Jump to content
Due to a large amount of spamers, accounts will now have to be approved by the Admins so please be patient. ×
IGNORED

Atari Star Wars Cockpit Restoration


Recommended Posts

It's been 8 years since I bought my last project cabinet, and 4 years since I finished restoring it. Over the 4 year gap I hadn't really considered doing any others, so I thought I'd thoroughly scratched the itch. Vindicators was probably a bad introduction to cab restoration but I came out of it wiser to exactly how much work is involved, and how trashed something can be in real life, even if it looks OK in photos.

 

Turn out the itch was still there, it just took something epic to re-awaken it! Something local that looked great in the photos, how bad could it be?

 

So I hired a van and dragged home the undisputed king of early 80s arcade machines, an Atari Star Wars Cockpit.

 

IMG-7772.jpg

 

IMG-7773.jpg

 

IMG-7775.jpg

 

IMG-7776.jpg

 

IMG-7778.jpg

 

The seller described it as working, but it was a long way from being 100%. The vector monitor had only partial vertical deflection, leaving the bottom half of the screen blank...

 

IMG-7785.png

 

... plus a slightly worrying lack of red in the image - hopefully not a bad tube.

 

It also had controller fault, the roll axis worked, but pitch didn't register at all, and there were no signs of any audio, not even a pop on power-up and no amplifier hum. Lastly the game crashed and rebooted a number of times during the time it was running too.

 

The first challenge was how to get a 196KG cabinet out of the van single-handed, at short notice and with a 30 minute deadline to get the van back before early bird pricing turned into a day hire.

 

IMG-7770.jpg

 

After a few failed attempts at ramps I ended up rolling it out onto its nose and driving the van forward before lowering the lighter back end onto its wheels.

 

Time to see what lies beneath!

 

IMG-7798.jpg

 

I had no keys for the locked coin doors, or the back panel, so I unscrewed the lid and got my first look inside in daylight, I'd had a quick look in the sellers garage but the lighting was poor and my phone torch weak.

 

Everything seems to be there...

 

IMG-7799.jpg

 

... a somewhat trashed looking 25" Amplifone monitor, which looks like it has got wet at some stage.

 

IMG-7807.jpg

 

Original deflection board, complete with burnt craters in the PCB from how hot these run.

 

IMG-7801.jpg

 

Original high voltage board, with fugly bodged-in line output transformer and gallons of bathroom sealant to keep the magic sparks in.

 

IMG-7809.jpg

 

Finally, right at the bottom in the dark....

 

IMG-7804.jpg

 

... a fruity mix of mouse and rat muesli.

 

Mouse crap is one thing, but they piss almost constantly, and chew anything they can get their teeth into.

 

Original power brick - rusty!

 

IMG-7805.jpg

 

Wiring harness - chewed up, and filthy.

 

IMG-7806.jpg

 

Audio Regulator II PCB - no damage, just dirty, being vertical probably saved it...

 

IMG-7808.jpg

 

...mounted to a rusted, PCB cage...

 

IMG-7819.jpg

 

IMG-8005.jpg

 

...which thankfully contained a pretty mint PCB.

 

IMG-7811.jpg

 

IMG-7816.jpg

 

An interesting mix of serial numbers in this cab. The two monitor boards and the power brick are all from machine SD02069 , which is probably the serial number of this cabinet, but the main PCB is SD02073, the vector PCB is SD00752 and the audio board is SD02070. Aside from the vector board all of these came from machines within 4 serial numbers of each other, so I can only assume this was owned by a company that bought a fleet of these and ended up cannibalising them for parts around to keep a declining number of them running.

 

Thankfully the PCB cage took all the damage from the squatters, mice may leak constantly but the volume is low, I guess there was never enough to drip through onto the PCB.

 

With the PCB cage out the extent of the mess was clearer.

 

IMG-7820.jpg

 

Power brick out...

 

IMG-7823.jpg

 

... a 20c discount at least, the cab just got cheaper.

 

As I had to roll this out of the van onto it's nose all the crap had rained down to the front of the cabinet.

 

IMG-7824.jpg

 

Apparently these cabinets are a magnet for rats and mice, presumably due to the wiring tunnel that's accessible at the rear wheels ans runs to the front compartment. It is pretty much a perfect artificial burrow for them, they like it, so they move in.

 

IMG-7829.jpg

 

IMG-7830.jpg

 

I tipped it back on its nose and gave it a good thumping until the sound of raining shit stopped.

 

While it was arse-end-up I went looking for other areas the mice had been camping, starting at the rear marquee light box. To drop the glass out I needed to remove 3 bolts, 2 could be persuaded to turn, the third would not budge and I ended up stripping the head.

 

Even dremeling a slot for am more persuasive screwdriver didn't work...

 

IMG-7848.jpg

 

... so I had no option to but to cut it off.

 

IMG-7851.jpg

 

Which got me in here...

 

IMG-7867.jpg

 

...not too bad, more mess but no major damage, all the crap would have rained out earlier, but that's why the bolt wouldn't move, it's in the corner that took the most mouse damage.

 

IMG-7868.jpg

 

With access to the rear wiring I could disconnect the wiring to the fluoro tube and the speaker interconnects which loosened things up enough to get to the actual connectors up inside the wheel well.

 

IMG-7857.jpg

 

Harness pulled out!

 

IMG-7888.jpg

 

Filthy, and chewed to hell in places, the mains wiring seems to have been especially tasty.

 

IMG-7889.jpg

 

Bath time, a soak in hot water + Napisan and a scrub until clean...

 

IMG-7890.jpg

 

...and an afternoon swinging in the breeze.

 

IMG-7941.jpg

 

With it clean the damage didn't look as bad, it's limited to three sections, the mains switch loop...

 

IMG-7939.jpg

 

...speaker connections towards the rear of the cabinet...

 

IMG-7940.jpg

 

...and not sure what this section is yet.

 

IMG-7944.jpg

 

To properly clean out the main compartment I needed to remove the coin box cage which seemed to be the foyer of the mouse motel.

 

IMG-7870.jpg

 

This is hooked over the bar between the two coin doors and clamped to the side of the cabinet by the bolts and catches on the inside. Without keys the simplest option for the coin mech door is to undo the nut that holds the locking arm in place and open the door from the inside.

 

IMG-7872.jpg

 

Hmm, there's something in there!

Edited by Womble
Link to comment
Share on other sites

  • Replies 82
  • Created
  • Last Reply

Top Posters In This Topic

Although disgusting it’s quite satisfying to relieve a machine of the rat and mouse shit

Ahh the sound of rat turds rattling up the shop vac hose :lol

 

Yeh I can't imagine what this would have smelt like when it was left on and all warmed up. :)

 

............

 

There's no access to the coin box or the lock on the lower door from the inside the cabinet, presumably to stop service guys pinching coins. These over/under coin doors have different lock on each door, one gets you access to the coin box, but not the guts, and the other gets you to the guts of the machine but not the coin box. I guess the service techs didn't need to get to their paws on the money and the coin collector wasn't allow anywhere near the electronics.

 

Drilling the lock out would have been an option, but also a bad idea as this sends shards of stinking hot metal down into the coin box area which was full of something. I didn't really want to burn this puppy to the ground on day one.

 

Thankfully tubular locks are very easy to pick...

 

IMG-7948.jpg

 

...you just need to put some rotational tension on the inner core and gently push each of the locking pins to find which one is binding, once you go round a few times it will unlock.

 

Using a tensioner made from a section of wire coat hanger with the end squeezed square in a vice...

 

IMG-7949.jpg

 

...it unlocked. You have to pick these locks twice to get the door open as at every 1/8 of a turn the pins all drop in and it locks again, and you need to turn it 90 degrees to unlock the door.

 

Success, WTF?

 

IMG-7951.jpg

 

One mouse nest, made of tights, plastic bags, tarp, and what was once the cabinet wiring diagram on the inside door.

 

IMG-7956.jpg

 

IMG_7957.jpg

 

IMG_7958.jpg

 

At the bottom was the remains of a drilled lock, the cardboard fuse block cover and a receipt dating from 2010, which I guess when the mouse house was last occupied.

 

IMG_7960.jpg

 

More discount, all up this cabinet was $3.80 cheaper than I thought it was! The coin box itself wasn't chewed up at all and the mess inside it was mostly superficial.

 

IMG_7961.jpg

 

Nothing a damn good wash in hot soapy water won't fix.

 

IMG_7964.jpg

 

With that out I could move on to removing the coin door and the cage.

 

IMG_7976.jpg

 

IMG_7978.jpg

 

Disgusting!

 

IMG_7981.jpg

 

IMG-7872.jpg

 

But 82,098 coins aint half bad.

 

IMG_7979.jpg

 

A good vacuum and blast with the sander sorted out the insides rather well, I guess there was never enough wet to really soak in.

 

IMG_7984.jpg

 

IMG_7986.jpg

 

The only other enclosed space to investigate was under the seat.

 

IMG_7903.jpg

 

IMG_7904.jpg

 

Thankfully not much had gone on in here, it got cleaned out and sanded.

 

IMG_7909.jpg

 

The marquee area got a scrub down too.

 

IMG_7910.jpg

 

All in all the mouse damage could have been a LOT worse, and as the cat had lost interest in the cabinet I reckoned it was clean enough to be let in the house.

 

At this point the tear down was complete, basically everything except the controller. Time to focus on the heap of filthy parts in the corner of the garage, before they get too comfy there, or I wake up and find 4 years have gone by like last time.

Edited by Womble
Link to comment
Share on other sites

So the power brick...

 

IMG_7998.jpg

 

....the coin door...

 

IMG_8073.jpg

 

...coin box and service panel...

 

IMG_7981.jpg

 

...were the main lumps, along with the PCB cage and the fan assembly.

 

As the coin box was mouse central it took the most of the damage. Apparently mice don't piss and crap in their nest, they wait until they are just outside to turn their taps on so this area took the brunt of it, as their only way in and out was via the coin chutes.

 

IMG_8009.jpg

 

This was beyond a simple wash as the paint was peeling and it was rusty in all the corners. Service panel off!

 

IMG_8013.jpg

 

IMG_8193.jpg

 

IMG_8196.jpg

 

IMG_8197.jpg

 

IMG_8199.jpg

 

Only option was a good sanding outside...

 

IMG_8201.jpg

 

...and in.

 

IMG_8202.jpg

 

The service panel and the fan bracket had the same paint job originally, so they were disassembled, ready for the same treatment.

 

IMG_8207.jpg

 

IMG_8208.jpg

 

I also did the coin reject flaps at the same time as their paintwork was shot.

 

IMG_8213.jpg

 

All back to bare metal, cleaned with acetone and left to dry in the sun...

 

IMG_8219.jpg

 

... before being propped up on some off-cuts from the Vindicator restoration years ago (the triangle pieces from the track edges make awesome paint props) and hit with metal primer.

 

IMG_8223.jpg

 

The worst sections on the coin box got an anti rust treatment, this stops the rust from spreading and converts it to a paint-able surface.

 

IMG_8229.jpg

 

IMG-8232.jpg

 

The coin doors themselves were in relatively good condition, but the paintwork was tired, scuffed, and badly chipped in places. I'm always in two minds with Atari coin doors as the original paint job is hard to replicate, originally the doors where painted matte black with a splatter coat of gloss paint over the top, giving tiny spots of glossiness. This not only gives it a strange "not matte, not gloss" look but also provides the texture. The texture on the coin gates is a wrinkle coat in satin black. The result is a major job if you want to strip back to bare metal as you lose the texture, and a very risky job that is likely to get ruined at the splatter coat stage. Some people have managed it by holding a spray can upside down and pressing the nozzle very lightly. I've tried this on cardboard and it really didn't work. So to keep the texture, and my sanity, I've previously just paint over the top in a satin black, but on these doors there were areas of damage to the original paint that needed treating first.

 

Both doors had break-in damage on them which was down to bare metal.

 

IMG_8180.jpg

 

I suspect someone had leaned on the lower door while it was open which had bent the hinge at the top...

 

IMG_8155.jpg

 

...so the door was lightly angled and rubbed on the frame. It wouldn't close properly without force which was rubbing into the paint in that corner

 

This turned into a right pain in the arse, as by the time I got the hinge aligned as best I could, 3 of the 5 tiny welds had failed and the hinge was now flapping in the breeze, making it even worse

 

IMG_8164.jpg

 

A few small rivets almost fixed the issue...

 

IMG_8172.jpg

 

... but it was still binding in that corner, despite everything looking like it was perfectly aligned. The original paint was pitted and there was some rust which may have been part of the issue. A session with the grinder cured the remaining issue.

 

IMG_8173.jpg

 

With the door finally able to clang shut without any assistance I was left with a range of chips and dings to prime, along with the ground-out corner.

 

IMG_8177.jpg

 

Most of the paint on the lower area had been eaten into by the mouse damage and flaked off, but a final clean, and a removal of the silver texter with iso alcohol...

 

IMG_8176.jpg

 

... resulted in a coin door ready for a bath

 

IMG_8184.jpg

 

IMG_8185.jpg

 

After they spent some time sunbathing, a session with masking tape and primer followed to sort out the damaged areas...

 

IMG_8224.jpg

 

IMG_8234.jpg

 

With the primer drying I turned back to the coin box, the anti-rust stuff had done it's job but left a lot of white powdery gunk that had to be cleaned off.

 

IMG_8236.jpg

 

With that scraped, soaked and sanded back it was looking pretty good, just needing some filler on the really pitted areas around the spot welds.

 

IMG_8241.jpg

 

Once hardened and sanded, I cracked out the Pringles and got to work.

 

IMG_8243.jpg

 

IMG_8245.jpg

 

Moving on to the coin door parts, these were washed, dried, and any loose paint picked off, before the first coat of satin black...

 

IMG_8249.jpg

 

IMG_8250.jpg

 

...and the whole left to dry for an hour or so.

 

Moving on to the front of the doors...

 

IMG_8258.jpg

 

...and the back of the frame.

 

IMG_8260.jpg

 

But by the time I got within an inch of finishing the job, with a final coat on the front of the frame...

 

IMG_8262.jpg

 

...the paint ran out, 30 more seconds and this would have all been finished.

Edited by Womble
Link to comment
Share on other sites

Emergency trip to Bunnings and another top coat required on everything.

 

IMG_8263.jpg

 

No big deal as I had forgotten the coin flaps anyway.

 

IMG_8265.jpg

 

With that lot drying...

 

IMG_8266.jpg

 

...the service panel and fan bracket got the same treament to match the original paint job.

 

IMG_8327.jpg

 

IMG_8330.jpg

 

The coinbox got the same treatment...

 

IMG_8546.jpg

 

IMG_8552.jpg

 

...multiple light coats, about 5 in all, and lots of drying time. With the cold weather on and off, plus work, this took weeks.

 

I'd also been cracking on with the power brick over the same period. I've met a few of these over the years and they usually look far worse than they are, in other cases the dirt is just 30 years of dust from the cabinet and simply brushes off. In this case it was as bad as it looked.

 

IMG-7822.jpg

 

There's no way this would have escaped the mice being right at the bottom of the cabinet and covered in tasty wiring, and the whole unit was filthy and badly corroded.

 

IMG_8001.jpg

 

IMG_8021.jpg

 

IMG_8025.jpg

 

Despite the carnage up above, the underside of these bricks is always pretty mint.

 

IMG_7999.jpg

 

I've yet to meet one of these with a non-damaged fuse block, or with the correct fuses installed. Over the years dirt and corrosion makes the fuse contact fail, until you get arcing and burning, which blows the fuses and the operators usually ram in whatever fuse they could find to fit.

 

Time to strip it down, damaged fuse block disconnected...

 

IMG_8028.jpg

 

the replacement fuse holders that were fitted to bypass the burning come off. This was actually a very neat repair by someone, the best job I've seen in a cabinet, but as I'm fitting a new fuse block they can go into my parts stock.

 

IMG_8032.jpg

 

Big blue removed...

 

IMG_8033.jpg

 

IMG_8034.jpg

 

...transformer comes off...

 

IMG_8038.jpg

 

Everything was going well until I hit the epicenter of rust.

 

IMG_8043.jpg

 

This is the bolt that holds the line filter to the underside of the plate, and is also the central earth point for the cabinet wiring. This was rusted so solidly the only option was to cut it off.

 

IMG_8046.jpg

 

I had hoped that getting the badly deformed upper part of the bolt off would allow what was left inside to turn. Nope - all I achieved was rounding off the nut.

 

IMG_8049.jpg

 

More cutting...

 

IMG_8052.jpg

 

...and drilling...

 

IMG_8053.jpg

 

...finally got it to give up.

 

IMG_8054.jpg

 

There's one molex pin to remove from the main connector before you can release the transformer from the base plate. Getting this out was a friken nightmare without the right tool.

 

IMG-8035.jpg

 

After that the crusty mains input connector unclips...

 

IMG_8055.jpg

 

... to release the transformer wiring.

 

IMG_8056.jpg

 

That just left the bridge rectifier and the terminal block to unbolt.

 

IMG_8057.jpg

 

Some people have fixed up badly corroded plates by sanding and painting the metallic gold effect back on, but this was so far gone I reckoned it was toast.

 

The issue was the much same problem with the coin mech parts, any horizontal area that the mice been on was corroded.

 

IMG-8070.jpg

 

There is a guy in the US who has remade new base plates for the power bricks, and you can get these coin mech parts from a few places, but the cost to ship to Aus is always the killer. Plus any salvaged coin mech parts are unlikely to be minty fresh looking..

 

Luckily I know a guy who got into sandblasting and plating for his own restoration projects, he now has a small sandblasting booth and electroplating tanks set up in his garage. He was also up for a challenge, so he got a box of dirty metal pieces to see what he could do.

 

IMG-8106.jpg

 

The stickers were trashed so I peel them off, somewhat reinforcing the challenge

 

IMG_8131.jpg

 

IMG_8134.jpg

 

IMG_8136.jpg

 

It's amazing how the stickers protected the areas they covered.

 

IMG_8137.jpg

 

The electroplating process dislikes cold weather as much as spray paint, and being August in Melbourne this was going to take a while.

 

 

----To Be Continued ----

Link to comment
Share on other sites

Great story and step by step breakdown. Looking forward to the next chapter. Are you logging man-hours? Interesting to here just how much time is involved.

 

 

Sent from my iPad using Aussie Arcade

Link to comment
Share on other sites

Great story and step by step breakdown. Looking forward to the next chapter. Are you logging man-hours? Interesting to here just how much time is involved.

 

Thanks!

 

The next chapter should be up later this week, that'll close out this phase, then its on to the electronics, I'm leaving the cabinet work to last for when the weather is a bit more predictable.

 

Haven't been logging time, not sure I want to know, but it will end up being a lot.

Edited by Womble
Link to comment
Share on other sites

  • 2 weeks later...

A few weeks later the plating was back, and the results were stunning.

 

IMG_8875.jpg

 

IMG_8879.jpg

 

Looks glorious - time to rebuild me some coin mechs! But first I needed to sort the reject button plastics.

 

IMG_8286.jpg

 

IMG_8289.jpg

 

A long soak in Simple Green got rid of the caked on grease. They ain't perfect, and I may replace them with NOS ones if I can find them, depends how they look when installed and backlit again.

 

IMG_8314.jpg

 

Re-assembly!

 

IMG_8884.jpg

 

Slight problem was that the bent reject frame was so bent that the coin flap hinges wouldn't fit properly into the recess so that flap was stiff and at an odd angle.

 

IMG_8883.jpg

 

Some action with a hammer, flattened it out but trashed the paint job. A few re-coats later and it was back in the game.

 

Next issue was the bottom half of the coin mech is held together by the bolts that hold the coin switches in place, and mine didn't work very well. Readings of 250-400 ohms when they were closed were never going to work well.

 

IMG_8889.jpg

 

Most switches I've met in cabinets are fairly easy to pop apart and clean, but even polishing and deoxit'ing the contacts doesn't always work. Replacing the switch is an option, if you can find them, but an easier option exists.

 

Just swap the never-used NC (normally closed) contact with the NO (normally open) contact that has the 30 years of wear on it.

 

IMG_8890.jpg

 

Giving everything a clean and a dexoit resulted in two minty looking switches that now gave a zero-ohm reading when activated.

 

IMG_8893.jpg

 

IMG_8895.jpg

 

The chunky bolts that hold the actual coin mech in place got a good polish as they were letting the side down...

 

IMG_8902.jpg

 

...but not badly as the wiring harness.

 

IMG_8906.jpg

 

The mice had been all over these and the tubing over the wires was covered in a thick greasy wax.

 

Baby wipes shifted that eventually and soapy water sorted the rest, which left the two rusty mounting screws as the final issue.

 

IMG-8922.jpg

 

A spin in the drill against some metal polishing lint later...

 

IMG_8929.jpg

 

... fit to be put back in, along with the service key hook. This is where the back door key lives, so the tech who has access to the service panel, can get in the back. It was completely rusted before but after a sandblast and plating step it completed the job!

 

IMG_9135.jpg

 

IMG_9139.jpg

 

IMG_8911.jpg

 

Coin door complete!

 

Power brick up next, shame about the transformer...

 

IMG_8001.jpg

 

... filthy, rusty and probably strongly mouse scented when warmed up.

 

A session with a wire brush and some polish got me some way...

 

IMG_8872.jpg

 

...but as my patience faded I resorted to sand paper.

 

IMG_8873.jpg

 

The copper strap looks like it is there to hold the transformer together, but it's purpose is actually to cut down the stray magnetic field leakage that would otherwise annoy the monitor. It was originally lacquered but that had let go in places so I stripped it all back on the upper faces and polished the copper back to a shine.

 

With the wires cleaned and everything rubbed down it got masked up...

 

IMG_8916.jpg

 

... and hit with a few coats of spray paint for a massive improvement...

 

IMG_8955.jpg

 

...then the masking was reversed and the copper strapping got a final polish and three coats of lacquer to keep it from re-tarnishing.

 

IMG_8968.jpg

 

Winning!

 

One big problem with working on cabs from the US is that all parts are imperial rather than the easy-to-source-locally metric options. A problem when you have a bag of rusty old parts that you would usually just replace. Replacing them with metric inst really an option as the old threads had cut into the base plate, and ramming new threads in would cut again and there isn't that much metal thickness to play with.

 

IMG_8935.jpg

 

The internet seems to be full of crazy sounding ways to treat rust, lemon juice, coca-cola, and apple cider vinegar. The vinegar one is pretty evenly split between people who say it worked like magic, and those who say it did nothing, but there was some in the kitchen. Worth a shot!

 

IMG_8938.jpg

 

After ten hours or so the rust was totally gone except on the one really badly rust k-nut, so I chucked in a few more parts and left it overnight, and everything turned black.

 

IMG_8939.jpg

 

Which thankfully just polished off.

 

IMG_8940.jpg

 

The result being an almost complete set of clean shiny parts to dry in the sun.

 

IMG-8950.jpg

 

The long transformer bolts didn't need anything as violent to get them presentable again, just a spin in the drill got them shining again.

 

IMG_8974.jpg

 

Reassembly time, with the assistance of fish...

 

IMG_8976.jpg

 

...so I could drop the wires through the plate while fitting the bolts.

 

IMG_8980.jpg

 

I'd had to cut through one nut and bolt I was one 8-32 3/4 inch bolt and the accompanying k-nut short. So the one originally mounting the bridge rectifier got redeployed, and a metric bolt from Bunnings now holds the rectifier in place as that doesn't screw into the plate, just through a hole in it.

 

Looking pretty sharp!

 

IMG_8981.jpg

 

I'd figured the capacitor clamp was good enough as it was the best bit on there originally, but against all the newly plated parts it looked rough as guts.

 

IMG_8972.jpg

 

So I got my mate to sandblast and plate it too!

 

IMG_9014.jpg

 

Also needed to clean up this puppy, the voltage selector plug. The transformer on this brick Atari's international one, that supports 100v, 110v, 220v and 240v mains inputs, with this little plug-in adapter selecting the correct taps for the region.

 

IMG_8977.jpg

 

The schematic pack lists the options for various plug options, and this one is the 220v version. Each region had different coloured wiring, and 220v wiring should be blue. Either through age, or by the powers of mouse, the wires were now bleached to a very pale pink.

 

New set of spade connectors go in to replace the crusty old ones and the missing ones, plus a new set of fuses.

 

IMG_9264.jpg

 

And while I'm gilding the lily on this I tracked a set of replacement decals, and the correct vulcanised fibre board to make a replacement fuse block cover for the chewed mouse-soaked original found in the coin box.

 

IMG_9236.jpg

 

Decals on..

 

IMG_9265.jpg

 

... fuse blast shield installed...

 

IMG_9276.jpg

 

... and that's the power brick done!

Edited by Womble
Link to comment
Share on other sites

The coin box and service panel paint jobs were done, leaving only the wiring harness to sort out.

 

IMG-8957.jpg

 

Before bath-time the switch, volume pots and the coin counter had to come off for a manual wash, as they wouldn't appreciate water.

 

IMG-8964.jpg

 

The harness went in the sink for a scrub and the pot got a clean by hand before being soldered back on.

 

IMG_9082.jpg

 

Was feeling protective of the paint job so a couple of fibre washers ended up preventing the nuts from chewing it up.

 

IMG_9084.jpg

 

The mounting plate on the coin counter was rusted so that went through the plating process before being refitted.

 

IMG_9128.jpg

 

IMG_9130.jpg

 

Service panel and coin box reunited and complete!

 

IMG_9132.jpg

 

IMG_9133.jpg

 

The mains-powered fan below the AR and main PCB was beyond saving.

 

IMG_8059.jpg

 

The dirt and corrosion may have been fixable but the badly chewed fan blades weren't. Even if I got it spinning again it would be out of balance, and I dread to think how many miles this has on the clock already.

 

IMG_8060.jpg

 

As this connects downstream of the power brick it needs to be a 110V fan, which means an import job. I could have modified the wiring to tap off 240V but as I'm putting this back to original condition I picked one up on US ebay that had cheap shipping, all up it was only about $15 more than a 240V one from Jaycar.

 

The rusty bolts all got a spin in the drill against polishing lint...

 

IMG-9073.jpg

 

... and the fan module assembled.

 

IMG_9275.jpg

 

The final part of this phase was the PCB cage. The upper face was badly corroded and even the best bits were looking very tired.

 

IMG-7819.jpg

 

IMG-8005.jpg

 

I flipped back and forth between three options, leave it as is, go the painting path, or get it stripped and re-plated. Leaving it as it is seemed a shame considering the rest of it was getting pimped. Painting was going to be a tonne of work and getting the right effect isn't easy or cheap. Or I could just bite the bullet and get the whole thing plated again. My mate's set up couldn't deal with anything this big so my only option was to go pro. Painting it would have cost me at least $50 in sanding pads, primer, paint as well as a lost weekend. Plating it was $90 and I could just drop it off on my way into work.

 

The plastic guide rails for the ARII and game PCBs came off for a wash...

 

IMG_8778.jpg

 

...and the cage was dropped off at Tudor Plating in Brunswick.

 

A week later I collected this!

 

IMG_9109.jpg

 

Mouse rust gone!!

 

IMG_9111.jpg

 

IMG_9115.jpg

 

Rails back in...

 

IMG_9118.jpg

 

... then out again as fitting them to the obvious holes means the 3 stack board won't fit.

 

IMG_9120.jpg

 

Four bolts back in to mount the ARII bracket...

 

IMG_9134.jpg

 

...and job done.

 

Stage 1 complete, garage corner cleared.

 

IMG_9267.jpg

 

Next up is the electronics, all the good stuff I had forced myself to keep off until I had got some progress made with the rest, that's all a few weeks away, in stage 2.

Edited by Womble
Link to comment
Share on other sites

  • 3 weeks later...

The wiring harness seemed like a decent place to start phase two. It had already had a good clean and the damage didn't look too bad.

 

IMG-9238.jpg

 

The mice had chewed into some wires and snipped out whole section of others, leaving only the bits they couldn't pull free from the cable ties.

 

IMG-9239.jpg

 

So it was just a case of spending the time slicing and dicing, and cutting out the damage and splicing in new sections.

 

IMG-9241.jpg

 

A great source for the right gauge wire, and a decent colour match for some sections was the old power lead itself.

 

IMG-9237.jpg

 

This section is the audio connection from the audio amplifier on the ARII PCB to the speakers, that runs in the wiring tunnel under the floor. The damage there was likely the only reason there was no audio from the cabinet, disconnected speakers tend to stay quiet. The bundle also carries the AC power feed to the rear marquee fluro tube.

 

IMG-9242.jpg

 

The middle section of the damage was the worst, needing lots of splices to put it all back together again...

 

IMG-9243.jpg

 

... including repairs to sections with gnawed insulation.

 

IMG-9244.jpg

 

The final section was more of a tangle as the original wiring had been modified to tap into the upstream 240V for a replacement fluro tube module behind the marquee.

 

IMG-9245.jpg

 

The two three-pinned connectors should plug onto the 110V connectors on the power brick J4A and J4B. One is for the marquee light and the other is for the fan. The marquee plug had the live and neutral wires cut out and spliced ...

 

IMG-9246.jpg

 

...to the mains switch harness to get 240V for a replacement fluro tube behind the marquee. The plug with the remaining earth wire was still connected to the brick on arrival, with the wires clearly exposed. Not a great idea when there is 110VAC on them.

 

IMG-7821.jpg

 

Originally this cab would have had a 110V fluro lamp, but when that died there wasn't any way to get a replacement locally here. As the transformer converts the 240V into 110V, the only option at the time was to tap into the mains directly, upstream of the transformer, to power a locally-sourced replacement lamp with the mains voltage it was expecting. The simplest way to do this was to cut into the mains switch loop which should be a separate harness.

 

IMG-9249.jpg

 

The mains power comes into the power brick direct from the mains cable, goes through the line filter and is appears on two pins of 4 at J2 connector. The switch harness inserts the switch between the two output pins of J2 before returning it to the other pair, which goes off to the transformer. For some reason it appears to have been made of particularly delicious wire.

 

IMG-9277.jpg

 

Half of it is totally ruined, to the extent I'm surprised this wasn't starting fires or blowing fuses as the strands of both the live and neutral are all torn out and tangled up. Luckily, even though half was trashed it is actually twice as long as it needs to be for this cabinet, I guess it is a standard part and designed for other cabinets where the power brick is placed further away from the main switch.

 

It was neatly done but the wires were very loose as the teeth inside the joiners had cut most of the strands in the wire

 

IMG-9257.jpg

 

2019 provides more options than the mid 1980s, 110V ballasts are easy to find online and cheap as chips from AliExpress, but the issue is more than the voltage, it's the mains frequency too. Anything plugged into an Aussie wall socket will get a 50Hz mains supply, and no combination of transformers will change it. A 60Hz ballast will probably work, but it will draw more current and will get hotter than it was designed to. The simplest option is to go for a modern "ballast bypass" LED tube, but find one that can take 100-230v input direct. These contain switch-mode PSU and those can be designed to take a wide range of input voltages. My biggest problem there is finding one that states 110V-240V on the packaging, anything with Australian specific packaging is unlikely to say anything other than the local voltage, even if the generic unit inside doesn't care.

 

With half a plan, I fixed the gouges...

 

IMG-9278.jpg

 

...thoroughly heat- shrinked it, cut the bad half of the harness off and re-terminated the connections to the switch with fresh spade connectors, and reunited it with the cleaned-up mains switch.

 

IMG-9279.jpg

 

Lastly it needed a new mains cable, which was just a case of moving the old connector on to a chunky new cable.

 

IMG-9508.jpg

 

One whole heap of done!

 

IMG-9280.jpg

 

Onto the PCBs!!

 

The ARII is the board that takes the stepped down AC voltages from the power brick, rectifies them to the range of DC voltages the game board needs. The right hand section contains the audio power amplifiers as the required + and - power rails are local.

 

IMG-9191.jpg

 

Great condition, just dirty.

 

IMG-9192.jpg

 

To the sink!

 

IMG-9194.jpg

 

To the sunshine!

 

IMG-9195.jpg

 

The advice from the US gurus is that if an ARII works there is very little that needs to be done to them, other than check the solder joints and to replace the 2N3055 power transistor on the heatsink. It's this transistor that does the heavy lifting on the board and gets very hot, by design based on the size of the heatsink. If they fail they can short out and dump high voltages into the game board, so with a 35 year old one on board, with thousands of hours one the clock, it is worth just swapping it out.

 

The 26th week 1983 vintage 2n3055 came off and a named-brand one from RS went in, born 39th week 2018. You can get these from all sorts of sources, for a huge range of prices, from 50c to nearly $10. They all claim to be the same thing but this part is very commonly faked so it pays to go to a trusted source and pay full price. This one was from RS and was 20x the price off the fleabay suspiciously cheap ones, considering on what this is going to be connected to it really isn't worth saving a few bucks and risking it.

 

IMG-9432.jpg

 

IMG-9434.jpg

 

The only suspect solder joint was at J4, the socket had been soldered in slightly raised off the board, so inserting a plug flexes the solder joint. Also there wasn't much solder so you could see the whole joint bending when you moved the connector.

 

IMG-9433.jpg

 

It came off, and was re-soldered back on flush to the board.

 

The expectation may be that this needs a cap-kit, but capacitors on the ARIIs are almost never bad, they were massively over specced and a high quality brand. They also had a easy life, were a long way from anything hot, and were not subject to the brutal treatment capacitors get in modern switch-mode power supplies. So after running the ESR tester over them and getting the all clear I left them be. I also passed on the sense mod, which overrides the auto-voltage system which can run away and damage boards if the harness contacts get really badly tarnished. Won't be a problem on this cabinet for a decade or two, and I'm planning on fitting the new over-voltage protection kit that's soon to be available for these.

 

On to the game PCBs, again these were working so only needed minor work.

 

IMG-9315.jpg

 

The stack breaks down to the CPU board, the vector board, the audio board, plus the stack interconnect board.

 

IMG-9327.jpg

 

The only battle scar on this board was the incorrect voltage regulator fitted over the scorch mark of it's predecessor on the vector PCB. These provide the reference voltages in the vector unit so have an easy life.

 

IMG-9337.jpg

 

It should be a 7815 but a 7812 with a pull up zener diode had been used to raise the ground reference, so it needed to be insulated from the pcb grounding layer below, hence the paper stickers.

 

This cleaned up surprisingly well (with the capacitor moved out the way for access).

 

IMG-9339.jpg

 

The replacement doesn't need to be isolated from the grounding plane, so as the plastic retaining lug was scorched and brittle it got a proper bolt fitting.

 

IMG-9341.jpg

 

The common failure point on boards of this vintage is the sockets, this era used single wipe sockets with only one spring contact in the socket pushing onto the inner face of the chip leg. The problem here is that the metal gets brittle over time and loses its springiness, reducing the contact pressure. The chip legs also get splayed out over time reducing the contact pressure even more, and layers of oxide build up on both faces and adds resistance. On more modern sockets you can scrape through the oxide by removing and re-seating the chip, and they have two spring contacts, one on each side of the leg. On these old brittle ones removing and re-seating often fractures the contact leaving things worse than before. The result is hundreds of intermittent contact points, and a range of resistances on random pins, which leads to an unstable board. This cabinet crashed a few times during the demo in the garage, which could be down to all manner of things discovered so far but sockets are high on that list.

 

Trouble is replacing them all is a tedious job, all up there are 906 pins to desolder and re-solder, without causing any heat damage to the fragile tracks and pads in process. It takes hours but I've more time tracking down the individual socket issues on other boards, than it takes to replace the lot and cure the issue entirely. It's an investment in sanity!

 

IMG-9426.jpg

 

IMG-9425.jpg

 

IMG-9428.jpg

 

I also wonder how many POKEY chips die from being in old sockets with poor contact on the power or ground pins. Its possible they end up drawing too much current internally, or getting power parasitically through control pins as a result. As these are now approaching US$40 each I'd prefer to keep them happy with a 20c socket each.

 

IMG-9462.jpg

 

The only original chip not going back on is the Analog Vector Generator (AVG) IC on the Vector board. These are dropping like flies to the extent it is rare to find a board with a fully working original, and they can damage the monitor if they don't fail cleanly. This one looked like it was mostly working but it was badly corroded and a liability. Thankfully modern replacements are now available, so one went in.

 

IMG-9466.jpg

 

IMG-9467.jpg

 

The board interconnect PCB is another weakness in the design when the boards are this old. It's just three 88 pin edge connectors connected in column that connects the three boards together. If these contacts are dirty or oxidised then problems and stability issues can creep in.

 

IMG-9456.jpg

 

Originally these had a foam pad stuck to the rear to prevent the pins shorting out on the metal PCB cage if the board-set was pushed too far in. The foam has always crumbled to dust on these by now, and the blue pad option is a good solution, but they used two layers making the sandwich far a bit thick, leaving the board sticking out of the cage too far.

 

IMG-9457.jpg

 

To inspect the state of the board it all had to come off, along with the remains of the glue from the pad.

 

IMG-9459.jpg

 

After a clean with glue remover the board turned out to be in perfect condition, not a single dodgy joint to be seen.

 

IMG-9461.jpg

 

So one of the blue pads was refitted and all the edge connectors got a treatment with DeOxit Red and Gold before being reassembled as complete a stack.

 

IMG-9511.jpg

 

With all that lot out of the way I was left with some stuff that actually needed fixing - the boards from the vector monitor, a strange unnatural beast, and something I had never met before.

Edited by Womble
Link to comment
Share on other sites

Great work so far! What technique do you use to remove the old sockets without damaging the traces and pads? What type of desoldering unit do you have?

 

It's a Xytronics DIA60 desolder station, basically a hollow tipped desoldering iron attached to a vacuum pump with a glass collector in the handle.

 

It's a generic model, it was rebadged and sold under many names, mines actually branded as a Dick Smiths T2270. I bought it in their fire sale back in 2007 when they were selling off all their components and useful stuff. I'm sure there are much better options around these days, if this ever dies I'll look to get something a bit less basic.

Edited by Womble
Link to comment
Share on other sites

It's a Xytronics DIA60 desolder station, basically a hollow tipped desoldering iron attached to a vacuum pump with a glass collector in the handle.

 

It's a generic model, it was rebadged and sold under many names, mines actually branded as a Dick Smiths T2270. I bought it in their fire sale back in 2007 when they were selling off all their components and useful stuff. I'm sure there are much better options around these days, if this ever dies I'll look to get something a bit less basic.

 

Thanks for sharing this information and your restoration.

Link to comment
Share on other sites

On to something a bit unusual, for me anyway, a vector monitor!

 

These were a bit of an evolutionary dead-end in arcade history, they were a simpler (kinda) way to get large fast moving images without needing a lot of processor power or expensive amounts of RAM. The main CPU and board logic is really only dealing with game geometry, which is converted into straight line information (vectors) and is fed via a series of op-amps to the CRT's deflection coils to swing the beam around drawing a rapid stream of individual straight lines on the screen to build up the image, up/down/left/right/diagonals etc.

 

This is in contrast to almost all other games where an image is built up in memory on the board and delivered to the monitor which draws it on the screen in a series of scan-lines from top to bottom, before returning to the top and starting again for the next frame. The result is an image style that is very unique and not something you can really emulate. By the mid 1980s with faster CPUs and falling RAM prices the industry went back to using more common raster monitors, possibly because vector monitors had gained a certain reputation for being unreliable and in some cases a fire hazard.

 

Star Wars machines were born with either a 19" Amplifone monitor in the upright, or a 25" for the cockpit version. Earlier Atari vector games had used the WG6100 vector monitor, but Wells Gardner couldn't keep up with the demand, leaving Atari with warehouses full of backlogged Tempest machines waiting for a monitor before they could be finished off and shipped. Atari's initial plan was to contract a second supplier, and Amplifone were the front runner, but in the end Atari just bought them outright, and cranked up production of their own design, hence the Atari logo on the Amplifone boards.

 

This in-house design is regarded as the better option for Star Wars, it was higher resolution than the WG6100 and supported a larger screen for the sit-down versions. Plus it apparently handles some of the effects better, such as the Death Star explosion which puts a lot of stress on the monitor.

 

Unfortunately the Achilles heal in the Atari design was the bright red line-output transformer (LOPTs), aka the flyback in the US. These were made in-house by Amplifone but failed in their thousands early on, while the machines were still young and should have been raking in the coins. Legend has it this is due to an incident at the manufacturing plant where a newly-in-charge Atari Exec refused to put out a cigarette when visiting from Atari soon after the acquisition. This set off the sprinkler system and gave the transformer wiring stock a good soaking. Not ideal as the LOPTs needed to be manufactured in a low humidity environment with no wiring or potting flaws. It's a good story but seems to have surfaced a few years after the event, and people who worked at Atari didn't hear it at the time. The contemporary story within Atari was that humidity during manufacture was the root cause, but it was due to them starting manufacturing during hurricane season, without controls in place to keep humidity low in the production area.

 

Whatever the cause it seems that the result was a potting compound that wasn't up to the task and it broke down under high voltages allowing shorts to appear at random points, in some cases blowing fuses, and in others actually setting fire to the cabinet.

 

Atari commissioned an emergency batch of LOPTs from another manufacturer but this took so long that many machines had their Amplifone hardware junked and a WG6100 monitor slotted. They were plug and play compatible and the industry couldn't wait for the official fix. By this time the Wells Gardner's production line had caught up and it was functioning Amplifone monitors that were now the rarer species.

 

Because the new stock of black LOPTs was too late for most Operators, much of the stock went unused. Up until the early 90s they were easy to find and cheap, but towards 2000 the stocks were dwindling and prices were going through the roof. So Amplifone monitors slipped into the uneconomical to repair territory, and the WG6100 was again regarded as the best monitor have, seeing as it was actually possible to repair one.

 

Today it has flipped once again, as replacement LOPTs are available again and prices have dropped from $200 back to $25 or so, and the Amplifone is regarded as the better monitor among collectors as it has a higher resolution and supports the bigger screen in the cockpit cabinet. Finally after 40 years the flaws in both monitor types have been well defined and the reliability of both can be hugely improved with some simple upgrades and modifications.

 

In my case this cabinet seems to have the original monitor boards and tube it was born with, as the serial numbers on the two Amplifone boards match the power brick in the bottom of the cabinet, but the original red LOPT is long gone. Both of my boards, deflection and the HV board, show signs of an eventful life with burn marks from previous faults and a tonne of hacks and bodges.

 

IMG-7919.jpg

 

IMG-7920.jpg

 

IMG-7926.jpg

 

The horrible mess on the HV board appears to be a common LAI hack to fit a LOPT from a WG6100 to the Amplifone, it was fairly major surgery and is going to be a pain to reverse.

 

IMG-7927.jpg

 

If anyone knows who B.K. in Cowra were, I'd be interested to know. I'm guessing these boards were sent off for repair at some stage and this may have been the owner/operator at that time.

 

The tube itself was filthy after 40 years of service and storage, mice peeing all over it hadn't helped either.

 

IMG-9285.jpg

 

The last time it was powered on back in the sellers garage there were two video faults, the lower half of the screen wasn't being drawn and that the image had missing or weak red. While it would have been good to find those issues specifically there is so much that needs to be done on these boards to address the design issues and bodges that it seemed far simpler to strip it back, clean it up, apply the upgrades - and then fix any issues that remain.

 

The community doco on these monitors is fantastic, especially @dezbaz 's site which is an absolute goldmine of information. Armed with the details of what needs to be done to service these, and what modification/upgrades I figured I'd give it a go and pray that the tube is actually still good. Thats one part that I really can't find a replacement for if that turns out to be hosed.

 

Considering the state of the HV board I took the easier of the two and started on the deflection board.

 

First to come off were the deflection transistors, these drive the deflection coils on the tube and it seemed mostly likely that my issue of half a screen was down to a bad transistor in this section.

 

All four of the TO-3 power transistors had been replaced since new, the X section during the mid 1980s based on the 8534 date code on the parts (week 34 of 1985), and the Y section sometime after 1999.

 

IMG-9155.jpg

 

IMG-9156.jpg

 

Three of the four pre-amp transistors were original, but Q15 was a replacement twisted to fit the non-standard pinout of the original parts. I can't find any reference to this type being a suitable replacement, but it's in the X deflection section which seemed to be working fine.

 

All the transistors came off...

 

IMG-9157.jpg

 

... and tested fine, so much for my theory that one of them was to blame.

 

One major source of trouble on these board sets are the wire links, designated as W1-W9 on the silkscreen. As these are single sided boards when a track needs to jump over another a link on the parts side of the board is used. In this case they used zero Ohm resistors, which are easier for the assembly machines to do with than bare wire links, as it just treats them like any other resistor. For some reason they are failing in their old age and in some cases test OK when the board is off, but go open circuit when working voltages are applied.

 

IMG-9158.jpg

 

The advice is to replace them all with actual wire links, they just need to be thick enough, so who ever used a single strand of wire here was probably pushing it. Judging by the burn mark the original W2 let go rather violently, or this wire glows red hot.

 

IMG-9160.jpg

 

The soldering across the board was horrible too, gloopy solder, burnt flux and torn-up traces and either too little, or too much solder used so in places there was no way to tell by eye if the joints were any good. I suspect someone went round this trying to shotgun fix by touching up random joints and applying loads more solder. This is a bit like trying to fix a car by tightening every single nut, bolt and screw. Unlikely to fix the issue, and very likely to cause others. They also used far too much heat as many of the pads have lifted from the PCB.

 

IMG-9151.jpg

 

All the messy soldering had to be removed so I could see what was underneath. Often people try to fix track damage by sloshing solder over the area, which often doesn't work so I was half expecting the collapse to be a result of bodgy work hiding a gap somewhere. Adding new solder to an old joint often covers up issues without fixing them, and adding far too much solder can dome over the pin potentially without bonding to it, like these.

 

IMG-9152.jpg

 

Despite the shotgun approach to soldering touch up this this dry joint on capacitor C3 had survived untouched...

 

IMG_9172.jpg

 

...despite the component body being covered in soot from a nearby explosion at R7.

 

IMG-9174.jpg

 

It's in the right part of the schematic to shut down the Y deflection, but its impossible to tell if this is the smoking gun. The cap was certainly loose and the leg could be pulled free easily. If this is the issue then potentially the monitor was flaky for years, with the fault coming and going until the contact points oxidised past the point of no return.

 

I moved on to strip off the wiring harness as a few of the wires were badly frayed at the board joint, leaving only a few strands hanging on, plus it needed a wash.

 

IMG-9163.jpg

 

The fuse holders were corroded and sooty so came off...

 

IMG-9189.jpg

 

... followed by the the large capacitors, for cleaning and testing off board, and the PTC which was probably OK but had lost it's lid.

 

IMG-9165.jpg

 

I went over the board sucking up the massive amounts of solder and re-soldering the messy joints and then went over it all with a brush and iso-alcohol to remove the charred flux residue. It then got a wash with simple green on the solder-side and spent a few hours swinging the sun.

 

IMG-9190.jpg

 

Some of the dirt turned out to be paint over-spray and couldn't be easily removed. I managed to get some off but avoided trying anything too violent.

 

With all the air gaps in the circuit now was a good time to go round the board testing stuff. Testing in circuit isn't guaranteed as other paths can interfere with the tests, but with so much removed a lot of things were already out of circuit despite still being soldered into the board.

 

IMG_9513.jpg

 

Practically everything across the board tested as fine and within spec. Some parts did need to have one leg pulled to be measured fully independently of other parts.

 

The final tally ....

 

- Q1/Q2/Q3 were all the wrong type, the should be 2N3904s but had been replaced with 2N4401s

 

- R3 and R35 should be 15 Ohm resistors, measured OK but both looked roasted and their coating was flaking off.

 

- Diode at CR3 was a long way out of spec showing a fwd voltage drop of 1.4v, rather than 0.7v. Still a diode but unwell.

 

- Diode at CR1 was dead, totally open circuit.

 

IMG_9472.jpg

 

Aside from R35 everything above was in the Y deflection system, I'm guessing something exciting happened and it was half repaired, maybe successfully for a while. The transistors had all been replaced, along with the burnt up resistors, but either they didn't find the bad diodes or they took a hit during the incident and failed later. Either way it's a fair chance the my Y deflection issue was the result.

 

I'd already had two hefty deliveries from RS and wasn't in the mood to wait another couple of days, but sometimes Jaycar does come up with the goods. Surprisingly enough they stocked the correct transistors along with the bog standard parts.

 

IMG_9512.jpg

 

The incorrect transistors might be perfectly fine as replacements in this application, but I've been led on wild goose-chases before so always go for original parts when I can get them.

 

Using equivalent parts in a repair is a bit of a gamble, in theory the ones on the board were equivalent types but they so differ in some respects - apparently the 2N3904 is better at really low currents, but the 2N4401 handles higher currents better. The problem is it's hard to know what specifications were important in the manufacturer's decision to use one part over another, it could be some critical specification, or something as mundane as the unit cost on the day of ordering.

 

Replacing the diodes was easier, the official parts list stating they should all be 1N914s, but this board was born with HN4148 diodes installed from the factory, going by the original solder on most the few that hadn't been re-touched. As those are still a very common type I robbed a couple of a scrap TV chassis and slotted them in.

 

The capacitors are standard mylar type and as I only needed 2, rather than RS's minimum order of 50, Jaycar was the easiest option.

 

Rebuild time - wire links up first!

 

IMG_9468.jpg

 

Fuse holders next, using an old fuse clipped into ensure correct alignment while soldering.

 

IMG_9469.jpg

 

I'd also picked up a replacement connector for J105...

 

IMG_9515.jpg

 

... as the original was in a bad way. This is the connector for the deflection coils on the tube and the original wasn't making a firm contact.

 

IMG_9514.jpg

 

The circular profile pins are common in older gear, which provide a very thin contact area compared with square profile pins. I guess this is why someone had slopped solder on them to fatten them up a bit. This actually doesn't work well as solder is a bad conductor when oxidised so makes a dreadful contact metal.

 

The connector on the coil wiring looks to be original but has square holes and fits this header perfectly. I just had to punch one of the pins out as one pin is blanked off in the connector and because the pad had been blasted off the board leaving nothing to solder to at that position.

 

IMG_9517.jpg

 

Y deflection section rebuilt!

 

IMG_9518.jpg

 

Spare PTC goes in...

 

IMG_9519.jpg

 

... and onto the bit I wasn't really looking forward to, fitting the deflection transistors, with mica insulators and heatsink compound. The advice is still to go mica and paste here, rather than the neater, cleaner, easier silpad option, as these get HOT!

 

It's a filthy job, so much swearing later...

 

IMG_9527.jpg

 

Pre-Amp transistors go in, on more mica, and with a ferrite bead on both MSPU57's base pins. Only Q14 had a ferrite when I got the board, I'm guessing all these transistors were taken off for testing at some stage and the one Q4 got lost. Q4 got one salvaged from a scrap a TV power supply.

 

I also managed to track down a replacement for the MPSU07 at Q15, these are a long obsolete transistor type and have been hard to find for many years. There are modern equivalents but they have a different leg layout which is why the BD139 replacement at Q15 had its legs crossed over. If I was missing more than one I would have considered replacing the lot with small adapter boards, but I guess I like doing things the hard way to retain the original design. So I took a punt on this one, from a UK seller of older parts online.

 

IMG_9529.jpg

 

Buying rare parts online is a massive gamble but this auction had a few things in it's favour. Firstly the seller only had one left of a pair, rather than sellers who claim to have 10, 50 or 100 suspiciously new and shiny ones. A lot of fake transistors out there are other parts that have their markings stripped off and are labelled to-order with what ever new part number want. The assumption presumably being that the buyer won't bother to return them when they don't work, or don't really know what they are doing if they are only ordering a small number.

 

In this case the part didn't look too new, the casing has a lot of tiny marks and nothing uniform to suggest its been polished. Basically it looks like it has been kicking around in someones stock for decades. Also the printing wasn't up to modern laser etched standards and looks like it was screen printed years ago. It also had the weird colour stripes printed on the back that my originals did. It could still be a fake, but it would have taken the faker much more effort than they seem to want to put in.

 

IMG_9531.jpg

 

On arrival I put it through the my parts tester and it came back as the right type of transistor, the hFE was a fair bit lower than the others, but it was within the range that the datasheet says is normal. The Vbe was a good match for the others and matching this parameter may be more important in this circuit if I'm reading it right. The first power up will be interesting either way.

 

They were reinstalled with the original bolts heat shrinked on TO220 mica plates cut in half. Only a couple of the original bolts had heatshrink on them and even that was just a few shreds, but I think that was the original method. Also had to improvise on the washers to keep the bolt head away from the transistor fin, as only one of the originals had survived, with the previous fixer using cardboard washers and shreds of plastic film.

 

IMG_9533.jpg

 

One down, three to go.

 

IMG_9534.jpg

 

Set complete, plus the large capacitors cleaned, tested and reinstalled.

 

IMG_9574.jpg

 

New fuses go in.

 

IMG_9577.jpg

 

The harness went back on last so it wouldn't be in the way, shame I'd forgotten to fit the replacement X and Y size potentiometers beforehand.. These are soldered across two resistors in what looks like a bit of a post-production bodge.

 

I ended up bending the three legs so they could be mounted neatly as a tripod to be soldered against the resistor legs.

 

IMG_9578.jpg

 

IMG_9579.jpg

 

All that was left was to fit the recommended diode mods to the rear of the board...

 

IMG_9580.jpg

 

... and add the zener diode mods on the deflection coil outputs.

 

IMG_9581.jpg

 

Deflection board done!

 

IMG_9585.jpg

 

Time will tell if it actually works or not.

 

--- To Be Continued ---

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

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    • No registered users viewing this page.

×
×
  • Create New...