Atari 130XE Glitchs and Hang

[enito]

Hi again, ok finally today i finished the install of new socket (better quality ones), was a bit hard some traces cut, but i tryed to make better as i can. I show photos of storie, but bad new, the 130xe continues bad, no basic boot (everytime i interchange basic rom), same effect like before the new socket install. Here photos and comments,

 

 

Here U5 removed, and put a new bridge, replacement of 2 cables bridge. I tested conituity and stuff no bad things. Only point 15 have a broken trace when i removed the stuff, but i fixed with a bridge in back side.

 

 

 

Here soldered, i know was not best work, but continuity works, and dont see shortcut of something.

 

 

Installed socket.

 

Badly the trouble continue :/ a hard work this, but i hope all this photos and way of work, can help too others newbies like me.

 

Greeets and thx for the time,

Btw, Chile 2x0 to Spain today

[1050]

Not sure what I'm looking at when I see at pin 10 and 11 of BASIC that the traces (top view thru socket in last photo) to them appear to have no copper in them and they are just green paint in places? Is pin 10 or 11 of the BASIC socket actually connected to relative socket pins of the OS? They sure don't look like it in this photo, so what causes the illusion if it is one - just curious. In other photo that shows part of this area, it's just fine or so it appears. You should get beeps in all those pins of bottom row of BASIC to bottom row of OS and all of the top row of BASIC except for pin 20, 21, and 24 with top row of OS rom.

 

Pin 15 of OS socket is also supposed to have a connection to the via just below it, one connection is not enough for the rest of the system to see those two pins of the roms. But now I see that the via in question has been reflowed which means you have been there and probably fixed it. It sure is one difficult 130XE, got a spare ANTIC?

[enito]

Hi, okey later today i recheck the pins, but im sure i checked, but i check again np. Now i can spare ANTIC from 800XL? i can remove and install socket.

 

Greets

[enito]

 

Not sure what I'm looking at when I see at pin 10 and 11 of BASIC that the traces (top view thru socket in last photo) to them appear to have no copper in them and they are just green paint in places? Is pin 10 or 11 of the BASIC socket actually connected to relative socket pins of the OS? They sure don't look like it in this photo, so what causes the illusion if it is one - just curious. In other photo that shows part of this area, it's just fine or so it appears. You should get beeps in all those pins of bottom row of BASIC to bottom row of OS and all of the top row of BASIC except for pin 20, 21, and 24 with top row of OS rom.

 

Checked, all fine, apparently was a photo shadow, not show reality but all this pins are fine.

 

Now i removed the ANTIC Chip, here:

 

 

Now in the 20 ,21 area, i see a copper, not sure if this was made in remove action, i need to check conducity of this before of install new socket. Now, i know 600XL have all socketed chips, if possible remove this ANTIC chip, better unsolder one of 800xl? for test?

 

Thx!

 

Greets

[1050]

Get one that is easy since they are all the same and one that works, do test this one in that machine too. Pretty good damage there, so do check it as best as possible. I see part of my problem is your lighting, Half of GTIA is so dark as to not even be there.

 

Took me long enough but I finally connected Google's animated themes all week long to the world cup!!! Time to get a life. Or even a TV.

[Mathy]

Hello 1050

 

You can always click on Google's doodles (the animated themes) to find out what's behind them.

 

Sincerely

 

Mathy

[enito]

Hi again, i checked today cupper traces and all fine, no cut, or something. I installed new socket, clean as possible. Installed a ANTIC chip from 600XL, and .... same selft test no basic :/

 

I attach like all trys photos, Greets

 

 

 

[1050]

Hi, Mathy! What will they think of next? Just when I can almost remember to right click on occasion, now this.

 

I was afraid you were going to say something like this enito, a guy with your dedication deserves a reward like a working 130XE for his efforts. You've replaced everything in the source of the BASIC enable signal possible and it still doesn't want to work right.

 

I do know that my 800XL does like two chips a bit different than the ones it came with. One of those is U18, a 74LS08 AND logic chip used for the phase 02 clock, when I use instead a 74HCT08 with a bit faster rise time and rail to rail output signal, they usually work a tad bit better than before, at least they seem to here. You can try this and see? I've noticed that yours has the factory GTIA timing fix jumpers connected to the U18 chip, these would be best moved to the bottom side of the board when you put a socket in for U18. This is just a shot in the dark mind you, no real reason to expect it to 'fix' this, but it's a remote possiblity just the same. Mainly, I'm out of ideas for anything else to try at this point. Anybody else care to chime in?

 

Just had a silly thought wrapping this up, what if any differences do you see when you boot it with OPTION pressed? Don't have an idea here, much less a theory, just curious. A little late now, but I wonder if this couldn't be a keyboard issue with the function buttons somehow. BASIC is supposed to be on and and fails checksum because the rom isn't turned on because the button says different than it's supposed to. OR the PIA is unable to turn it on (bad PIA). Grasping at straws here and time for bed.

[1050]

Should have done this to start with since it's the easiest to do, measure the voltage on pin 11 of the PIA, U23. If it is 5 volts then it's a bad PIA since the test would set it to low and the PIA is not able to comply with the command. BASIC is failing it's checksum because it's ram and not rom. Replacement PIA number to buy is 68B21 these days since the old 6520 is no longer made. Several mentioned this possiblitly and I completely overlooked just how easy it would be to double check if in fact it IS even working right to start with. My bad there, sorry.

[enito]

Hi Again

 

 

Just had a silly thought wrapping this up, what if any differences do you see when you boot it with OPTION pressed? Don't have an idea here, much less a theory, just curious.

Same effect, selft test :/

 

I try to move the jumper to the downside area like you say. Now,

 

 

Should have done this to start with since it's the easiest to do, measure the voltage on pin 11 of the PIA, U23. If it is 5 volts then it's a bad PIA since the test would set it to low and the PIA is not able to comply with the command. BASIC is failing it's checksum because it's ram and not rom

 

I get very low measure, i take a photo,

 

 

0.15V dont know if this one is correct...

 

Greets !

[1050]

Ok then it appears the PIA is pulling it down like it should, so now do you get the same reading at pin 18 of the MMU? You have the meter set perfect for these measurements. 0.15v is lower than TTL logic levels (0.8v) need it to be, so this is a good thing.

[enito]

Okey, today when i back to home i check

[enito]

Okey, do measurement, and same result 0.15V, i see the continuity of PB1 are good ans voltage too.. I attach photo of measure moment.

 

greets

 

 

pd: i push DH for frozen the moment for take photo. My gf today not here for do the photo

[enito]

I discovered a desoldered one leg of ceramic capacitor, see in the photos, of U18, can be the trouble...

 

 

 

 

No, i soldered the one,

 

 

And same effect, no basic :/

Edited June 24, 2014 by enito

[1050]

Sorry, I got busy. Looks like square one to me again then. That capacitor is probably not needed on the GTIA timing fix done by the factory, the one I'm using for reference has the timing fix but does not have the capacitor where yours does. Try changing out the 74LS08 for HCT08 next? Last shot in the dark for me - I dunno what's wrong with this one.

[enito]

Hmmm this one is strange, dont know what can be the next step...i thinked freddie, but ...if freddie was bad, i dont get power i think.

[1050]

If your meaning was that if Freddie was bad, then you wouldn't get past power up screens, you got that part right. You wouldn't be seeing any graphics, let alone the self-test screen itself. You might see a color but that would be about it and same for GTIA and ANTIC, they are just so involved in the machine that a bad one will NOT boot. Try the HCT08 replacement suggested, it might affect the systems rom timing by sending ANTIC a sooner and cleaner phase 02 clock that allows it to send the same improved REF\ signal to the MMU that is being used for it's clock to enable the BASIC rom a bit faster that it is getting right now. This problem has timing issues symptoms all over it with the odd pattern shown in your first posts and even with only the BASIC rom failing to turn on in time - faster HCT08 may be the answer after all. Once this is done then I'm out of ideas for the next step too.

 

Maybe you said before and I didn't notice, but is this a PAL machine? I assume it is.

[enito]

Ok i try to ordee replacenent HCT08 . No this s a NTSC machine.

[enito]

Inspecting my NTSC machine i discovered dont know if this normal, a area without solder and with this a bridge broken. I marked with black circle the area with doubt.

 

 

Thx

[1050]

He, he. I was wondering if you would ever spot the GTIA timing fix done by the factory. The one I use for reference is done this way too, they take the signal from pin 15 of U2 the LS138 used in combination with the MMU to provide chip select signals and send it thru the normally unused gate on U18 the LS08 AND gate to delay it just a tad on some 130XEs. As I understand it, a batch of new GTIA chips required a late select signal for who knows why compared to earlier ones and this what the factory came up with to provide that. Some 130XE don't have it, some do.

 

I can't imagine this is the problem but if you feel like it you could solder a wire across the drill hole and remove both jumpers from U18, tape the wire ends temporarily and you have the old system back - but if GTIA really does need a slower select signal then you just messed up? Easy to put back the way it is right now though. Jerzy's schematic (02.gif) shows it as if it's not a patch job but it is.

 

What I'm trying to imagine is X number of drill presses each outfited with a precise jig to take the 130XE motherboard and drill one half hole just here. If it was only one drill press and one guy, he was plenty busy!!

[enito]

HI lot of time not write here, but today i have some time for do more tests.. I change from working FREDDIE chip and PIA, all goes to socketed , and ....same results, no basic and selft test :/ now im waiting to arrive the 74HCT08 chips....

[enito]

If your meaning was that if Freddie was bad, then you wouldn't get past power up screens, you got that part right. You wouldn't be seeing any graphics, let alone the self-test screen itself. You might see a color but that would be about it and same for GTIA and ANTIC, they are just so involved in the machine that a bad one will NOT boot. Try the HCT08 replacement suggested, it might affect the systems rom timing by sending ANTIC a sooner and cleaner phase 02 clock that allows it to send the same improved REF\ signal to the MMU that is being used for it's clock to enable the BASIC rom a bit faster that it is getting right now. This problem has timing issues symptoms all over it with the odd pattern shown in your first posts and even with only the BASIC rom failing to turn on in time - faster HCT08 may be the answer after all. Once this is done then I'm out of ideas for the next step too.

 

Maybe you said before and I didn't notice, but is this a PAL machine? I assume it is.

 

Hi, finally chips arrived the replacement for U18, now is need to do it the fix to gtia (the bridges ones)

 

thx