after the CE & OE go high and although this is a very short period, if the system is not reading

until the end of the valid period, then a faster chip may have removed the data before it has been fully read.

 

Why Cartridges are OK may just be down to the way the MMU handles access, without information on the

internal timing of the MMU it's a bit difficult to see why Cart's are ok.

 

This is my best guess why it doesn't seem to work with fast chips.