The Star Wars (Atari) repair and maintenance topic

[Belike]

Wow, just a fuse holder, lol.
I guess most of us would gave this up, thanks for the repair and for the great explanation.
Send you a pm. soon.

[Belike]

Seems like I only had 3 working deflection pcbs for 1 day.
After a 3-4 months break, I fired up my Star Wars and saw some flames coming from the deflection pcb.
Fortunately the back door was open and I stood behind the cab, so I turned it off after a sec.
Seems like C703 burned, it is a 0,47K100 resistor or something and I can see also some burns on R711 and R712.
I didn't dare to install the other deflection pcb from my ESB, any ideas where to start?

[level42]

6100s........Maybe we should just burn them all.....

What board version is it, P314,P327 or Something else ?

First check the big transistors on the chassis with your multimeter in diode test setting. John just did a good video about it:

http://youtu.be/nxuRSEtjaCo


Oh and C703 is a capacitor not a resistor.
Rxxx = resistor
Cxxx = capacitor
Dxxx = Diode
Fxxx = Fuse
Qxxx = transistor...........don't ask me why it's not a T

[level42]

To go back to the one I fixed:

This was the picture I meant Etienne:



I also noticed a nice hint that Béla wrote on his thread about this on KLOV forum:

Switched off the machine, disconnected that plug, cleaned the F700 fuse under it and when I turned it on again, there was just a vertical thin green line in the middle of the monitor for 10 sec, than nothing.

My theory:

Corrosion is usually not so much a problem because....the stuff that corrodes is in the open air. So f.i. for a fuse, the holder and the fuse caps may have corrosion but as long as you leave it as it is (and it works) there's no problem because there is no corrosion between the metal where the holder and the fuse touch each other.

BUT....when you take out the fuse, and put it back turned around a bit, the corrosion will now be between the holder and the fuse

Pretty sure that this is what has happened

However, it was good to replace them anyway.

While we're on this subject: there were other parts like transistors and diodes that showed some corrosion on the leads too. But as long as they are in their original soldered spots there's no problem.
At one point I noticed (by feeling, carefully bending) that one of the small transistors was not "fixed" like it should. By just moving it around a couple of times one of the leads broke.
So I removed the old solderings and put the transistor all the way down to the board (I hate it how they kept the leads so long on the 6100's, another quality issue). However, once I did this I could not solder back the transistor legs anymore. The solder would not "grab" on the leads.....the reason was the corrosion on the legs.
Just one solution: new transistor

SO......corrosion is definitly something to keep an eye on...I haven't seen it much on PCBs but this one obviously has been in a moist surrounding for quite some time...

[Etienne MacGyver]

The solder would not "grab" on the leads.....the reason was the corrosion on the legs.
Just one solution: new transistor

Or high grid sandpaper, (depending on the cost and availability of the transistor ofcourse)

That fuseholder indeed looks corroded, and i think your explanation is the correct one.

Again, i guess its not worth the trouble to sandpaper that, and just throw in a new holder and fuse, but i think sandpaper also solved the issue.

the theory of the exposed parts corroding is correct, but i think also the corrosion will grow in between the fuse and holder as its not soldered or hermatically enclosed, so there is air there and like battery acid, it will grow in there

like you said, best thing is just comletely eliminate the corrosion and replace it complete

[level42]

Yeah well I have the small sized fuses in plenty and the holders were still here

You are partially right about the corrosion, however, if it turn the fuse in between my fingers I can easily see what spots have been in contact with the holder (those have no or minimal corrosion) and the spots that have been in the open air clearly.

I really think if I had put back the fuse "the right way" the issue might have been gone.

Anyway, this was the best solution that will work again for some 30 years. Also IF the fuses blow now you can get those size fuses almost everywhere for very little money, getting the US sized one's is harder and/or more expensive.

[Etienne MacGyver]

Yep, right thing to do was replace it completely, and indeed, for future replacements its alot easier for us europeans than US size !

all in all, perfect repair ! 

[Belike]

Thanks for the tip André, I watched the first half hour of John's video, it's a really good one, his explanations are great for noobs like me.

[Belike]

It is a P314.
I watched John's video carefully, removed the monitor from the cab and tried to test the big transistors on the metal frame.
There are 6 of them, 3 pieces of 2N3792 and another 3 2N3716.
I measured them one by one, 3 of them look like ok, the situation with the other 3 is the following:
2N3716:  with the black lead of the multimeter on the base and the red on the emitter the value is 1228, there should be no reading
2N3716: no readings at all with all six measure types
2N3792: no readings at all with all six measure types

Seems like 3 transistors gone at the same time, is it possible?

[level42]

So you got your multimeter in diode test ? and what do you mean with no reading at all, zero's on the screen and a beeping sound?

By the way.....this was another thing that could have put me off the right track with fixing your board that I got here.....if you set the multimeter in diode check and you measure a resistance like a couple of ohms, it will still beep.....you'll think nothing is wrong with it.

So,  the only proper way to measure fuses 100% is from soldering pad to soldering pad and in Ohms setting only !!

But that's just something else...

[Belike]

Ok, some pics would say more than 1000 words.
I used the multimeter (crap one) in diode test mode, if I don't touch anything, there is a number "1" on the display.
This is the metal frame with the transistors:

The 3716 should show a value between 500-900 with the red lead on base and black on emitter.
This is the good one:

The suspected bad one with the same lead setup:

The 3792 should show the same value with the leads exchanged:
The good one:

The bad one does not show the same "1" as the 3716, it shows 000.

I tried it in Ohms setting, I got the same values.
Also tried between the base and collector, same values.

[level42]

The one showing the 000's is DEAD

Full short on all legs....replace it

The one showing the 1 ....remove it from the socket and measure again....there could also be a bad contact on the socket.

The blown transistor "pulled down" the burnt resistors...Because the transistor died, a much higher current was running through these parts and this made them burn.

You'll have to replace those burnt parts too and have to check the transistors on the deflection PCB too.

[level42]

Your "first" PCB has just past a 2 hours test in attract mode on my SW. I think it is fit for the trip home.

The flickering spot killer LED stays though....not sure if that is a problem or not, I asked at KLOV.

[Belike]

Many thanks my Friend for your advises and for the pcb.
I found some 3792 in Budapest, but have to order the 3716s from ebay.
Or anybody has a few pieces for sale here?

[level42]

I can get replacements...(See my "equivalents" thread) but at this moment there is some discussion on KLOV about those replacing transistors causing some small issues on the picture.

It has to do with the "parasitic capacitance" of the transistors....this is a specification not often listed on datasheets but could have quite some influence.

In fact, I'm about to replace one of the transistors on my original Amplifone deflection board with an original 2N3716 to see if the picture will improve.
I think the only "big" issue is that when SW shows a lot of text you'll see some "squiggles" in the short vectors that make up the text.

I have always rendered this as something common on color vectors on SW machines because I can't remember seeing one without them....

On my Amplifone deflection board I have still 3 transistors that have been there from the factory (all RCA and production date are very close to SW's production year) but one has been replaced with another one. So, I'll put a 2N3716 in there to see if there's any improvement.

If not, I can get plenty of replacements or indicate them for you so you can source them locally.