130XE brought back from the dead - Advice re glitching display

[Beeblebrox]

Tiny update. Still no joy restoring colour.

 

@kheller2One thing I did was test if I get a signal out of the RF modulator going into the RF aerial socket - Interestingly after tuning in the TV I get the ready prompt - again in black and white. Tweaking the trim pot has exactly the same effect as on composite and also S-video (eg the pink screen at a certian biting point but that's it). So that is another clue - RF is in black and white as well.

 

TBH I am getting a little overwhelmed as learning about schematics and electronics on the fly, cross referencing schematics with the silkscreen maps to locate components. Also making silly mistakes. (Trying to locate the C53 cap had proved completey elusive until I clocked I was looking at the XEGS schematic and my board doesn't have one! Another time I lost the image completely with the exception of a white line - then realised I'd left the CPU on the workbench! Doh!) 

 

The other thing I am struggling with is that I only have a digital multimeter. I have no oscilloscope. A lot of the references to checking signals coming from the various ICs and components I am not able to check. I am reading from various sources and field manuals, AA threads.

 

I've replaced resistors 202, 205, R203, R45 and 201 testing after each install. Nothing. Still nothing.

 

Initially Q1, Q2 and Q3  I just removed, took a reading them popped back on the board, (as per post #38 all have around the 0.900v for each which according to what I've researched means they are working.)

 

Then I decided to replace Q2 given it's importance in the Chroma circuit  - no difference with a new 2N3904 transistor in place.

 

A new batch of 74LS74 chips arrived today which I tried in socket and it makes no difference. (Even though I didn't think it was that chip causing the issue given my reports earlier on in this thread with swapping out the chip already with ones from my other A8's and alos the chip working. Again I am confident the socket for this chip and others is ok after testing connections along traces for all concerned).

 

Also I checked that the Chroma pin of the DIN socket had continuity with the board and the first component in the line to it. It did.

 

I've been following the Chroma line of components from GTIA pin 21 and testing and replacing as per above only.

 

Having no frame of reference with multimeters, this being my first digital one, it looks like mine has turned out faulty as not able to swtich to diode and capacitance mode testing. Very annoying. Will have to order another model.

 

So whilst the likes of L6 100uh inductor and L31 (FB) seem to have a confirmed continuity in place I can't get a resistance reading so I'll have to leave them alone for the mo and move on. 

 

I also decided to remove the old socket for the CPU and install a precision socket. (I took the liberty whilst the old socket was removed and PCB cleaned to check the continuity of all 40 vias and traces coming off them on both the topside and underside of the board. All was ok. I installed a precision socket and rechecked. With CPU in place in the new socket all is working as before. (Wasn't expecting it to make a difference).

 

So whilst I know I have a lot more to research and checking to do, I am a little burnt out. I really wanted to nail it this weekend too. Ah well.

 

[TGB1718]

2 things:-

25 minutes ago, Beeblebrox said:

Initially Q1, Q2 and Q3  I just removed, took a reading them popped back on the board, (as per post #38 all have around the 0.900v for each which according to what I've researched means they are working.)

For future reference:- When testing transistors, you also need to reverse the connections and make sure there is no reading (as described earlier)

25 minutes ago, Beeblebrox said:

So whilst the likes of L6 100uh inductor and L31 (FB) seem to have a confirmed continuity in place I can't get a resistance reading so I'll have to leave them alone for the mo and move on.

Resistance of inductors is usually very low, so getting no reading is normal. (it's basically a coil of wire)

 

I got one of these, it can test most passive components, resistors, capacitors, inductors.

It can also test diodes (as in the picture), most types of transistors.

 

It's quite clever, you don't need to put components in in any special way as long as the

component legs are in different testing slots, it sorts out which leg is which, tests and tells you the pinout.

Edited September 25, 2021 by TGB1718
Update

[chevymad]

Wouldn't a UAV bypass everything and be cheaper/easier then what you've done so far? 

 

Edit, if you're just in it for the challenge, when I was having issues I bought this pc scope. Not terribly expensive as compared to a good multimeter. 

 https://www.amazon.com/gp/product/B00HC4UP52/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1 

Edited September 25, 2021 by chevymad

[Beeblebrox]

5 minutes ago, chevymad said:

Wouldn't a UAV bypass everything and be cheaper/easier then what you've done so far? 

@chevymad Fair point although I've not spent much tbh. Ram chips I bought a bulk load of which I can use for my other A8s too and it's good to have spares. It's also a learning curve and good excuse/focus for me to gain more knowledge about electronics and the A8 hardware. I do like tinkering and it's likely I'll always have an A8 to fix.

 

With this 130XE I really want to see if I can get to the bottom of it as it is. I am really pleased I managed to bring a dead 130XE consigned to the spares pile back in the 90s to life!

 

As mentioned earlier in the thread I do have a UAV rev D installed in my main driver 800XL but I am loathed to rip it out and install it in this 130XE. That's simply because I'll need to pop it back in my 800XL at some stage and then I'll be back to square one with the 130XE.

 

Why not buy another UAV? Well the other consideration is I am in the UK and sourcing UAV from Brewing academy works out quite expensive with postage, import duties, etc. I got a quote from them the other month and dismissed the idea. Plus it would take ages to arrive. 

 

Time will tell if I can get it to work but the AA community have been great with the advice.

[Beeblebrox]

1 hour ago, TGB1718 said:

2 things:-

For future reference:- When testing transistors, you also need to reverse the connections and make sure there is no reading (as described earlier)

Resistance of inductors is usually very low, so getting no reading is normal. (it's basically a coil of wire)

 

I got one of these, it can test most passive components, resistors, capacitors, inductors.

It can also test diodes (as in the picture), most types of transistors.

 

It's quite clever, you don't need to put components in in any special way as long as the

component legs are in different testing slots, it sorts out which leg is which, tests and tells you the pinout.

@TGB1718   ooooo gettin me one of those straight away - nice lil toy! Thanks for the suggestion.

 

Re transistors - yup I am pretty sure I flipped and tested each as you suggested and didn't get a reading in each case.

 

Regarding inductors I read this:http://www.learningaboutelectronics.com/Articles/How-to-test-an-inductor

 

Quote: "

The inductor should read a very low resistance across its terminals, only a few ohms. If an inductor reads a high resistance, it is defective and should be replaced in the circuit.

If an inductor is reading very, very small resistance, less than an ohm (very close to 0Ω), this may be a sign that it's shorted. Functional inductors normally read a few ohms, greater than 1Ω and normally less than 10Ω. This is a healthy range for an inductance value. Outside this range and this is normally a sign the inductor is bad. So a resistance check is a simple but effective method for finding out if an inductor is defective or not."

 

 

[xrbrevin]

2 hours ago, TGB1718 said:

 

i have a variation of this device. it came in kit form and was fun to build

it also displays the rating of the tested component so you can tell if it is out of spec

[TGB1718]

41 minutes ago, xrbrevin said:

i have a variation of this device. it came in kit form and was fun to build

Same with this one, was very cheap and fun to build, even more fun when it worked  

my only criticism, I wish I could change the screen colours

[Beeblebrox]

I've long since come to the general conclusion that many of us, myself included, like the building of something more than using it after it's been built. (I am a bit obsessive like that).

[TGB1718]

It's even more fun when you design something, build it, code it and it works   

[+kheller2]

Putting a UAV in will immediately tell you what half of the video system to troubleshoot.  If you get black and white you know the problem is upstream. 

[Beeblebrox]

53 minutes ago, kheller2 said:

Putting a UAV in will immediately tell you what half of the video system to troubleshoot.  If you get black and white you know the problem is upstream. 

@kheller2good point. I suppose I didn't appreciate it would narrow things down as I assumed it was just replacing it all. OK, I'll carefully extract it from the 800XL. (prob make new wiring for the XE save me having to u solder lines from the 800XL.

I'll need to dig out new instructions for the hook up points for the XE. Prob a job for next week. If it works I'll then have to work out what components uav replaces as such so I can discount the ones it doesn't replace as working... (if that makes sense). Brain fried. Thanks ?

Edited September 25, 2021 by Beeblebrox

[bfollowell]

6 hours ago, TGB1718 said:

I got one of these, it can test most passive components, resistors, capacitors, inductors.

It can also test diodes (as in the picture), most types of transistors.

 

It's quite clever, you don't need to put components in in any special way as long as the

component legs are in different testing slots, it sorts out which leg is which, tests and tells you the pinout.

 

 

Not to get the thread too side-tracked, but I'd really like to know where to pickup such a device, if any of you that mentioned having one could pass a link along. It looks like it would be a useful piece of test equipment to have and a cool little project to put together.

 

Thanks in advance!

 

Edited September 25, 2021 by bfollowell

[Beeblebrox]

1 hour ago, bfollowell said:

 

Not to get the thread too side-tracked, but I'd really like to know where to pickup such a device, if any of you that mentioned having one could pass a link along. It looks like it would be a useful piece of test equipment to have and a cool little project to put together.

 

Thanks in advance!

 

@bfollowell

Seems you can get them in diy kits, or assembled. In the states on ebay they are dirt cheap:

https://www.ebay.com/itm/174034093041?_ul=GB

 

??

[chevymad]

This  is the one I have.  https://www.amazon.com/Transistor-DROK-Capacitor-Capacitance-Automatic/dp/B01MS1FOYM/ref=sr_1_1_sspa?dchild=1&keywords=electronic+component+tester&qid=1632610348&sr=8-1-spons&psc=1&smid=AFHAE9RJVUMB&spLa=ZW5jcnlwdGVkUXVhbGlmaWVyPUEzUE9XNjVEMDRQREk5JmVuY3J5cHRlZElkPUEwNzI4Mjg4M1RXTjVWRkdNTU1DOCZlbmNyeXB0ZWRBZElkPUEwNTU2NTY1M0dUQzE0TkNJUE9IMSZ3aWRnZXROYW1lPXNwX2F0ZiZhY3Rpb249Y2xpY2tSZWRpcmVjdCZkb05vdExvZ0NsaWNrPXRydWU=

 

[manterola]

On 9/23/2021 at 7:13 PM, Beeblebrox said:

With the removed (white) original 130XE trim/colour pot I was getting readings starting 450ohm and going up to 2.5v as I turned it

Not sure if you are measuring at the same point, but you should get up to 9v or so in the pot. Check the circuit I posted earlier in this thread. 

Also compare the behavior of the voltage as turning the pot, with the voltages of computer that's working correctly. 

 

 

Edited September 26, 2021 by manterola

[Sugarland]

@Beeblebrox have you tried continuity tests of the chroma and other traces to be sure they are functioning?

 

If so then you need to reach 100% improbability for fixing the issue. In that case, it should work.  Or rather, it's been working this whole time but you're suffering from changes to the environment and/or suffering morphological structure including black and white video hallucinations.

 

[Beeblebrox]

 

Nailed it!!!!!   Huge relief. 

 

Now here is a lesson I've learnt..... from the outset always try replacing a component rather than just testing it, and take more thorough notes along the way.

I got myself tied in knots at times and trying to do and read too many things at once only causes brain ache.

 

So after a good 5hrs today spent replacing a lot more components on the 130XE with my donor board, and after each swap testing to see if the damn colour has made an appearance....... the last component I replaced after revisiting it only bl00dy worked! Typical!

 

Today I new I was in for the long haul, with cat and other half locked out of the bedroom/work area.

 

I set to it reading up some more on the forum, schematics and troubleshooting guides. I fired up the iron and replaced L31, C50, C44, L6, the entire RF module, Q6, (relating to the GTIA pin 16 input via the crystal), again testing after each one.

 

I then decided to revisit the transistors and replace Q2. No change. 

 

Went to replace Q3 with the donor board Q3 2N3904 transistor. Powered up the XE and the screen at last had some flickers of colour!

 

    

 

Encouraged (and in disbelief tbh), I tweaked the trim pot hoping for something and bingo, blue ready prompt!  (So the donor trim pot I installed the other day is fine btw).

 

So transitors Q3 and the other Q2 and Q1 were components I'd tested before with the multimeter. Being a novice at this I should have popped in those components from the donor board - but no... I just popped em right back in and moved on at the time! Doh! 

 

(Hasn't helped that my multimeter had been clearly malfunctioning and as of this morning completely stopped being able to test for continuity and then promptly died on me .......and is now a useless plastic brick.  Sending it back for a refund.)

 

So all in all in addition to the components I replaced today I also replaced 99% of the components in the colour video circuit as well as others, including :

 

Both crystals, the 4050, 74LS74, 74LS138, most 40pin sockets, 555, R204, R45, R201, R205, R202, R203, R38 trim pot, and Q1 and Q2.

 

The whole board with the exception of a few small ICs is now socketed.

 

I may see issues down the line given the amount of work I've done of this, (so much soldering and heat stress can't be good). Before I screw down the case I'll go through and give it another clean to remove any solder debris, and double check for anything in the way of particularly bad solder points, etc.

 

However the main object of this whole exercise was to bring a good ol Atari 130XE back to life that was previously likely gonna end up as a donor board. And of course in the process further my understanding of electronics and Atari hardware. 

 

I also tried the Simple XE mod I've done on 2 x of my other XEs to improve the sharpness, reduce jailbars, etc. (Did this yesterday just in case it made a difference and in the vain hope it might actually help restore colour ...(well... i was getting a bit jaded and needed something heh heh!) 

 

I am really pleased I didn't have to rip out my UaV Rev D board from my 800XL to pop in this machine. Granted I totally see the point but it is happily installed in my driver 800XL and is working well.

 

To recap what essentially was wrong with this machine was:
 

• Half the crappy Micron MT ram chips had gone bad, which is the main cause the machine appeared dead all those years ago.
• A bad via on pin 2 of ANTIC was initially causing screen glitching.
• The transistor Q3 (2N3904) on the Color circuit from GTIA pin 21 leading to pin 5 chroma of the DIN socket was malfunctioning. 
   

So lesson learned. I've gained a good amount of beginner insight into Atari hardware and generally how schematics work.

 

Thanks so much to all who replied and gave assistance over the last week, very much apprecaited. 

 

Right I need a pint of Ale!!!!?

 

This is on S-video:

         

 

   

     

 

 

 

 

 

Edited September 26, 2021 by Beeblebrox

[flashjazzcat]

Nice one.  I'm gonna say I called the bad transistor in my post from Thursday, although I did hedge my bets with two other guesses too.

 

[TGB1718]

Well done, perseverance paid off, I think many others would have given up

[xrbrevin]

every day is a learning day..

great success!

[+kheller2]

And now any future Atari repairs will seem simple in comparison.  ?

[+DrVenkman]

1 hour ago, flashjazzcat said:

Nice one.  I'm gonna say I called the bad transistor in my post from Thursday, although I did hedge my bets with two other guesses too.

 

Yep. And on Thursday morning I basically said the same:

Quote

If it appears that the color pot is working, and you are confident there are no issues with the GTIA socket, then you're down to the discrete components in the video circuitry, which can be a pain to diagnose. I'm not familiar with the PAL output circuitry, but if there are transistors involved, it seems to me that - being active components - are much more likely to fail than are resistors and ceramic caps.

 

 

Anyway, glad the issue is finally sorted. 

[Beeblebrox]

12 minutes ago, DrVenkman said:

Yep. And on Thursday morning I basically said the same:

 

Anyway, glad the issue is finally sorted. 

@DrVenkman @flashjazzcat @TGB1718  and others - you called it indeed. As I say rookie mistake was not replacing the Q3 at the time of testing it. I just popped it back in thinking it was good and moved on.

 

My biggest problem is I have a terrible short term memory - which means if I glance away from the PCB /component I am working on or a schematic I am looking at, I instantly lose it and have to find it again. Makes things take twice as long. (I am terrible at Where's Wally (and over in the states Where's Waldo (I believe)) as a result heh heh!

 

Still I have absorbed what I consider for my novice status a satisfying amount of new knowledge and experience.

 

@kheller2 And now any future Atari repairs will seem simple in comparison.  ?

Yup - makes the upgrades I've been installing a breeze.

 

So - onto business...   Anyone wanna buy a newly refurbed 130XE?! AA mate's rates!  (Kiddin!) 

 

Edited September 26, 2021 by Beeblebrox

[xrbrevin]

most digital multimeters have a transistor tester to show the "HFE" rating. also, the cheapo component tester kit (discussed earlier) can show more specs.

you can do a continuity logic test on them but this will only test for short or open, not 'partial failure - operating out of spec'

[+DrVenkman]

I bought this little widget a few months ago and it works great for testing stuff out of circuit, including diodes, transistors and capacitors. 

 

https://www.amazon.com/gp/product/B08K3BGKXC/ref=ppx_yo_dt_b_asin_title_o00_s00?ie=UTF8&psc=1