Jump to content
IGNORED

Fixing a dead 130XE board, which bank of RAMs to fill first ?


Level42

Recommended Posts

I've had a dead 130XE board laying around for many years. Not sure if I previously worked on it (cutting out some dreaded mT RAMs) or that I got it is way but there were already 3 or so RAMs cut away.

 

Just for fun I powered it up but it gives a black screen only.

 

Since I bought a hot air soldering station recently for insane little money I figured I'd give it a try to remove all the RAMs.

 

This was the first or second I tried....would have gone a lot better without the iPad in my way and all the others were out in no time...

 

 

I also managed to remove ANTIC this way but I was much more careful about that one of course. Teste the ANTIC in a 800XL and at least that one works just fine.

 

Now my question: can I fill up just one bank of RAM (or maybe just a single RAM chip) to further diagnose the board as I have a feeling that there might be more issues than just the RAM.

 

If so, which bank to fill ?

post-25272-0-93277500-1508569233_thumb.jpg

post-25272-0-26195600-1508569254_thumb.jpg

  • Like 1
Link to comment
Share on other sites

Each chip only supplies a single bit on that system so you need to populate the entire column.

 

Fairly certain the outside 8 chips represent the normal 64K so you'd want to put chips in there for the system to be able to start normally.

Also pretty sure the inside column can be left empty although given the machine has an EMMU and is setup to operate as 128K it wouldn't behave the same as a 128K machine if extended Ram access was attempted.

Normal result if an extended bank is accessed on a 64K machine is that main Ram will be accessed. But with a depopulated 128K machine you'd likely get $FF or address echoes returned.

 

Most software should still work fine though.

  • Like 1
Link to comment
Share on other sites

Excellent, thanks For the quick reply Rybags :)

 

It's just for testing if the main (hardware) stuff works.....

 

If it runs on a single bank the RAM was the only issue.....and I will fill the second bank then of course.

If it won't run with one bank I can still test mos of the rest of the system.

If I manage to fix it then I will still populate the second bank of course.

 

But I want to socket everything I replace on it naturally and saving (temporarily) 8 sockets is a good thing.... also time wise :)

 

 

Hmmm....thinking about the RAM again: I actually think it IS possible to put in just one RAM ic. Of course, the machine will go straight into the memory test, but it will show one good and 7 bad RAMs......or does the men test require a minimum of RAM to work ?

 

I remember getting this effect on a 800XL late model with mT RAMs.....

 

Isn't it also usual to get a red screen if there is only a RAM problem ?

Edited by Level42
Link to comment
Share on other sites

Mm, I can simulate this on a 800XL of course.....remove all RAMs but one and see what it does....

Nope....that doesn't work...I guess it depends on WHAT is defective an a certain RAM..

 

 

Just for fun, I just tested all the RAMs that I removed on a 800XL (one by one of course).....it's amazing ALL but one are dead......guess which one is the working one ? yes indeed....the NEC !!!!

I realize I probably put more heat on them then they are designed for (SMD is designed for hot air, through hole parts are not and RAM is probably pretty sensitive) but it IS pretty indicative that it is naturally the only non mT RAMs to have survived....

 

Were those RAMs actually really manufactured in the USA as the state on the case ? If so, it makes you wonder how they screwed up.....or could this be an example of early counterfeit chips ?? I wouldn't be surprised.....Tramiel bought whatever was cheapest...as long as the thing worked for a while and Mr.Tramiel had your money in his pocket hewas happy as he had already showed at Commodore

Link to comment
Share on other sites

Thinking about that all but one RAMs are dead ......i think it's not unlikely that someone supplied some incorrect voltages to this board....or maybe the PSU crapped out at one point in time...

 

I mean, I came across a 800XL with one bad mT RAM, but all others were still,working.....A 100% death rate in one machine isn't that likely, even with mT RAMs....right ?

Link to comment
Share on other sites

Haha, yes for me the marking mT an an Atari RAM is enough to get rid of it......well I did keep a bunch working ones as a back-up just for....you neve know and I bet they come in handy now just for testing,

 

 

By the way I LOVE that hot air solder station....should have bought one years ago.....(but they weren't as cheap then).

I started removing all solder from all the RAM holes using my trusted solder wick and iron but then figured there would be a faster and cheaper way....I simply fluxed two rows of pads, got all the pads solder melted with the hot air gun and "gently" whacked the PCB to my desk....the result was a bunch of solder rests on my desk and all,holes open nice and clean....went SO fast !!!

 

Now....if this board runs again I'm tempted to socket everything......just because :D

Edited by Level42
  • Like 1
Link to comment
Share on other sites

Interesting video on using that little hot air station. I hope you figure a way to remove the chips without damaging all those little capacitors along the way though. It's bad enough to have install all those little sockets, but sourcing and replacing a bunch of tiny little axial caps would be a real pain in the ass.

  • Like 1
Link to comment
Share on other sites

Well actually those weren't up against the heat....most of them broke of between the casing and the pin...kinda weird....

But they don't have to be axial, those caps,are very simple ceramic filter caps, any shape wil be fine.....you buy these things like a 100 for 50 cents.....so I actually removed them all, even those who did not break during the process as I wouldn't trust them anymore anyway......

Link to comment
Share on other sites

You should equally be able to non-destructively remove those Rams, ie don't cut pins, just desolder in place. Though I guess timewise it might be quicker the other way.

Oh but I did ! The video is really a bad example...and even that one I got back in shape where it concerned it's pins.....

Edited by Level42
Link to comment
Share on other sites

I like reading about projects like this because I have a 130XE with RAM issues on my to-do list. I have been putting off a repair attempt due to the bad reputation of this particular PCB.

I expect to end up removing all the MT chips. Oddly enough, there was only one non-MT dram on this board as well. Top chip of bank 1 is NEC.

 

Hot air guns are indeed much cheaper these days. Any tips on using one on a 130XE to avoid damaging the delicate board and traces? I'm considering that route instead of iron and wick.

Link to comment
Share on other sites

True.

 

This is a green board and there were absolutely no problems with the traces. I set the temperature at 370 degrees (temp that my Weller always uses...) and flow pretty high, almost max. I used a pretty narrow nozzle to concentrate the heat around the pin rows...

Key is to keep moving the nozzle.

 

The board browned a bit on some places but nothing bad.

 

In fact, I just removed the large chips, a cartridge connector and the small Atari C0 chip from a Tramiel era 800XL PCB and this was a lot harder to do. Luckily I didn't have to care for the board anyway....All large chips tested and working but I really needed a lot of time/heat to get the solder flowing and I sure did use plenty of flux (which is absolutely REQUIRED !!!!, No flux = failure for sure). With the smaller IC the board developed a kind of "bulb" under the solder mask I guess....but as I said, the board was written off before I started. But the 130XE I worked on had much easier running solder...that 's for sure.

Edited by Level42
Link to comment
Share on other sites

Pics after removing all RAMs and caps.

 

The browning is probably only the solder mask....nothing structural. It's not flux residu....I thoroughly cleaned everything with alcohol and paper towel.....even tried a magic eraser (sponge) a bit....

post-25272-0-19540800-1508631787_thumb.jpg

post-25272-0-34265300-1508631808_thumb.jpg

Edited by Level42
  • Like 2
Link to comment
Share on other sites

Proceed with caution. It's not too hard to blister and delaminate these old boards using hot air. Some boards can take it and some can't.

Definitely will. First thing I noticed when I got this 130XE is how clean and new the case and PCB looked. I would venture a guess that something went wrong early in its life before it was boxed and stored away.

 

The power supply that came with it tests good, so no idea what damaged the memory..possibly just poor quality ram.

 

I may get a Hakko FR-300 when finances allow, and I convince myself that there will be future projects which merit the expense! ;)

Link to comment
Share on other sites

Proceed with caution. It's not too hard to blister and delaminate these old boards using hot air. Some boards can take it and some can't.

Very true Jon, but I've had traces lift up with arcade PCBs by using soldering iron as well....

 

In theory all PCBs should survive at least some heat as boards were wave soldered back in those days....

 

I bought this hot air station. It's sold under a number of brands but usually with the same model number. This one has a fairly decent quality PCB inside, there's lots of info about it on the internet, incl, firmware hacks etc.

 

http://s.aliexpress.com/i63meiiM

 

I must say, for the money, it's an amazing quality.

Edited by Level42
Link to comment
Share on other sites

By the way, I already did a much better job on the ANTIC on the 130XE board. It's also a matter of getting some experience I gues, the RAM s were my first try with this method ever....

 

There's absolutely no browning to be seen there and the chip came out beautifully :)

Link to comment
Share on other sites

By the way, I already did a much better job on the ANTIC on the 130XE board. It's also a matter of getting some experience I guess, the RAMs were my first try with this method ever....

 

There's absolutely no browning to be seen there and the chip came out beautifully :)

post-25272-0-25474200-1508661833_thumb.jpg

post-25272-0-91981300-1508661850_thumb.jpg

  • Like 1
Link to comment
Share on other sites

Very true Jon, but I've had traces lift up with arcade PCBs by using soldering iron as well....

 

In theory all PCBs should survive at least some heat as boards were wave soldered back in those days....

I'm really talking about the fibre layers of the board blistering and separating. An XEGS to which I was forced to apply hot air the other week seemed indestructible (although the board had a lot of "give" in it when hot, before stiffening up again once cool), but prior to that I attempted to remove a keyboard connector from a donor 65XE board and the area around the connector was utterly destroyed. On the face of it, hot air is a neater and possibly quicker way of removing ICs, but there's a reason you can buy $200 electric pump thru-hole desoldering tools. Wave soldering doesn't put sustained, localised heat into the PCB.

  • Like 1
Link to comment
Share on other sites

Yes of course, these boards were never designed for hot air....a heating plate (hot plate) which pre-heats the entire thing to a reasonable temperature (like 100 degrees) would be nice, but are crazy expensive.

 

Totally understand what you mean about the blistering.....this is the 800XL Tramiel era board.

 

The last picture shows where i surely overheated it, the honeycomb structure shows it...see arrow.

post-25272-0-37749600-1508669665_thumb.jpg

post-25272-0-23948100-1508669678_thumb.jpg

post-25272-0-99204200-1508669696_thumb.jpg

Link to comment
Share on other sites

The main difference for sure was either the solder they used at the factory or the way that both boards spent their lifetime and maybe environmental issues influenced the condition of the solder....fact was that the 130XE's solder ran much easier and especially earlier.

 

In fact I was amazed that all the chips survived the process (tested them).

 

Note that the 130XE board was just laying around....spent years and years on my attic and was full of dust before I began.

I gave it a nice luke warm bath using water and dish washing soap.....and that's another great use of a hot air station: I set it to the lowest temperature (100 degrees) and max. flow to,do,the final drying of the board. It managed to blow away the last rests of water which otherwise hangs between rows of resistors and IC pins because of the capiler effect. Also works great to make sure the connectors and colour pot were al dry right away.

 

Sadly there's no telling if this board was manufactured later or earlier than the 800XL board and also no idea of where it was made.

 

 

They should fire the guy who fucked up the Atari logo like that though......oh wait.....:D :D

post-25272-0-59340900-1508670222_thumb.jpg

post-25272-0-45591600-1508670237_thumb.jpg

Edited by Level42
Link to comment
Share on other sites

P.S. as you can see I also salvaged the cartridge connector from the 800XL board.

I was afraid the plastic could melt but it survived without a problem. The pins are extremely long on this connector so it was heating while gently pulling it out bit by bit....

 

Any tips on removing the mechanical joints of the SIO connector ? Want to remove that too (and the joystick connectors).

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...