Atari 130XE yellow screen issue

[runstop]

Hey everyone! I've just started my journey with Atari 8bit computers lately (I've had a 600XL sitting in my closet for close to 20 years, but I just got a 130 and an 800XL). My 800XL works like a treat and I've installed an ultimate video in my 600xl and that is awesome. However, my 130Xe boots up with a yellow screen. I've been trying to figure out what might be wrong with it with no luck. I'm sure I'm missing something obvious.. 

 

According to https://www.manualslib.com/manual/1248642/Atari-130xe.html?page=47#manual it might be U27, U1-U9, U12, or U13. So.. I swapped out every ram chip with TMS4164's (tested in my little ram tester), swapped the SALLY, Antic, GTIA, OS and BASIC roms with known good ones from my 600 (and tested the 130xe chips in the 600 to be sure), swapped the MMU over as well. I'm not really a hardware expert so I've just been following datasheets and trying to figure out the issue with my cheapo aliexpress oscilloscope, but no luck so far. Anyone have an idea where I should look next? The Freddie chip maybe? I would have to source one first. 

 

Thanks!

[Beeblebrox]

You say you have swapped out ic's. Given the XE boards are not socketed from factory had you socketed it yourself? If so with the exception of ram chips did you socket the others ic's in turn, checking after each any changes? 

 

Reason I ask is it's never good practice to go and socket multiple chips then test. You might have introduced an error, (especially if not used to working with a soldering iron or with these pcbs). Much harder to find the cause(s) if you have. 

 

My first port of call for a yellow screen would be checking ram, cpu, OS rom, and MMU. 

 

Also if you have socketed double check as much as you can your solder work? 

 

What type of sockets were used? 

 

Also are the ram chips Mt branded? They are notorious for failing. 

 

Can you post images of the top and bottom of the PCB? 

 

Also, where are you based!? 

 

 

 

 

Edited May 2, 2023 by Beeblebrox

[TZJB]

 

15 hours ago, runstop said:

Hey everyone! I've just started my journey with Atari 8bit computers lately (I've had a 600XL sitting in my closet for close to 20 years, but I just got a 130 and an 800XL). My 800XL works like a treat and I've installed an ultimate video in my 600xl and that is awesome. However, my 130Xe boots up with a yellow screen. I've been trying to figure out what might be wrong with it with no luck. I'm sure I'm missing something obvious.. 

 

According to https://www.manualslib.com/manual/1248642/Atari-130xe.html?page=47#manual it might be U27, U1-U9, U12, or U13. So.. I swapped out every ram chip with TMS4164's (tested in my little ram tester), swapped the SALLY, Antic, GTIA, OS and BASIC roms with known good ones from my 600 (and tested the 130xe chips in the 600 to be sure), swapped the MMU over as well. I'm not really a hardware expert so I've just been following datasheets and trying to figure out the issue with my cheapo aliexpress oscilloscope, but no luck so far. Anyone have an idea where I should look next? The Freddie chip maybe? I would have to source one first. 

 

Thanks!

 

Just to clarify, when you say your 130XE boots to a yellow screen, I take it there is no Ready prompt?

 

Assuming there is no Ready prompt, has that yellow screen been constant throughout your round of chip replacements?

 

You may need to go back to basics and continuity test each Row Address (RA) pin of Freddie back to each DRAM chip. i.e. Freddie pin 32 should be continous on all DRAM pin 5s. Pin 31 on all DRAM pin 7s etc. It just needs one break or bad connection to cause something like this.

 

Another trick to try is to swap the CAS banks by crossing over R110 and R111 to see if the secondary bank works as the primary. You don't need the secondary bank for it to boot to BASIC.

 

Edited May 2, 2023 by TZJB

[runstop]

Hey everyone, thanks so much for the replies!

 

@Beeblebrox: Yes, I've been desoldering each chip and socketing it, and then checking that chip in my 600xl (and vice versa). I desoldered each MT ram chip individually and swapped each one out with a tested TMS4164 chip and then tested that. I used higher quality round hole sockets for the bigger chips, but only had dual wipe sockets for the ram, but nothing changed at all between swaps. At least I have a nicely socketed board now LOL. 

 

@TZJB thanks for the tips. I will check all the pins between Freddie and the ram chips tonight and see if I missed something. 

 

Oh, I'm in Ottawa, Canada! I collect commodore and apple stuff mainly, but I'm loving these atari 8bits that I never had as a kid. 

 

John

[Beeblebrox]

Hiya John

Great, sounds like you methodically did the socketing, etc the right way. The build quality of the XE pcbs compared to the XL pcbs is pretty stark. I have socketed many A8 pcbs and the XE boards are far likely to result in lifted traces and issues. 

 

So see how you get on with tzgb's suggestions. 

 

The other thing you can do is if you have a Star raiders cart try using that. Irrc it bypasses the os among other things. 

 

Also highly recommended getting a sys check II unit. Great for trouble shooting. 

 

 

[runstop]

Thanks, everyone! I've tested everything tzgb suggested. Everything from the freddie and the MMMU to the ram looks good. I did test swapping R110 and R111 to see if that helped, but no luck. I'm a bit stumped! I'm going to contact Jurgen to get a syscheck II. I'll no doubt be picking up more of these fun machines in the future and it will be good to have anyway. I'll circle back when I get one and give an update (unless I figure out it in the meantime!)

[TZJB]

20 hours ago, runstop said:

Thanks, everyone! I've tested everything tzgb suggested. Everything from the freddie and the MMMU to the ram looks good. I did test swapping R110 and R111 to see if that helped, but no luck. I'm a bit stumped! I'm going to contact Jurgen to get a syscheck II. I'll no doubt be picking up more of these fun machines in the future and it will be good to have anyway. I'll circle back when I get one and give an update (unless I figure out it in the meantime!)

 

Thanks for the feedback. Sorry you are still having problems.

 

Since you have verified the DRAM connectivity, it wouldn't hurt to now extend continuity testing to the CPU and other main chips. A0 - A15 and D0 - 7 are common on many chips. Attached are the full set of Jerzy Sobola 130XE diagrams if you don't already have them, which are mainly correct and useful for continuity testing.

 

Also make sure the clocks are working when the yellow screen is showing. Freddie pin 2 CLK should have the clock from the crystal circuit which is output on Freddie Pin 37 OSC and connects to GTIA pin 28 OSC via an inductor.  Next check Antic pin 34 PHI0 is providing PHI0 through a gate of U18 to SALLY pin 37 PHI0 followed by the two clock outputs from SALLY at pin 3 PHI1 and 39 PHI2.  PHI2 sets the clock for almost everything else in the Atari and connects directly to Freddie pin 5, so a bit of a circle there really. A buffered Phi2 signal via a different gate of U18 should also appear at POKEY pin 7, ANTIC pin 29, and at PIA pin 25 so should appear sharper than the raw signals directly from SALLY.

 

If you don't have an oscilloscope, you could still check for fast clock pulses with a logic probe maybe? Hope this helps.

 

 

ATARI130XE-SCHEMATIC.PDF

Edited May 5, 2023 by TZJB

[runstop]

@TZJB thanks for the suggestions and for the attachment! This is a big help. I do have an oscilloscope and I'll start testing when I get a chance this weekend. 

[_The Doctor__]

swap PIA, Pokey and check the supporting glue chips.

[Beeblebrox]

3 minutes ago, _The Doctor__ said:

swap PIA, Pokey and check the supporting glue chips.

here's a question. What are glue chips?  I've never come across the term and would love to know.

[_The Doctor__]

A chip containing a small number of simple logic functions that is used to connect or 'glue' larger, more complex chips together. The 74xx 74ls 7F etc ...

crucial. Timing in the Atari and to the periphery is ruled be these...

On a side note...

A great example of glue logic mastery is inside the ICD Multi I/O... still love the device.

Edited May 5, 2023 by _The Doctor__

[runstop]

On 5/5/2023 at 9:39 AM, TZJB said:

Also make sure the clocks are working when the yellow screen is showing. Freddie pin 2 CLK should have the clock from the crystal circuit which is output on Freddie Pin 37 OSC and connects to GTIA pin 28 OSC via an inductor.  Next check Antic pin 34 PHI0 is providing PHI0 through a gate of U18 to SALLY pin 37 PHI0 followed by the two clock outputs from SALLY at pin 3 PHI1 and 39 PHI2.  PHI2 sets the clock for almost everything else in the Atari and connects directly to Freddie pin 5, so a bit of a circle there really. A buffered Phi2 signal via a different gate of U18 should also appear at POKEY pin 7, ANTIC pin 29, and at PIA pin 25 so should appear sharper than the raw signals directly from SALLY.

Hi TZJB (and everyone!),

 

So I've been busy trying to debug this after I got my amazing sys-check II in the mail today. SCII is working great on my 800XL and my 600XL, but there is still no display on my 130xe. Here's what I've found so far with my oscilloscope:

 

Freddie Pin 2 has the 14.3mhz clock

Freddie Pin 37 has a 3.58hz signal on it. Is this right?

GTIA 28 - same 3.58hz signal.

 

Antic 34 - 1.79mhz 4.5vptp

That same signal follows through onto sally pin 37 and the same on pin 3 and 39, but voltage is around 4v ptp.

..and the same onto pin 5 of the freddie. 

 

u18 pin 6 and and pokey pin 7 have the same signal, but the votlage seems to drop a bit under 4v. and same onto the PIA at 25.

 

It feels like there are clock issues and voltage drops, but I'm not entirely sure what the proper values should be.. but I'm having fun learning! Any suggestions on what might be the issue or what I should check next?

 

 

 

[+DrVenkman]

I don't remember the clock frequencies for PAL systems off-hand, but the minor voltage drops on various clock signal pins is normal. Those signals are all generated by other components off the 5VDC rail and several of them pass through other chips or components on the way to your meter. So long as the clock signals are stead and regular (and in-spec) they're fine. 

[runstop]

Ok, thanks! This is an NTSC machine, but everything looks stable. Hmm!

[+DrVenkman]

3 minutes ago, runstop said:

Ok, thanks! This is an NTSC machine, but everything looks stable. Hmm!

Ah, sorry. No idea why I was thinking it was PAL.  Do you have a copy of the SAM'S COMPUTERFACTS for the 130XE? If not, here you go:

 

http://www.atarimania.com/documents/130XE - Sams ComputerFacts.pdf

 

I'm a big fan of the various SAM'S volumes for different machines and devices. Great troubleshooting flowcharts, pinouts for everything with expected voltages and key 'scope waveforms to be expected. Excellent, to-the-point nuts and bolts diagnostic stuff. 

[TZJB]

12 hours ago, runstop said:

Hi TZJB (and everyone!),

 

So I've been busy trying to debug this after I got my amazing sys-check II in the mail today. SCII is working great on my 800XL and my 600XL, but there is still no display on my 130xe. Here's what I've found so far with my oscilloscope:

 

Freddie Pin 2 has the 14.3mhz clock

Freddie Pin 37 has a 3.58hz signal on it. Is this right?

GTIA 28 - same 3.58hz signal.

 

Antic 34 - 1.79mhz 4.5vptp

That same signal follows through onto sally pin 37 and the same on pin 3 and 39, but voltage is around 4v ptp.

..and the same onto pin 5 of the freddie. 

 

u18 pin 6 and and pokey pin 7 have the same signal, but the votlage seems to drop a bit under 4v. and same onto the PIA at 25.

 

It feels like there are clock issues and voltage drops, but I'm not entirely sure what the proper values should be.. but I'm having fun learning! Any suggestions on what might be the issue or what I should check next?

 

 

 

 

Yes Freddie pin 37 OSC should be the 14.31818 MHz frequency divided by 4 = 3.579545 MHz so looks good. This should go straight to GTIA pin 28 and be the same 3.579 Mhz. This is output from GTIA pin 29 FO0 from what I can make out and fed into Antic Pin 35 FO0. This is output on Antic pin 34 clock O0 which is OSC divided by 8 = 1.78977 MHz, and feeds directly into Sally pin 37 O0 and Sally then generates outputs pin 3 clock O1 and pin 39 clock O2 which is probably the most important clock in the machine. Note that there are phase differences between the various O clocks but they are the same frequency 1.78977 MHz.

 

Clock O2 should go unbuffered from Sally pin 37 to Freddie pin 5 O2 and the EMMU U34 pin 15. U18 pin 3 buffers clock O2 in order to feed Antic pin 29 O2, GTIA pin 30 O2, PIA pin 25 O2, Pokey pin 7 O2. It also ends up on J4A the Cartridge interface pin 5.

 

Logic signals above 3 Volts are normally recognised as high so 4 Volts is good. U18 pin 6 is clock O0 whereas Pokey pin 7 is clock O2 as is the PIA pin 25 so U18 pin 3 must be outputting clock O2 correctly. Now just check clock O2 on the other main chips referencing the paragraph above.

 

I still suspect that you may have a physical break in at least one of the board traces or at one of the IC pins somewhere so it may come down to tracing every line. Since you have checked and verified continuity for the DRAM Address lines to Freddie the next place to start is the 16 Address lines that go to every large chip, followed by the 8 Data lines. Also check U19 the 555 timer is performing the RESET from pin 3 when the machine is initially powered on.

 

I have attached the 130XE Field Service Manual to assist you. Choose the circuit diagram relevant to your model.

 

130XE Field Service Manual.pdf

 

[runstop]

Thanks, everyone! I'm going to try to find time to work on it this weekend and will report back. 

[TGB1718]

On 5/17/2023 at 11:20 PM, runstop said:

u18 pin 6 and and pokey pin 7 have the same signal, but the votlage seems to drop a bit under 4v. and same onto the PIA at 25

Don't worry, those voltages are well in spec. See below.

 

“Acceptable” input signal voltages range from 0 volts to 0.8 volts for a “low” logic state, and 2 volts to 5 volts for a “high” logic state. 

“Acceptable” output signal voltages (voltage levels guaranteed by the gate manufacturer over a specified range of load conditions) range from 0 volts to 0.5 volts for a “low” logic state, and 2.7 volts to 5 volts for a “high” logic state:

 

[runstop]

Thanks everyone! Sorry, real life has gotten me recently here.. but I'm back looking at this. It looks like the 02 signal is fine everywhere, but I am getting some strange jittery readings on the EMMU. The 02 seems stable, but the CASMAN line seems weird and the CASBNK is just at 5v (is that correct?). I wish I had another C025953 chip as I'd just swap it out and test it. I can't find any shorts anywhere on the board, but I haven't tested everything yet. 

[TZJB]

15 hours ago, runstop said:

Thanks everyone! Sorry, real life has gotten me recently here.. but I'm back looking at this. It looks like the 02 signal is fine everywhere, but I am getting some strange jittery readings on the EMMU. The 02 seems stable, but the CASMAN line seems weird and the CASBNK is just at 5v (is that correct?). I wish I had another C025953 chip as I'd just swap it out and test it. I can't find any shorts anywhere on the board, but I haven't tested everything yet. 

That's unfortunate this and FREDDIE are the only chips you don't have from a 800XL. Is CASIN from FREDDIE pin 35 CAS oscillating correctly?

 

However, you can bypass the EMMU to test it as a 65XE.

 

See the 130XE Field Service Manual I posted above page 30 section 2-12. Fit a 20 pin IC socket for U34 EMMU, just in case you might need it later for a GAL16V8 mentioned later. Now jumper pins 1&16, 2&17 and 6&15. You now have a 65XE/800XE model. There is a picture on page 31 for clarification.

 

If you think you need another C025953 chip you have the choice of acquiring another one or making your own from @Mathy website https://mathyvannisselroy.nl/special stuff.htm using the JED file and a GAL16V8/74LS95B combination. The 74LS95B ensures ANTIC compatibility but the memory will work as normal without it.

Edited May 29, 2023 by TZJB

[Beeblebrox]

@runstop  soz if asked before, but where are you based?

[runstop]

8 hours ago, TZJB said:

That's unfortunate this and FREDDIE are the only chips you don't have from a 800XL. Is CASIN from FREDDIE pin 35 CAS oscillating correctly?

 

However, you can bypass the EMMU to test it as a 65XE.

 

See the 130XE Field Service Manual I posted above page 30 section 2-12. Fit a 20 pin IC socket for U34 EMMU, just in case you might need it later for a GAL16V8 mentioned later. Now jumper pins 1&16, 2&17 and 6&15. You now have a 65XE/800XE model. There is a picture on page 31 for clarification.

 

If you think you need another C025953 chip you have the choice of acquiring another one or making your own from @Mathy website https://mathyvannisselroy.nl/special stuff.htm using the JED file and a GAL16V8/74LS95B combination. The 74LS95B ensures ANTIC compatibility but the memory will work as normal without it.

 

CASIN on FREDDIE pin 35 was reading strange (jumped from 1.34mhz to 1.79mz) so I went ahead and socket U34 and pulled out the chip. I jumpered 1/16, 2/17 and 6/15. Same results.

 

When I pull out C061618 the CASIN reading on the FREDDIE is stable. Is this jumping around expected with that chip? Sorry, this is my hardware newb side showing though..   I've tested the C061618 in my 600xl and it works fine. I'm thinking there's something wrong around here, but I'm just not quite sure what it is. 

 

@Beeblebrox I'm in Ottawa, Ontario Canada where it's been lovely and sunny for the first time in a while.

 

[TZJB]

On 5/29/2023 at 10:12 PM, runstop said:

 

CASIN on FREDDIE pin 35 was reading strange (jumped from 1.34mhz to 1.79mz) so I went ahead and socket U34 and pulled out the chip. I jumpered 1/16, 2/17 and 6/15. Same results.

 

When I pull out C061618 the CASIN reading on the FREDDIE is stable. Is this jumping around expected with that chip? Sorry, this is my hardware newb side showing though..   I've tested the C061618 in my 600xl and it works fine. I'm thinking there's something wrong around here, but I'm just not quite sure what it is. 

 

@Beeblebrox I'm in Ottawa, Ontario Canada where it's been lovely and sunny for the first time in a while.

 

 

Canada sounds like London. We currently have some lovely weather too, although the temperature is a bit variable. But that's the weather for you...

 

Back to the matter in hand. Is there any sound as you boot up the 130XE?

 

I would not expect FREDDIE pin 35 CAS output to fluctuate as you were seeing. You haven't said but I advise that while you are still faulting this board, the second DRAM bank should be removed.

 

Since you have the oscilloscope out, it wouldn't hurt to check how every DRAM pin is looking now and recording discrepancies. Bear in mind that each DRAM chip should have the DATA pins 2 and 14 connected on each individual chip, each representing a Data Bus bit connecting to the Data Bus D0-D7, which connects to every large chip. i.e U9 pins 2 & 14 should go to U8 SALLY pin 33 etc. I can remember breaking one of these links between pins 2 & 14 once and the machine not working.

 

If you wouldn't mind, please upload high resolution pictures of both sides of your 130XE. Sometimes we can spot things that you may have missed.

[runstop]

Hey everyone! So sorry for the radio silence on this.. real life, etc.. But I'm looking at it again here this morning. I've checked all the dram chips and it does look like U9 2&14 are indeed connected, but the weird thing is I get nothing between U9 2 and Sally 33, nor do I get connectivity between Sally 33 and Antic 30. Very strange. I am going to bodge it here to see if I get anything. All the other dram connections to sally and antic seem fine.

[runstop]

Here is the board. Maybe something stands out?