The Dumb Things I've Done...

[+kheller2]

I thought this would be an interesting and education thread to start about the dumb things one does when working on Atari's.  This isn't meant to be a bashing session, its meant to be a somewhat sympathetic thread and to show that we all make mistakes -- even face palm ones.

 

I'll start with something that i just did this past week!

 

So, I'm working on upgrading some 600XLs to 64K.  It is a rather simple upgrade (2 wires, a jumper and 2 ram chips).  Well, in my haste to make something sexy vs functional I thought I could alleviate said wires on the chip side by just cutting some ground traces to the pins on the solder side and running the two wires needed there. 

 

Off I go cutting away with a model knife and test away with the meter. Odd thing is one of the two pins I have to isolate from ground still is connected to ground!  How can this be?  So I grind away more.  Still no progress.  I then get out the Dremel and grind that thing deep.  It is still connected to ground even though I've completely gouged a hole between the pin and the original ground trace. And then I wonder...

 

Is there something on the socket side under the chips and socket?  Yeah .. dumb me. It is then that I noticed that Atari used some of the grounded pins I was isolating to continue ground on the chip side, even suppling ground to a few other chips!  So I just completely broke this thing.  It is an easy fix, but instead of my having a sexy board, it now has to have those original wires I didn't want on the top side, and now two bridge lines on the bottom to fix what I wrecked.

 

Note to self: Always check the top side for hidden traces.  I normally do, but for some reason I just didn't this time and it bit me.

 

[flashjazzcat]

I did this once years ago when I used to employ an unused gate on the CD4050 to buffer the sync signal on VBXE machines. I forget which machine it was (one of my own, fortunately), but one of the spare inputs or outputs was tied to GND, so I cut the trace before wiring things up. It was only months later that I noticed I'd completely screwed up the legacy video output. When I looked more closely, I'd completely (and completely unexpectedly) isolated the GND pin of the CD4050 by cutting that trace, leaving the whole thing floating.

 

[ryanr256]

I posted this on another thread, but it bears repeating:

 

Installed a Newell 256k memory upgrade into a 1200XL. Didn't put it back together completely, just plugged in the keyboard and LEDs to test it. Connected it to the power supply and turned it on and heard a "pop". No power. Looked it over and didn't see anything obvious. Tried another power supply. Same results.

 

I popped the fuses on 4 power supplies troubleshooting before I realized that the motherboard was sitting directly on the metal shielding. I was so focused on the upgrade being the issue that I overlooked the obvious.

 

Now I have to open the power supplies and replace the fuses.

 

-Bob

Edited March 2, 2021 by ryanr256

[+gnusto]

When thinking of this kind of thing with Atari I always go back to my first example. It was 1980-ish and I was upgrading the memory on my then Atari 400, which was all I could afford at the time. Somehow I had gotten some cut rate single chip solution, and completely absent any reasonable guidance like instructions or a service manual I had my 400 apart and ready to apply my upgrade. It was a solder job to install of course, I dont remember which upgrade it was other than 48K, but my hazy memory said it needed a lot more soldering the more modern ones I see nowadays.

 

Let's just say early teen me was not quite the electrical engineer. I borrowed my dads soldering gun and tools, and off I went with enthusiasm.

 

Cold solder joints doesn't begin to describe what i ended up with (keep in mind, this is on resistor traces on the back side of the board, tiny spacing). I mean these things were arctic joints. They looked like little mushroom clouds on the back of the board. Somehow I had enough intelligence to try to turn it on before I fully reassembled it, and I got nothing, a video carrier but no CPU output. I fiddled with it for a bit and finally got sane and asked my dad for help.

 

There's this sigh that only dads can let out, which lasts for a couple of seconds but manages to to jam in a lot of information, along the lines of "what the hell is this? How did you even think this was acceptable? Are you even my son, who would have never made this atrocity!??". That is the sigh my dad let out when he examined the motherboard. I was pretty embarrassed after he explained how it was wrong, and probably just wanted to be left alone. But in full dad mode, he then preceded to de-solder everything, pull the traces, and start over from the beginning making me take every step methodically and verifying everything himself at each step. I just wanted to ? fre(0) but he was on a mission to teach me how to effectively solder. It probably took the rest of the day.

 

Dad was a cool dude.

 

[manterola]

48 minutes ago, kheller2 said:

Off I go cutting away with a model knife and test away with the meter. Odd thing is one of the two pins I have to isolate from ground still is connected to ground!  How can this be?  So I grind away more.  Still no progress.  I then get out the Dremel and grind that thing deep.  It is still connected to ground even though I've completely gouged a hole between the pin and the original ground trace. And then I wonder...

 

Tha happened to me, exactly. In my case the 600XL motherboard did not included sockets,  I did not want to desolder so I thought cutting traces was a good idea, it was not...

 

Another one, I remember trying to fix a mylar based keyboard. I took all screws out and put them back several times, checked everything with continuity, etc. At the end I realized that the problem was around or close to the keyboard connector, so I cut the final part of the mylar, and it worked!!! All keys perfectly working.

The problem was that when I put everything back to together I realized that the mylar was too short to get to keyboard connector, so I couldn't close the Atari 800XL case.

 

Those two I remember, I am sure I have a lot more.

 

Edited March 2, 2021 by manterola

[mimo]

I put a 6502 chip in a 6520 socket in a xegs after desoldering and replacing just about all the ICs 

It was about 3 years later when raiding the"dead" motherboard for spares that I realised my mistake.

Popped in a 6520 and I had a working xegs. 

Didn't help with my need for spares though 

 

[+Nezgar]

Oh the feeling of hastily putting in an 2732 EPROM backwards in a 1050 and it actually makes a flash in the EPROM's glass window when you turn it on.... Couldn't be more painfully obvious that I toasted the EPROM!

[Frankie]

I had my 800 apart for an Incognito install and I plugged the board in backwards. Turned it on and fried my Incognito.

 

[Awch]

I spent days trying to install a UAV in an 800xl. It's a very simple install but it just wouldn't work. Went over it multiple times confirming my installation with the manual and the long thread here on AA. Moved on to an upgrade of a 400 in frustration. Once that was done I took another crack at the 800xl. I was looking at it through magnifying goggles ('cause I'm old) when I realized that the UAV had Rev C printed on the board and the instructions were all for Rev D. Ugh!

[chevymad]

Decided to replace my vbxe in one machine with a sophia 2. So I pull the vbxe and proceed to insert the sophia into that socket. My original Candle u1mb didn't like that. 

[Mark2008]

Yeah, it's definitely understandable.  

 

I have no Atari stories, since I suck that bad at it, but I did get the idea to repair iPhones back when repairing the original iPhone was still a thing.  Repairing glass and yes - batteries (soldering!).   As soon as people at work heard about that, iPhone requests came flying out of the wood work....I didn't make a single cent on the business and it was such a headache.   

 

The basic idea was if you were super careful you could replace only the glass overlay, not the lcd.

 

Real world, that may or may not work and it wasn't airtight.  Nobody wants an iPhone with a hair that gets behind the glass. So I was working all night to not make someone happy....

 

Battery - you can replace the battery, but you'll just be putting another non-working battery in the machine, since there was no source of quality batteries - the chinese sources available to me, would sell me anything.

 

ok, enough of that story....need to block that back from my memory

 

 

 

 

 

 

 

[flashjazzcat]

A professional recently installed VBXE in the GTIA socket of a 130XE, couldn't understand why it wouldn't work, started a thread about it with photos, and it took two dozen posts before any of us noticed what he'd done. Sadly the thread got deleted, which is rather regrettable since no-one gave him a hard time about it. That was the best one by far of recent times, anyway.

Edited March 3, 2021 by flashjazzcat

[Mclaneinc]

 

My story is from a different angle, I'd inherited a Sony receiver and it was of no use to me so decided to put it on Ebay, its then I noticed the volume was very low so withdrew it and looked for a fix, turns out it was a caps thing so I found and ordered a legendary capacitor pack. That arrived and I broke it all down and carefully changed all the SMD caps. I put it all back together and it didn't work so after several deconstructs and rebuilds I was stumped. I then chanced upon a russian video on YT where the guy was doing the same repair and had videoed it. Turns out in breaking it down the first time I'd removed and lost a tiny connector between 2 boards, unable to fabricate it I then had a heap of parts still sat in a box worth nothing..

 

That's pretty dumb, failed at the first hurdle and yet I'm normally the guy that takes pics of the boards and carefully puts all the parts in little boxes...Just shows how easy it is..

[JayoK]

1. Not an Atari story, but in a former role (20+ years) ago, I received a brand new (expensive) Dec Alpha server for my company and it needed to be commissioned. As typical back then, it came in pieces and it has to be assembled (CPUs, disk controllers, etc, etc) all into the chassis. I had completed assembly and powered on the box. As it started through it's bootstrap, it detected no disks. Hmm.. so I looked into the chassis and saw that the SCSI cable wasn't pushed home fully. Stupid moment commenced:

 

I reached in, and pushed the cable home, meanwhile my "metal" watch touched something, and an small bang with a smell of burning followed. Uh-oh... panic... on/off, check cables, etc. Burning smell continued... turned-off and think.

 

Call to support, reported as DOA and an engineer was dispatched. He arrived took one look at the server and then at me... said... "DOA? Yeah right."  New PSUs were installed and an invoice followed for a 2 x PSUs and Support Engineer... I had to explain.

 

I haven't worn a watch since.

 

Edited March 3, 2021 by JayoK

[TGB1718]

35 minutes ago, JayoK said:

I received a brand new (expensive) Dec Alpha server

Dec Alpha's, built a ton of those in my days  but not done anything like that

[+x=usr(1536)]

Years ago, I was given a 1050 that was said to be fully-working.  My own stupidity caused me to never find out if that was the case or not, however.

 

With it plugged in and resting on a table, I happened to nudge the drive and heard something rattle around inside.  Without thinking, I picked it up at an angle, an action that was shortly followed by loud bangs and the smell of caps releasing their goo.

 

Opening it up revealed two things: a loose case screw and a pair of shorted main caps on the power supply circuit.  The caps were replaced but it never worked properly after that, so it hung around for a while as a source of parts before being passed on to its next owner.

[+DrVenkman]

About a year and a half ago I was swapping my 1200XL for my 1088XEL at my main retro-desk. I unplugged the 1200XL, set it aside, then moved the XEL into the spot. I reached for the XEL’s 5VDC PSU lead and plugged it in. Except it wasn’t the 5VDC PSU lead I grabbed, it was the 1200XL’s 9VAC lead, which has the same size plug. A quick pop and hilarity ensued. I was very, very lucky that the only thing destroyed was one or more logic chips in my XEL-CF2 CF card interface and the Sandisk 8GB card inside. All of the Atari chips - including both POKEYs, thank Jebus - were good. I used the event as the opportunity to replace the CF2 interface with the improved CF3 version and replaced the dead card and all is well. 
 

The XEL power jacks has been changed on the current bill of materials to prevent this but I still haven’t changed the plug and power supply on mine. ? I did put a brightly-colored cable wrap on the XEL lead however, to hopefully catch my eye. 

[Gunstar]

For me, first, the same thing as @DrVenkman, plugging the wrong power cord into a video converter box. I had an RGB to VGA and video to VGA, one is powered by 3.3V the other 5V. I plugged the 5V into the 3.3V device and fried it. $80 down the drain...I ordered another. I still have the original though, and it just needs a new transister so one of these days I'll get around to fixing it...when I need it, instead of buying another.

 

Second, while trouble shooting and swapping IC's on my 1200XL I'd just done a PBI mod too, I swapped in a bad MMU I thought was good (partially working only it turns out) then I found out I had the source connection for PHI2 line for the PBI soldered to a wrong pin, fixed it, but still issues. I went over every PBI line about a dozen times trying to figure out where else I might have gone wrong or a bad solder connect or something. After hours and hours over many weeks, I finally remembered I'd swapped that MMU and inserted the original and everything worked! Lesson learned, if you swap something and it doesn't fix the problem, swap it back before moving on.

 

Those are the only stupid mistakes I can recall atm. But of course I'm sure there were others. The fact is, with all the upgrades and mods I've done over the years, in the end, 99% of the time turned out to be some stupid and simple mistake and otherwise they were installed correctly the first time.

Edited March 3, 2021 by Gunstar

[flashjazzcat]

Oh crap... Herb just reminded me of my magnum opus mad skillz moment. Last April or so, I accidentally got the12V and 5V regulators on my 1088XLD swapped and pumped 12V through every chip on the board when I first turned it on. Fortunately I left the U1MB out for the first power-on test. Michael had labelled the regulators as I recall, so I have no idea what the Hell happened here (although I blame the UK government for terrorizing and pyschologically torturing everyone for the past eleven months).

 

Still haven't been able to figure out what blew (I appear to have a complete set of spare discreet logic ICs), and having rectified the error, the system boots to a PMG stripe down the screen with U1MB installed. Maybe I can diagnose it when I get a scope, which will be shortly.

 

[+orpheuswaking]

When I got back into Atari I ended up throwing out a bunch of carts that I thought didn't work on my 800XL. Nope they just needed a 400/800 or a translator disk. At that time I had no idea. Still kick myself about that. 

[+mytek]

49 minutes ago, flashjazzcat said:

Oh crap... Herb just reminded me of my magnum opus mad skillz moment. Last April or so, I accidentally got the12V and 5V regulators on my 1088XLD swapped and pumped 12V through every chip on the board when I first turned it on. Fortunately I left the U1MB out for the first power-on test. Michael had labelled the regulators as I recall, so I have no idea what the Hell happened here (although I blame the UK government for terrorizing and pyschologically torturing everyone for the past eleven months).

Reminds me of the time I was connecting scope probes to the MPBI port on my XEL with it powered up and ended up shorting out a critical line on the U1MB as a result. End result: dead, very DEAD U1MB ?.

 

Jon I forgot to ask you when you first mentioned this happening - was the MIDI S2 module plugged in at the time?  I'm hoping not.

[JayoK]

2 hours ago, TGB1718 said:

Dec Alpha's, built a ton of those in my days  but not done anything like that

My first real job -- I learned so much that day

[VinsCool]

Specifically Atari, I had done something dumb but I don't think it was that bad.
I had one of my 800xl disassembled and cleaned up thoroughly, and once I was done I was meticulously reassembling it, tightening ever screws together... then put everything back on, and wiped the surface one last time.

Somehow that made the keyboard no longer work properly, with an entire row unresponsive and the A key stuck at power on.
Thankfully I only had to loosen the screws under the keyboard to bring it back to normal, so I assume they really were "loose" on purpose, probably to not make the keyboard contacts too close to the keys.

I also had done several stupid things in the meantime for all sort of electronics, but never really damaged anything beyond repairs, the most recent one has to be shorting my VIC-20 with the video cable I use for my Atari, trying to find which one was actually carrying the video signal...
I certainly found which one was carrying the voltage, oops. Nothing was damaged at least, it simply made my monitor flash for a second, and the VIC-20 to reset.

Another pretty cursed one I did was to use a nail filer and a knife to take off some plastic from a C64 shell, since each halves weren't fitting together, and I had no other parts around, so I made them fit regardless... ?
Sure that was pretty stupid, but in the end it did work and also didn't look too bad from the outside, haha.

[gollumer]

Installed the U1MB upgrade in my 1200XL, and when I fired it up, it wasn't working. I double-checked the instructions, and my work, and couldn't figure out what was wrong. Since I live fairly close, I drove it down to The Brewing Academy. Took about a minute to realize my fading eyesight had caused me to be 1 pin off when re-attaching the keyboard cable. Easiest "fix" ever!

[flashjazzcat]

2 hours ago, mytek said:

Reminds me of the time I was connecting scope probes to the MPBI port on my XEL with it powered up and ended up shorting out a critical line on the U1MB as a result. End result: dead, very DEAD U1MB ?.

Ouch.

2 hours ago, mytek said:

Jon I forgot to ask you when you first mentioned this happening - was the MIDI S2 module plugged in at the time?  I'm hoping not.

It was not, fortunately. I'm now at the stage where I'm not sure if I made some critical, invisible error with a passive component or whether I'm somehow managing to always have at least one destroyed IC installed.

Edited March 3, 2021 by flashjazzcat