600XL Memory Test Issue

[asommer]

Hi all, I've got a problem with a 600XL and I'm not sure how to solve it. When the machine is started up it immediately goes into the memory test and the 2nd ROM test is red as shown in the screen shot.

 

Here's the things I've done:

1. Swap GTIA with 800XL and chip works fine.
2. Swap ANTIC and things work.
3. Swap Pokey and things work.
4. Swap Sally and things work fine.
5. Swap PIA and things work fine.
6. Swap BASIC ROM and things work fine.
7. Swapped OS ROM and the 800XL had a black screen.
8. Performed the composite mod from Best Electronics.
9. Ordered new CO61598B from Best Electronics.

The new CO61598B is giving the memory test failure. I did the composite mod to the 5-pin din and after the mod was done it booted to the Ready screen once, then the memory test failure each time after that.

 

I've tried the new CO61598B in the 800XL and it does the same thing. I haven't tried the CO61598B from the 800XL in the 600XL because I'm afraid that something in the 600XL is blowing out the chip. Is there anything I can check to make sure that there isn't something in the 600XL causing the problem with the OS ROM?

 

Thanks for your help, I appreciate you reading through this long post.

[+Nezgar]

I'm suspecting the OS socket in the 600XL needs replacing, or at minimum some contact cleaner... but that is indeed worrisome if the new ROM has been compromised similar to the original...

 

If you have an EPROM programmer, you could try reading the OS ROM as a 27C128, and verify the integrity of the resulting saved dump.

 

Likewise, a 27C128 EPROM is pin compatible with the CO61598B, so you could test with those programmed with the XL OS in the machine until issue is determined without risking damaging the original.

[shoestring]

Sounds like a bad socket or support logic fault to me.

Check the MMU and the 74ls138 as well.

[asommer]

Nezgar, thanks so much for the reply and advise. I'll definitely look into the 27C128 EPROM.

 

shoestring, thank you as well. I didn't check the MMU before so I'll check that too. After cleaning the socket :-)

[asommer]

So tried contact cleaner in the socket and it did boot to BASIC one time, but then after that the same memory test failure. I then replaced the socket and it got the same memory test error.

 

It's almost like it it's completely cold it'll boot, but once it heats up (or if there's electricity recently through the board) it errors out. Could it be a capacitor problem?

 

None of the caps are bulging or leaking that I can tell.

 

Also, have some 27C128 EPROM and burner on order so maybe it's a ROM chip issue...

[shoestring]

I would check U2 and U3 if you've already replaced the socket.

Edited November 28, 2018 by shoestring

[asommer]

Thanks again shoestring, I appreciate your help.

 

I switched the MMU with the one from a 800XL and the same memory test failure happened in the 600XL. The MMU from the 600XL did work in the 800XL too.

 

I'll report back in a few days when the EPROMs arrive :-)

 

Thanks again!

[+Nezgar]

 

Also, have some 27C128 EPROM and burner on order so maybe it's a ROM chip issue...

When you get the programmer, read out your existing mask ROMs, and compare your checksums against all known official ROM's here: http://www.wudsn.com/productions/atari800/atariromchecker/help/AtariROMChecker.html

 

600XL usually had a Rev 1 XL OS, 800XL usually Rev 2. Rev 1-3 all had the same part # chip though I think. If nothing matches, the ROM is most probably bad.

[shoestring]

've tried the new CO61598B in the 800XL and it does the same thing

 

I missed this.

 

 

 

 

I'm afraid that something in the 600XL is blowing out the chip

 

Check the +5v voltages at vcc on the sockets, then with the machine off measure the resistance between vcc and gnd on the same socket. If you had a dead short the machine wouldn't even power up. Since you replaced the socket, check your work on the solder side.

 

You most likely just got another dodgy os rom or killed it, it's a mask rom so they can't take much punishment at all.

Edited November 28, 2018 by shoestring

[+DrVenkman]

Since the trouble started after you installed a DIN and did a video mod, I'd like to see photos of the board after you did the work; is possible you've got a short somewhere and don't realize it.

 

As for capacitors, if they're not leaking or physically split, it's about 98% certain they're fine. I've got 14 A8 computers, plus two 5200's, two 7800's, and five 2600's. I've had one tiny electrolytic cap go bad in the video circuit of one of my 800XL's, and I've had the poly caps go bad on my Vader. That's it. All the big electrolytic power caps on every machine I've got are fine.

[_The Doctor__]

Nice catch, after the mod... failure... might be miss-connection causing damage.

[asommer]

DrVenkman, I appreciate the reply and apologize for the confusion.

 

The memory error was happening before I did the video mod. I used the TV out with both a 2600 cable and an RCA cable and the picture was super touchy. If I unplugged it and plugged it back in I had to wiggle the cable to get the picture back. So since I was going to do the video mod anyway (though I would liked to have the system working first to rule out any mistakes I potentially made) I went ahead and did the video mod. Since doing the video mod the video out to a modern flat screen TV has can worked absolutely fine.

 

So the memory problem was definitely happening before the video mod. That being the case there could definitely be a problem with my execution of the mod. Here's some pics of the mod area and the ROM socket. If you can determine a problem with the install please let me know. I'd totally like to fix it :-)

 

Also, great to know about the capacitors, after watching a bunch of YouTube videos on old 8-bit computers I didn't want to rule anything out. but I am far from an expert.

 

shoestring, I checked the +5V on the ROM it was good and there is no short to ground. Thanks for that suggestion, I was curious about the voltages too.

[_The Doctor__]

picture is blurry for PBI non component side, verify that solder bridge is supposed to be there near the "fingers" of the card edge... I don't remember it being bridged as it appears..

[+tf_hh]

Interesting issue

 

When I read the posts before right, you´ve changed OS-ROM, BASIC-ROM and MMU already. So I would focus on smaller parts. One possibility - I´ve had this issue some years ago - is a missing or defect (yeah, possible!) resistor. Check R4 (3k, near by the MMU) and the voltage of pin 14 @ MMU. This pin always should read high (something more then 2.5 volts is fine) as long no real PBI device is attached and running.

 

When pin 14 - called MPD, MathPack disable - is left open or grounded, then the address area $D800-$DFFF will be reserved for external use by a PBI compliant device. This will cause in a ROM checksum error, because the floating point math routines located in the same area can´t accessed and self-test will report an error. Check this pin...

[_The Doctor__]

Just looked at another one and it is bridged so that's okay

No Broken socket wipes?

No chip pins bent under or not in the holes?

No shorted or cold solder joints?

No bent over through hole leads touching?

Edited November 29, 2018 by _The Doctor__

[1050]

Change U7 74LS08 for 74HCT08 and give that a try.

[asommer]

tf_hh, thanks for the info about pin 14, I appreciate it. The pin did read a strong 5.09 volts so I guess it's good.

 

_The_Doctor__, Nope, no broken swipes, bent pins or cold solder joints. Also, don't see any through hole leads touching.

 

1050, thanks for the suggestion have some 74HCT08s on order and will be here in a few days.

 

Also, still waiting on some new EPROMs and burner.

Edited November 30, 2018 by asommer

[asommer]

I've received the EPROM burner, but the chips are on back order or something.

 

I'm trying to read the chip, but I'm not sure I'm doing it right. I've got a Minipro TL866 II+ and I set the "Select IC" to 27c128 @DIP28, then when I go to Device > Read it gives me a "Pin Detect error!".

 

Not sure if I'm doing something wrong in the software, or if that means the chip is bad and unreadable?

 

Thanks again for all your help and suggestions.

[shoestring]

Try reading the good one form the working machine and see if you get the same error.

 

Chip is most likely bad or pin might not be making good contact with the device. If you've already tested it in the working machine and reproduced the same error, then there's your answer.

[+Nezgar]

In the minipro software, there's a tickbox for "Check ID." Since it's not a real EPROM, uncheck that box, and try reading again.

[asommer]

shoestring, thanks for the reply. Since I'm new to EPROM burning and the software for the burner I wanted to make sure I'm not making some silly mistake first.

 

Nezgar, also thanks for the reply. I tried with the "Check ID" unchecked, but it didn't work until I unchecked the "Pin Detect" checkbox... totally should have noticed that before. But like I said I'm new to this :-)

 

So I was able to extract a 16k .bin file and ran the MD5 on it and it matches the 06daac977823773a3eea3422fd26a703 - SYSROM_BB01R2. From the comment:

 

"Production ROM, OS from Atari 800XL and early Atari 65XE/130XE"

 

 

I take it that the ROM is for an 800XL and may not work with a 600XL? It could totally be that I ordered the wrong part :-)

 

On the upside with the checksum matching that seems to indicate that it is a good chip... right?

[+Nezgar]

shoestring, thanks for the reply. Since I'm new to EPROM burning and the software for the burner I wanted to make sure I'm not making some silly mistake first.

 

Nezgar, also thanks for the reply. I tried with the "Check ID" unchecked, but it didn't work until I unchecked the "Pin Detect" checkbox... totally should have noticed that before. But like I said I'm new to this :-)

 

So I was able to extract a 16k .bin file and ran the MD5 on it and it matches the 06daac977823773a3eea3422fd26a703 - SYSROM_BB01R2. From the comment:

 

 

I take it that the ROM is for an 800XL and may not work with a 600XL? It could totally be that I ordered the wrong part :-)

 

On the upside with the checksum matching that seems to indicate that it is a good chip... right?

 

Yep, that's a good MD5 for the XL/XE OS Revision 2. Yes, this seems to imply your mask ROM should be good!

 

There's many 'stock' revisions of the XL/XE OS from Atari, mainly 6: 1200XL 10, 11, 600XL/800XL/65XE/130XE had XL/XE Rev 1/2/3, and the XEGS Rev 4. All of these will run just fine in any of those computers. (Specifically, your 800XL OS will work just fine in a 600XL)

 

I'm still running the original MiniPro with the older v6.6 software, and that "Pin Detect" is not there in mine, but good to know it worked!

 

Cheers,

Nezgar

[asommer]

Christmas update,

 

I finally received my 74HCT08 and replaced the 74LS08. Have the same issue unfortunately. The 2nd ROM test still fails.

 

I've burned EPROMs, but I think I've made some type of mistake during the process because when I use one of the EPROMs I get a greenish/black screen and no boot at all. I'm pretty sure I've not burned the EPROMs correctly.

 

I programmed the EPROMs by loading the .bin file I saved from the ROM and clicking Device > Program > then the Program button. Does that sound correct?

 

Thanks all

[Simius]

This issue may be also due to corrupted RAM.

[asommer]

I have the 64k upgrade kit, would you recommend doing that?