Peri Noid Posted December 28, 2023 Share Posted December 28, 2023 (edited) I need your assistance in repairing a faulty XEGS. It gives me a black screen (synchro present, clocks are fine). I initilally, by default, replaced the RAM chips with known good ones (tested on another board) but it didn not change the behavior (the removed chips were faulty anyway). The Diagnostic Cart boots and seems to work (so I can assume, the CPU and Freddy seem to be fine). But during the RAM test I'm getting a report about a faulty bit 3 (counted from 1 so it's D2) in address area C000 (so just under the ROM). The ROM test passes. I desoldered the ROM anyway and put it into an XE board - it worked so I'm assuming, it's fine. I also tried to use ROMS from XE/XL machines - the behavior was the same, only the reported faulty bits differed from chip to chip. My next suspect was the MMU but... XEGS has a dedicated MMU, which is not compatible with XL/XE systems. And I have no another XEGS MMU to swap and check. And I'm kind of stuck here 😕 Looking at your experience and knowledge - does anyone have a suggestion, what else can be wrong here if it's not an MMU? Does there exist a JED file for a GAL, that would allow me to create a GAL replacement for the C101686 MMU? I tried to find one but couldn't (no problem with a standard C061618 but it's different). Edited December 28, 2023 by Peri Noid Quote Link to comment Share on other sites More sharing options...
+Nezgar Posted December 28, 2023 Share Posted December 28, 2023 1 minute ago, Peri Noid said: Looking at your experience and knowledge - does anyone have a suggestion, what else can be wrong here if it's not an MMU? Does there exist a JED file for a GAL, that would allow me to create a GAL replacement for the C101686 MMU? I tried to find one but couldn't (no problem with a standard C061618 but it's different). XEGS MMU JED is in this post here: 1 Quote Link to comment Share on other sites More sharing options...
Peri Noid Posted December 28, 2023 Author Share Posted December 28, 2023 Thank you very much. So... it didn't help, it's not the MMU 😕 Any other ideas? Thanks in advance... Quote Link to comment Share on other sites More sharing options...
xrbrevin Posted December 28, 2023 Share Posted December 28, 2023 have you considered the OS ROM? they can be flaky Quote Link to comment Share on other sites More sharing options...
Peri Noid Posted December 28, 2023 Author Share Posted December 28, 2023 Yes, I have. As written above, I desoldered the OS chip and due to lack of another XEGS, I put it in an XE board - and it worked. So, I suppose, it's not the ROM chip. Quote Link to comment Share on other sites More sharing options...
Beeblebrox Posted December 28, 2023 Share Posted December 28, 2023 (edited) Given the is rom and mmu in the xegs are both different chips to the standard XL/XE counterparts I am surprised the os rom worked when you put it into the xl/XE. I would still suspect the xegs mmu and os rom chips if you are 100% sure the replacement ram chips and their sockets are good. I had a failing xegs which was sorted when I replaced those chips. The other way you can test it is by installing u1mb in the xegs. Since it replaces the os rom and mmu anyway it's a good way of testing it. If you have an u1mb you can transplant into it from another machine, you can just socketed the mmu and os rom on the xegs and hook up the cpu 4 signals. I've done this in xegs consoles 4 or 5 times now. Edited December 28, 2023 by Beeblebrox 1 Quote Link to comment Share on other sites More sharing options...
Peri Noid Posted December 28, 2023 Author Share Posted December 28, 2023 I think I have one U1MB that I can transplant for the tests. I'll try. Thanks for the tip. Quote Link to comment Share on other sites More sharing options...
+scorpio_ny Posted December 28, 2023 Share Posted December 28, 2023 1 hour ago, Peri Noid said: I need your assistance in repairing a faulty XEGS. It gives me a black screen (synchro present, clocks are fine). I initilally, by default, replaced the RAM chips with known good ones (tested on another board) but it didn not change the behavior (the removed chips were faulty anyway). The Diagnostic Cart boots and seems to work (so I can assume, the CPU and Freddy seem to be fine). But during the RAM test I'm getting a report about a faulty bit 3 (counted from 1 so it's D2) in address area C000 (so just under the ROM). The ROM test passes. I desoldered the ROM anyway and put it into an XE board - it worked so I'm assuming, it's fine. I also tried to use ROMS from XE/XL machines - the behavior was the same, only the reported faulty bits differed from chip to chip. My next suspect was the MMU but... XEGS has a dedicated MMU, which is not compatible with XL/XE systems. And I have no another XEGS MMU to swap and check. And I'm kind of stuck here 😕 Looking at your experience and knowledge - does anyone have a suggestion, what else can be wrong here if it's not an MMU? Does there exist a JED file for a GAL, that would allow me to create a GAL replacement for the C101686 MMU? I tried to find one but couldn't (no problem with a standard C061618 but it's different). What is the condition of the PCB and sockets? Is there any corrosion on any traces? Quote Link to comment Share on other sites More sharing options...
Peri Noid Posted December 28, 2023 Author Share Posted December 28, 2023 35 minutes ago, scorpio_ny said: What is the condition of the PCB and sockets? Is there any corrosion on any traces? THe PCB looks acceptable. It's not... mint, but has no sign of visible corrosion. At least I havent's found any. Quote Link to comment Share on other sites More sharing options...
Beeblebrox Posted December 28, 2023 Share Posted December 28, 2023 (edited) 1 hour ago, Peri Noid said: I think I have one U1MB that I can transplant for the tests. I'll try. Thanks for the tip. No worries. If it doesn't work then it'll point to either possible bad sockets or traces near or related to those two chips/sockets, or other chips, traces, etc, etc, perhaps. You can solder the 4 x cpu signals to the underside of the pcb directly to pins 40, 39, 35 and 36. I pass the 4 x wires through the top of the pcb in holes around the colour pot, to then solder them. Edited December 28, 2023 by Beeblebrox Quote Link to comment Share on other sites More sharing options...
Peri Noid Posted January 4 Author Share Posted January 4 I fitted a U1MB into this XEGS and... it didn't boot. And it's not the Pokey's fault since I desoldered it and checked somewhere else. What to check next? Any suggestions? Quote Link to comment Share on other sites More sharing options...
Beeblebrox Posted January 4 Share Posted January 4 I am increasingly thinking this is an issue with the pcb, (traces/vias/solder/socket) rather than a chip or passive component. Quote Link to comment Share on other sites More sharing options...
+tf_hh Posted January 4 Share Posted January 4 On 12/28/2023 at 10:03 PM, Peri Noid said: I need your assistance in repairing a faulty XEGS. It gives me a black screen (synchro present, clocks are fine). I initilally, by default, replaced the RAM chips with known good ones (tested on another board) but it didn not change the behavior (the removed chips were faulty anyway). The Diagnostic Cart boots and seems to work (so I can assume, the CPU and Freddy seem to be fine). But during the RAM test I'm getting a report about a faulty bit 3 (counted from 1 so it's D2) in address area C000 (so just under the ROM). The ROM test passes. I desoldered the ROM anyway and put it into an XE board - it worked so I'm assuming, it's fine. I also tried to use ROMS from XE/XL machines - the behavior was the same, only the reported faulty bits differed from chip to chip. I think this is a very early version of a CPS SALT or SuperSALT cart and won´t work correct with the XEGS. All CPS SALT/SuperSALT versions I know of didn´t really test anything above 48K ($C000). And the "error" message is clear, $04 is the content of the first byte in the XEGS operating system (ROM file offset $4000). I would suggest the PIA is bad. The standard RAM test (48K) is okay, the ROM is okay, ANTIC and GTIA also seems to work. MMU you´ve checked. So PIA should be the one. Just watch PB7 (pin 17) with scope and power on the machine without diagnostic cart. Typically defect PIAs shown a continious low signal or remain always high. All XL/XE operating systems must follow this behaviour: After power-on the PB7 go high, after a few tenth of a second to low, also for a short time, then high again. This is used to mirror the selftest into $5000-$57FF. O.S. uses part of the selftest routines during cold start for a quick RAM test and detected existing RAM size. Without the O.S. will hung up. A diagnostic cartridge (special bytes in it´s header) will started before this test is done, so that´s why it starts. 2 1 Quote Link to comment Share on other sites More sharing options...
Peri Noid Posted January 4 Author Share Posted January 4 I'm using this cart, which is dedicated for the XE line actually: You were right, @tf_hh. It was the PIA. After replacing, the machine is operational again. 2 Quote Link to comment Share on other sites More sharing options...
Beeblebrox Posted January 4 Share Posted January 4 Great news. 1 Quote Link to comment Share on other sites More sharing options...
+MrFish Posted January 4 Share Posted January 4 20 hours ago, Peri Noid said: I'm using this cart, which is dedicated for the XE line actually: There's a newer version of this cart available. It was just fairly recently dumped. It coincides with this version of the 130XE FSM. I need to get the ROM posted to my site, but haven't gotten around to it yet. [Edit] Cartridge ROM added to my website: 130XE FSM / Production Test Cartridge 2 2 Quote Link to comment Share on other sites More sharing options...
Peri Noid Posted January 4 Author Share Posted January 4 Thanks. I saved it for future needs. I'd need to reprogram my SALT cart (I bought a ready one with 4 versions of SALT ROM). Quote Link to comment Share on other sites More sharing options...
+MrFish Posted January 4 Share Posted January 4 BTW, the main screen shows "Atari 130XE Tests" (in the screenshot above); but this is only the display it shows for an XE system with 128 KB or more. If you boot it with a 65XE or XEGS with 64 KB, you'll get the following screen. 3 Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.