Atari 130XE Glitchs and Hang

[enito]

Sure, i take some photos of different angles, hope can see fine. I put flux in bad solder you say me...greets and thx!!!

 

[1050]

Thanks, the MMU socket pins sure look a lot better now, but now I see a single wire hanging off pin 17 of the OS rom mighty close to pin 18. Do you get any beeps from your beeping multimeter between pin 17 and pin 18 when there is no chip in the OS socket? Might you consider changing your two red wire jumpers for a single wire jumper across all three contact points? Solder top and bottom connections first with just enough rope to reach top and bottom thru vias and then just melt the insulation away from the wire at pin 17 near the end of the pin so it doesn't even get close to board at that point, and make the connection that way? This would insure no stray wires dangling into areas they shouldn't be. I see you reworked pin 17 and I don't see the orginal bridge anymore, but that repair jumper should be done cleaner. Data line 7 and Address line 11 (pin 17 and pin 18) crosstalking just could make a pattern of failure like you saw with the memory tester, and/or the patterns on the screen when running Demos - I was sure this was going to be it. I keep hoping for good news...

[1050]

Was also thinking that since it's pretty easy to do you might want to check continuity for all side by side pins on those installed sockets without the chips in them. We can see the bottom side pretty good but are blind on the top side, this board is not right somewhere, odds are it's where it's been modified. Don't worry about late reply - life has a way of happening, I have one, you have one.

[enito]

Hi sorry, for this late, some stuff in real life, but i back. I do a better joint point, but problem continue selft test in boot :/ no basic. I re tested with multimeter and no join. I attach photos of the work..

 

 

This one i gotta win one day i have other ataris, but wanna learn how repair this one greets!

[1050]

You should have corrected me for my wrong method of pin counting much earlier, how embarassing. I've been calling out the wrong pin numbers from the start. Pin 17 by my old count is actually pin 26, Address line 13 going to the MMU, CPU, ANTIC, and Freddie as well as the OS rom where the damaged trace is.

 

Extra length of new jumper on top bothers me some over by the MMU, you could have just poked it into the hole of that via to take up the extra length instead of having it dangle over the other trace, too easy to make contact with the other trace and you can't see the contact if there is contact sometimes.

 

There appears to be an unusual amount of solder around pin 22 of the BASIC rom viewed from the top side, can you work underneath the board holding it in it's normal position and reflow that pin in the hopes that the excess solder will wick down the pin on onto the soldering iron by gravity? Another blob at pin 26 in the OS rom socket from the top view seems to have run down as far as the pin shoulder on the chip, but this shouldn't cause any crosstalk, wouldn't hurt to reflow it to remove the excess either? Please remove chips before soldering or they may not want to come out later...

 

Would bending the socket pins outward while hot with the iron allow the solder bulge on the pin to move away from that one bare uncovered trace near pin 10 of the MMU - they run very close together there and with the trace being bare like that it's just asking for crosstalk trouble. Nail polish would fix the uncovered part, but I fear that there is flux in the area and that might causing the crosstalk to begin with - it really needs to be wire brushed absolutely clean with solvent before applying nail polish as doing otherwise would seal in the conductive gunk which is exactly what we don't want to do. But, it's also too close to keep the polish away from the pin unless you removed the solder with some braided copper desoldering wick first. Then the hassle of scraping the dried nail polish from the pin and via it lives in as well not to mention resoldering. You don't have to do any these things on my account, just pointing out trouble spots as I see them. If you are sure there is no continuity there then fine, just know that it sure looks like a real trouble spot in my opinion, and I WOULD fix it if it were mine.

 

Have you tried cleaning the board with a toothbrush and automotive brake cleaner in a spray can or similar high powered solvent? Shoot it under the sockets and scrub the pins on the bottom real good with the toothbrush dripping solvent to get rid of flux residue which can be conductive. Then rinse and let dry completely. In difficult places like the bare trace running so close to the MMU pins mentioned above, I've even used a small fine bristled stainless steel brush to clear out flux gunk that sometimes just won't move any other way. Once you get it cleaned and I learn how to count, things are bound to improve? I do like the new wire choice much better, if it's tinned under the insulation then you have the perfect wire to solder on top of slightly ruined traces, once you pull the insulation off. And your soldering is pretty much top notch now, nice and bright just like it ought to be.

[enito]

 

ou should have corrected me for my wrong method of pin counting much earlier, how embarassing. I've been calling out the wrong pin numbers from the start. Pin 17 by my old count is actually pin 26, Address line 13 going to the MMU, CPU, ANTIC, and Freddie as well as the OS rom where the damaged trace is.

 

 

Yep, i notifyed with some numbers of pins added in photo, for to be clearify about pins, but i understand about this little wrong number. no problem About check i checked with multimeter all continuity about 26 to A13 in MMU,CPU,ANTIC and Freeddie OS too and all fine.

 

 

There appears to be an unusual amount of solder around pin 22 of the BASIC rom viewed from the top side, can you work underneath the board holding it in it's normal position and reflow that pin in the hopes that the excess solder will wick down the pin on onto the soldering iron by gravity? Another blob at pin 26 in the OS rom socket from the top view seems to have run down as far as the pin shoulder on the chip, but this shouldn't cause any crosstalk, wouldn't hurt to reflow it to remove the excess either? Please remove chips before soldering or they may not want to come out later...

 

Done i do a better reflow in this area (photos in the end), And i care about remove chips before all.

 

I checked too the pin near of 10 MMU and better reflow. I applyed now with ISOPROLYC AlChool with teeth brush clean, chip inserted, check, same thing self test I tested too another BASIC same, MMU too. Now i have i doubt if a way for check if Freedie is working fine? some voltage check or something? Antic i doubt have troubles, OS too.

 

Here the photos, thx for your big patience, but i think this thread can help to me and others learn more about fixing this jewels

 

Greets

 

 

 

 

 

[enito]

Hi again i discovered something in Freddie chip with CAS pin. I checked pins using schemas of freddie and found the pin 35 (CAS) goes to memory area (PIN 15) (CAS) not have continuity. For to be sure , this one is need to be fixed? can be this the trouble? ..hope yes.

 

Greets!

[1050]

I have pin 6 at the second memory MMU U34, a CO25953 being connected to Freddie CAS\ pin 35, NOT to pin 15 of U34 as you are stating. So probably no joy there.

 

If you can find 93% or higher alcohol at the same store you regularly buy it at, you will find it's three times more effective at the very least in removing difficult gunk, perhaps a pharmacy? Standard 50% is pretty weak for cleaning in comparison. I didn't even know they made or sold a higher grade for most of my life.

 

I'm begining to think the problem is under those added sockets, with your new skill set this should only take some time to remove, clean well, inspect very closely, and do again? In agreement with this thread helping others with their soldering skills too. The best test for Freddie I can think of is to swap resistors 110 and 111 on their output sides which trades main bank for extended bank as outlined previously on page one of this thread. The memory test should give same results if Freddie is OK. Never heard of a Freddie gone bad either. Running out of things to suggest.

[enito]

 

I have pin 6 at the second memory MMU U34, a CO25953 being connected to Freddie CAS\ pin 35, NOT to pin 15 of U34 as you are stating. So probably no joy there.

 

Hmmm strange, i do check too from U34, i guided from Schems of 130XE and say freddie have CAS output, thats i thinked was the trouble...bad :/ (Please read next post, some explain i do

 

 

If you can find 93% or higher alcohol at the same store you regularly buy it at, you will find it's three times more effective at the very least in removing difficult gunk, perhaps a pharmacy? Standard 50% is pretty weak for cleaning in comparison. I didn't even know they made or sold a higher grade for most of my life.

Yep i have some lot, cause i work in chemical enterprise and here they can give me some isoprolyc alchool

 

 

I'm begining to think the problem is under those added sockets, with your new skill set this should only take some time to remove, clean well, inspect very closely, and do again? In agreement with this thread helping others with their soldering skills too. The best test for Freddie I can think of is to swap resistors 110 and 111 on their output sides which trades main bank for extended bank as outlined previously on page one of this thread. The memory test should give same results if Freddie is OK. Never heard of a Freddie gone bad either. Running out of things to suggest.

 

Hmm can be, the ones i used in MMU and OS was a little cheap and badly one, instead of basic socket i put lately. The problem of remove with sucker of my air hot station is can be more damaged in remove the traces near. And hmmm freddie if was bad i think computer start only with black screen. I try to check again traces near or something.

 

Greets and thx!

Edited June 12, 2014 by enito

[enito]

I have pin 6 at the second memory MMU U34, a CO25953 being connected to Freddie CAS\ pin 35, NOT to pin 15 of U34 as you are stating. So probably no joy there.

Ehh i mean the freddie CAS area Pin 35 not goes to memory area i mean U9 to U33 Area, not get continuity to CAS of this memory area PIN 15, this one is fine???

[1050]

OK, it's not supposed to though, CAS\ signal from Freddie must first go thru the second MMU memory bank controller, U34. From there it goes to U33 and all of the ram chips in that particular column from the resistor 111. Which happens to be one of the two that can be used to swap memory banks for the Freddie test. At the memory chip pin 15, it's called CAS\ but that is not the signal line named CAS\ as ouput from Freddie, that signal line only goes to U34 and is used as an input there, output CAS\ signal lines from U34 are then named CASMAN\ and CASBNK\ to indicate the main and extended memory bank columns of ram chips that they control. So it's normal to find no connection between Freddie CAS\ and memory chips CAS\ pin 15. You need to be looking at yet another diagram, 130xe_04.gif is the name of mine.

[enito]

I see tonight i check cas from U34 , for to be sure if this area is fine with memorys...

 

yep i have the 130xe_04.gif, only i thinked freddie care too about CAS cause WRT RAS (near of cas) in freddie, goes to mems u9-u33 area.

 

Still fight

[enito]

OK, it's not supposed to though, CAS\ signal from Freddie must first go thru the second MMU memory bank controller, U34. From there it goes to U33 and all of the ram chips in that particular column from the resistor 111. Which happens to be one of the two that can be used to swap memory banks for the Freddie test. At the memory chip pin 15, it's called CAS\ but that is not the signal line named CAS\ as ouput from Freddie, that signal line only goes to U34 and is used as an input there, output CAS\ signal lines from U34 are then named CASMAN\ and CASBNK\ to indicate the main and extended memory bank columns of ram chips that they control. So it's normal to find no connection between Freddie CAS\ and memory chips CAS\ pin 15. You need to be looking at yet another diagram, 130xe_04.gif is the name of mine.

 

Apprently i get good news, i checked MMMU U34 PIN6 to CAS pin15 of set of (u9 to u33) and nothing. Okey i checked from PIN6 OF MMU to U33 PIN 15 (CAS) and nothing. I opened a 65XE i know is not the same machine , but very similar, and for start i checked cotinuity of freddie CAS to PIN5 CAS of mems (u9-u33) and i get continuity. I think the cas way is broken and need to be repaired...is correct master?

 

Greets

[enito]

Edit for add, CAS Freddie to CAS MMMU have contact with multimeter, only this ones are lost with u9 to u33 pin 15.

[1050]

Not master and NO, you seem to be stuck barking up the wrong tree. Freddie makes the CAS\ signal and sends it to U34 for further processing and it comes out of there as TWO possible signals which are now renamed to CASMAN\ and CASBNK\ - CAS\ no longer exists. So you won't find it going to the memory chips no matter where you look for it. BUT CASMAN\ goes to one column of ram chips and CASBNK\ goes to the other via resistors 110 and 111. 65XE doesn't have U34 so it's Freddie CAS\ signal of course goes to the ram chips directly. The job of U34 is to decide which column of ram chips will be accessable by either or both of the CPU and ANTIC depending on how port b is set. This is done by denying the CAS\ signal to the ram chips that will NOT be read or written and allowing the CAS\ signal to go only to the column of ram chips that will be read or written. The chips that do not get the CAS\ signal are then in refresh mode and can not be accessed. This access or no access is done inside each ram chip as designed, they are quite clever little bugs in fact, a lot going on inside of them too.

 

Pin 6 of MMU to Freddie CAS\ pin 35 has continuity means good

 

Then pin 11 of MMU to pin 15 of U26 thru U33 ram chips has continuity means good

AND pin 10 of MMU to pin 15 of U9 thru U16 ram chips has continuity means good

no problem, quit trying to find one here. Waste of time, the machine would never boot at all if there was something seriously wrong here.

 

These last two is where the CAS\ signal from Freddie went to, but only one at a time can be used which is the job of U34 to direct traffic of data in and out of both sets of memory depending on how port b is set to operate the memory banks.

[enito]

Right, i see, leaerning i continue finding the wrong stuff...and take it easy, only i read the schems and thats, now i know cas is not read from freddie

[atari8warez]

Sorry for barging in, but from the photos I've seen, I tend to agree with 1050 on his comment that the motherboard area and the traces and vias under the installed sockets may to be the source of the problems.

[enito]

Sorry for barging in, but from the photos I've seen, I tend to agree with 1050 on his comment that the motherboard area and the traces and vias under the installed sockets may to be the source of the problems.

No, welcome hmmm can be but checking this traces with multimeter without the chips i get good contacts... i checked more of twices ...strange...

[flashjazzcat]

I'll bet there's a short circuit somewhere under those sockets. I'd apply no-clean flux and heat the pins up again with the iron.

[enito]

I'll bet there's a short circuit somewhere under those sockets. I'd apply no-clean flux and heat the pins up again with the iron.

thats a good idea, sorry for my ignorance can you show me example of no-clean flux?

is like a paste?

[flashjazzcat]

Just type "no clean flux" into eBay. You'll see it in small bottles or pens. It's a liquid which helps keep the job clean and ensures the solder sticks to the metal parts. My thinking here is that you have some tiny bits of solder bridging pins under those sockets. Just a theory, but worth a try.

[enito]

thx

[1050]

Yes, thanks to everyone commenting, more than welcome. Desoldering can be done with desoldering braid also which gives you the same tempurature control that you have on your iron which appears to be more than adequate for 130XE work. You want to be sure that the braid is clean or otherwise fluxed well itself in order for it to draw the solder from the pin and hole you are working on. Often works well to push a corner of the braid right in the hole with the iron in order to get the braid to start drawing solder away. And you should always double check that each pin can be wiggled in it's own empty hole after cooling off and before trying to lift the socket - rarely the socket will fall out on it's own, but most of the time a pin or two wants to stick anyway. Just be careful and don't apply too much pressure and so traces and vias don't get pulled loose. On the other hand, life happens and traces are pulled - I still think it needs to be done.

 

I think we are dealing more with a gob of crud with some minor conductivity across pins rather than a genuine solder short which sould sound off with the beeping multimeter (got to get one of those) being used, a weaker connection like I suspect may not be enough to sound off the mulitmeter. I know from my work using acidic flux, it must be CLEAN, CLEAN, CLEAN or it won't fire up.

[enito]

Hi, today i started removing the U3 (One of the bad quality socket installed) , and dont any damaged trace or something. Is need to removed the pins , but the hard work was made. Here i add a photo. Next i try to remove U5 (too bad quality socket). Basic was installed a good quality socket, dont think this one can have trouble.

 

Greets

 

[1050]

Good man, I see that unmoved jumper I spoke about a few messages back. Unsolder it and stick it in the hole in that same via so it doesn't overhang that other trace and I'll quit nagging. Wouldn't take 10 seconds with warm iron and would make the job look better, while giving me peace of mind.

 

Using just this photo, blow it up and note the little solder overlay between pin 3 and the trace running past it from pin 18 - that's all it would take to make it go south like it's doing. If it's going to happen it will happen at a very close spot just like this, so they deserve extreme scrutiny. BUT I also see similar white patches in the same areas of other vias that must be reflected light and so this is probably a false alarm since I'm not able to turn the board to a different angle myself and verify that there is no solder there in reality. We can't say that about what is happening under that overhanging green jumper wire, but I do see 10 mm of crud along the wire itself, this should be cleaned until there is no sign of crud there or around the via.

 

You can cheat in this sort of situation where now you need good clean holes to put more pins back in and I would use a properly sized drill to just drill out the solder even in socket vias that are relatively clean already. This saves heating the board as well. For acytelyene\oxygen welding they sell tip drills so one can drill out the gas holes that often become plugged, the set comes with a pin vise about the size of a large pencil to grab the drill and then you turn it with your fingers to drill into the hole. The drills are then kept inside the pin vise and there are several sizes to choose from, very handy for just this kind of work. You do not want a drill size that cuts the board, just the solder in the hole. The drill should easily fit into a good open hole, this way you won't be drilling out your vias which is not good at all.