A Tale of three XEs - hardware mods and issues help needed

[Beeblebrox]

@VoxelThanks. OK, then not sure what it was as was pretty happy with components and soldering at the time. I think before I attempt it again and solder them back on, I'll wait for the 910 ohm resistor to arrive.  Thanks for your continued help. ?

[Voxel]

Just to clarify as previously advised by TGB1718, the 10K ohm + 1K ohm in parallel gives 909.0909 ohm's and in case you needed assurance your soldering of the two resistors looked fine.   So if you feel ready before the 910ohm resistor arrives you should be fine.

[Beeblebrox]

1 hour ago, Voxel said:

Just to clarify as previously advised by TGB1718, the 10K ohm + 1K ohm in parallel gives 909.0909 ohm's and in case you needed assurance your soldering of the two resistors looked fine.   So if you feel ready before the 910ohm resistor arrives you should be fine.

@Voxel many thanks for clarification - I've replied to your PM btw.  

[marauder666]

the 2 resistors in parallel are around 910 ohms, you need 910K ohms. 1000x bigger.

 

I've got some 910K ohms resistors, PM me and I'll send you some.

 

[Beeblebrox]

41 minutes ago, marauder666 said:

the 2 resistors in parallel are around 910 ohms, you need 910K ohms. 1000x bigger.

 

I've got some 910K ohms resistors, PM me and I'll send you some.

 

@marauder666 many thanks for the offer, v. kind of you, but it was actually just a typo in my last post. I meant 910K, just the K dropped off due to my cack-handed typing skills.   I've been using in parallel a 10k and 1k resistor combo with my first attempt to give me the same as the 910k resistor. (I've got a few of the latter coming from China which I ordered the other week after failing to source them in the UK). thanks anyway. I am hoping my first attempt at this mod failing was just down to some bridged soldering. I will likely plan to re-attempt the mod when the 910k resistor turns up shortly. Thanks again 

 

[marauder666]

The 10K in parallel with the 1K is 910 ohms.  Not 910K ohms.

[Beeblebrox]

49 minutes ago, marauder666 said:

The 10K in parallel with the 1K is 910 ohms.  Not 910K ohms.

Doh!! See this is why I failed Maths GSCE BITD.   ok, so in fact to run the parallel resistor trick I'd have needed a (EDIT) 1M and 100K  combo to give me the 910k (or thereabouts)?  You can tell I know diddly squat about electronic circuit boards can't ya?  I have the 910K coming this weekend which I ordered here the other week as I say. Thanks for ya help.

Edited June 21, 2021 by Beeblebrox
typos

[marauder666]

1m and 10m

[Beeblebrox]

 

2 minutes ago, marauder666 said:

1m and 10m

Sure, I saw my mistake but ya beat me to it. Thanks. Annoyingly the 910k I ordered just arrived in the post this afternoon.......and they've sent me 910 ohm by mistake due to what appears to be bad labelling. Totally jinxed. Anyways I placed an order for a pack of 10m in the last hour which will arrive tmrw. Thanks for the help. 

[Beeblebrox]

Ok bit of an update. Thanks to all who have helped me over the last fortnight with my A8s. Much apprecaited.

 

Firstly the stock 65XE

 

I opened up the keyboard and the mylar is shot. I've investigated whether or not it's fixable and after some advice from the forum I've decided to by a replacement mylar at some stage. (Simius sells them).

I did the resistor mod here as well just in case but it's looking like it's definitely the keyboard mylar at fault here causing the function key issue. I temporarily borrowed the keyboard from my other XE to check. I played Rescue on Fractalus on cart and start and reset worked during the game.

 

In the meantime annoyingly I have noticed that whilst the RoF game cart works fine, I have no luck with my SIDE3 cart. I just get a grey screen when I tried to booting up in side loader menu mode. In SpartaDos mode I just get the blue screen with the square cursor and nothing happens. (Not text or command line).

 

So either there is a connection on the cart port that is needed by SIDE3 but not the ROF cart that needs cleaning or something else is up. I'll research cleaning the cart port contacts further first I think.

 

Next up, the 130XE with Ultramon and RS232

 

This one has worked out really well after a few hiccups - so again thanks for all your collective help. I did the 130XE vid fix mod here (page 32) earlier this eve, adding back the missing video circuit components and it booted up ok. Must admit I didn't really notice any difference, then I tweaked the colour pot as advised earlier and sorted the reds out. But I am glad I did the mod in the end as it can only improve the overall video signal. Plus it's given me some much needed soldering confidence. 

 

The 65XE (320K) Stereo A8 - my main Atari

 

I've discovered a graphics issue which isn't immediately aparrent until you see some demos running. So in this case I was viewing the NEAR and UNMEC demos by Agenda. (Amazing demos btw). Most of the segments in the demos run with no issues, but others consistently result in garbage screen outputs. So I am wondering if this is a sign of either bad RAM, or a faulty GTIA or ANTIC chip? If I run them on my other A8 they work fine and don't exhibit these issues.

 

So with the NEAR demo, the excellent subway graffitti parallax scrolling segment right at the end is partially corrupted. 

 

It should look link this:

 

 

but I get this:

 

 

Also slightly earlier on in the NEAR demo another segment which should look like this:

 

 

I get this:

 

 

The rest of the NEAR demo screens, including similar cube effects over city scapes all work fine. Go figure?!

 

In the UNMEC demo another segment should look like this:

 

 

yet I get this where you can still vaguely see the round sphere moving:

 

 

 

Ultimate1MB

 

I went ahead and purchased the U1MB upgrade with the intention of uprading the stock 65XE. This is because I'd be unable to upgrade the other two A8s above because they are already heavily modded. (The 130XE with Ultramon/RS232 utilizes the OS chip and the 65XE 320K stereo obviously has the 320k upgrade already, and now potentially a graphical glitch issue).

 

Until I can acertain if the stock 65XE is a viable machine to upgrade I am a little stuck. I guess I'll just have to get the SIDE3 cart working reliably on it.

 

 

 

[+Nezgar]

1 hour ago, Beeblebrox said:

The 65XE (320K) Stereo A8 - my main Atari

Just to confirm.... when you boot to BASIC and type BYE, a memory test should result in 40 blocks. When you boot with OPTION held down and no drive attached, it should also land in self-test, but memory test should result in 48 good blocks. If you still get 40 blocks, or the last 8 blocks are bad, the MMU might not actually be disabling the BASIC ROM.

 

The original RAM in XE's is not very good, and might be worth attempting to replace. Normally, in a 320XE the upgraded ram is all installed in the second column, so the original RAM should still be accessible, and not covered by other chips.

 

Another idea is if it's timing, would be to try swapping out the 74LS08 clock buffer for an HC08 or F08 - the extra RAM in the system might be putting extra load /delay on the pulses causing garbage reads on the bus in some conditions...

[Beeblebrox]

5 hours ago, Nezgar said:

Just to confirm.... when you boot to BASIC and type BYE, a memory test should result in 40 blocks. When you boot with OPTION held down and no drive attached, it should also land in self-test, but memory test should result in 48 good blocks. If you still get 40 blocks, or the last 8 blocks are bad, the MMU might not actually be disabling the BASIC ROM.

 

The original RAM in XE's is not very good, and might be worth attempting to replace. Normally, in a 320XE the upgraded ram is all installed in the second column, so the original RAM should still be accessible, and not covered by other chips.

 

Another idea is if it's timing, would be to try swapping out the 74LS08 clock buffer for an HC08 or F08 - the extra RAM in the system might be putting extra load /delay on the pulses causing garbage reads on the bus in some conditions...

@Nezgar Many thanks for the tips. I'll run the mem test (Bye) initially and see what happens. I do have spare 2 x XE boards which I could potentially salvage components from in the event I need to replace the old RAM chips. I'll do some research. So you don't think the graphical glitches are faulty GTIA or ANTIC, because it's working in displaying other graphics fine? thanks 

[Voxel]

If in stock mode (64k) the XE display is functioning well (with <64k programs) then your GTIA and ANTIC should both be fine, but if the demos are accessing 128k or more then you'll need to check the memory upgrade.  If so the images may be suggesting an issue with the bank switching (or bank switching modes used) in the demos, indicating the memory above the stock 64k is not being correctly accessed indicating a possible fault with the memory upgrade.  I'd check the wiring, may just be dry solder joints or a flaky connection.  Otherwise as suggested above you may need to swap out chips to check for the source of the fault.

[Beeblebrox]

1 hour ago, Voxel said:

If in stock mode (64k) the XE display is functioning well (with <64k programs) then your GTIA and ANTIC should both be fine, but if the demos are accessing 128k or more then you'll need to check the memory upgrade.  If so the images may be suggesting an issue with the bank switching (or bank switching modes used) in the demos, indicating the memory above the stock 64k is not being correctly accessed indicating a possible fault with the memory upgrade.  I'd check the wiring, may just be dry solder joints or a flaky connection.  Otherwise as suggested above you may need to swap out chips to check for the source of the fault.

@Voxel great, thanks for the insight. I'll have a look at the solder joints in those areas as part of my investigations. 

[Beeblebrox]

On 6/24/2021 at 12:12 AM, Nezgar said:

Just to confirm.... when you boot to BASIC and type BYE, a memory test should result in 40 blocks. When you boot with OPTION held down and no drive attached, it should also land in self-test, but memory test should result in 48 good blocks. If you still get 40 blocks, or the last 8 blocks are bad, the MMU might not actually be disabling the BASIC ROM.

 

The original RAM in XE's is not very good, and might be worth attempting to replace. Normally, in a 320XE the upgraded ram is all installed in the second column, so the original RAM should still be accessible, and not covered by other chips.

 

Another idea is if it's timing, would be to try swapping out the 74LS08 clock buffer for an HC08 or F08 - the extra RAM in the system might be putting extra load /delay on the pulses causing garbage reads on the bus in some conditions...

@Nezgar Hi. Ok, just ran the mem tests on the 65XE (320K) Stereo A8.

Booted to basic and got into the test by typing BYE. Got 40 blocks no errors. Rebooted holding down OPTION, got 48 blocks, no errors. 

 

Had another look at the Ram chips and sure enough the new chips are installed (in sockets) on the right, with the old chips on the left:

 

 

At this stage I prob won't consider replacing all the old RAM chips just yet. (Bit of a mammoth soldering job - although I do have access to some chips I could salvage from another board).

 

You mentioned replacing the 74LS08 clock buffer. Interestingly whomever did the upgrade sems to have replaced that already with an SN74HC14N if I am looking in the correct place:

 

 

 

Edited June 25, 2021 by Beeblebrox
typos

[Beeblebrox]

Ok  bit of an update re the Stock 65XE

 

So:

 

• Saturday afternoon with no interruptions - check
• Soldering iron powered up - check
• Cat locked in another room to save her from any soldering iron related mishaps (or me from any surprises that might cause me to melt my Atari!) - check.
• 1 x not inexpensive bit of 21st Century kit (U1MB) ready - check
• Newly acquired chip sockets - check
• Stock Atari 65XE patient on table - check.
• Borrowed working keyboard from my other A8 ready for transplanting - check
• Incontinence underwear donned - double check!!
• Mild headache from the 2nd AstraZeneca Covid jab I had 1st thing this morn - check! 

 

Desoldered the MMU and OS chips with relative ease, (which amazed me as I botched other chip removal whilst practising on a dead A8 board!!).

New 20pin and 28pin chip sockets soldered in place and the ribbon cables fitted in with no probs at all. (What's going on!?!)

Wired up the CPU (latter was a bit touch and go and at one stage I had bridged pins 35 and 36 with too much solder. Managed to get it all off although I really thought all that heat would have done the CPU some damage).

 

Mounted the U1MB above the monitor port and RF unit in an existing PCB hole. (I had to remove a plastic screw point from the casing with a knife to accomodate the head of the mounting bolt). Case actually fitted back on first time, (minus the top metal sheild plate).

 

 

After checking my none to shabby soldering work () and making sure nothing was bridged, etc, I hooked up the bare board and tentatively ..... powered it up.... it only bl**dy worked didn't it!!   - I was so prepared for an epic fail too!!

 

 

Managed to get into the U1MB Bios ( holding down HELP whilst powering on).

 

Despite cleaning the cart port contacts and checking the soldering, I still can't use the SIDE3 cart on this 65XE!!  (I still can't figure out what is wrong with the cart slot given it loads Rescue on Fratalus, Lode Runner, etc). The SIDE3 just boots to a very dark grey screen in menu mode and square cursor on blue screen in spartados mode). 

So instead I was using my SDrive Max. I loaded up a few games and demos to check it was running ok. Atari Blast! loaded in around 3-4mins, Agenda Cyberpunk demo in around 2.

 

Managed to flash FJC's latest June BIOS to the U1MB. (Highly recommend FJC's video here)

 

One major bonus I wasn't expecting is that in the bios after flashing the June firmware, there was a setting for "Hi speed OS" among others (SIO settings, etc). Next time I booted the SDrive Max it was super quick. Atari Blast loaded in less than 1.30min this time and Cyberpunk in less than 30seconds! Very happy as I didn't realise the SDrive Max would benefit from the U1MB upgrade in this way at all 

 

So given my 65XE (320k) Stereo A8 has graphics issues and I've borrowed the keyboard from my 130XE Ultramon/RS232 A8, I think I'll propmote my newly upgraded 65XE 1MB A8 as my main machine for now.

 

Gonna do a lot more research and look into some of the suggestions people have kindly made above to see if I can work out what the cartridge issue is. I really want to use my SIDE3 with U1MB on this machine.

 

Also gonna order a replacement mylar (and a spare), from Simius.

 

Today is a good day 

 

 

 

 

 

 

Edited June 26, 2021 by Beeblebrox
typos

[Beeblebrox]

Mini update.

 

65XE (the one with the new U1MB recently installed).

 

This eve I've been busy installing sockets for the GTIA, PIA, and POKEY chips. The GTIA in advance of the Sophia 2 I've decided to go for. The others cause I just thought - what the hell! (Good soldering practice). Thankfully all sockets and chips worked.

 

After a bit of digging on AA it seems my SIDE3 cart issues (prior to any work I've done incidentally) is possibly down to the following, where replacing the (U18) 74LS0 chip with a 74F08 chip might help:

 

 

The SIDE3 cart has no issues being recognised on my other XE A8s, so it appears this might be the issue.

 

I've purchased sockets and one of these so hopefully it'll work: https://www.ebay.co.uk/itm/Logic-IC-AND-Gate-Quad-2-Inputs-14-Pins-DIP-74F08-/133631709211

 

Looks like when it arrives I'll be removing the old 74LS0 chip, popping a 14pin socket in, installing the new 74F08 chip and soldering the factory wiring back as it was:

 

 

 

Fingers crossed it works. This is the only A8 I have I could install the U1MB into, so it will be a major PITA if I can't use the SIDE3 with it!

 

 

 

Edited June 30, 2021 by Beeblebrox
adding more info

[TGB1718]

Have you updated SIDE3 to the latest firmware, couldn't see if you have done that .

 

Currently Version 0.40 https://atari8.co.uk/wp-content/uploads/SIDE3_Loader_0.40.zip

[Beeblebrox]

34 minutes ago, TGB1718 said:

Have you updated SIDE3 to the latest firmware, couldn't see if you have done that .

 

Currently Version 0.40 https://atari8.co.uk/wp-content/uploads/SIDE3_Loader_0.40.zip

@TGB1718 Hi. Yes, that was one of the first things I did when FJC's new firmware came out last month. It flashed fine and SIDE3 has worked flawlessly with my other 2 x XEs but just not with this XE it seems.

 

Incidentally I tried it on it before I installed the U1MB btw - with nothing but a joystick plugged it. I just get a black/grey screen in loader mode and a blue screen with square cursor (no text or command prompt) in SpartadosX mode. No difference with the U1MB installed, same blank screen. Nothing. 

 

I am doing one of two things actualy - I have a USB powered 5V 3A cable coming so I'll try this first. This is based on the fact that the original Atari power brick I have been using to power all my XEs in turn is rated 5V 7.5va  (which equates to 1.5a). Because of some quirk with this particualr XE, at the lower end, 1.5a may not be powerfull enough to run this XE with SIDE3, let alone with U1MB installed too. (Or using SDrive Max when the SIDE3 is inplugged). It should do but from what I've read on AA not all XEs chipsets and quality are created equally. 

 

If this makes no difference I'll do the 74LS0 to 74F08 chip swap and factory rewire and see if I get any joy as Jimbaloid did.

 

thanks

Edited July 1, 2021 by Beeblebrox
typos

[TGB1718]

I am using an original Power Brick with my 130XE, U1M, SIDE3, S-Drive MAX and Printer Interface however I do have

an SIO splitter which I do power up separately, so peripherals are not loading the XE.

[Beeblebrox]

15 minutes ago, TGB1718 said:

I am using an original Power Brick with my 130XE, U1M, SIDE3, S-Drive MAX and Printer Interface however I do have

an SIO splitter which I do power up separately, so peripherals are not loading the XE.

It's a monster isn't it?   I could probably weight train with two of em!! 

 

I saw the SIO splitter for sale on Retrolemon. Tbh if I did dig out my old Atari peripherals, (if I could ever find my 1050 drives), I am not sure how much use I'd get outta them. I intend to use the SIDE3 as my main device given the speed, ATR/HDD functionality and the excellent FJC loader interface. (Occasionally using the SDrive Max). The 1050 would be for initially transferring boxes of my old 5 /14 still stacked up on the Attic from BITD.

I must admit I do find plugging and unplugging the SIO connectors a major PITA. Always worried I am gonna crack the board one day. Hmmmm.

 

Just outta interest do you have an old dot matrix printer hooked up or is the interface for a modern printer?

 

cheers

Edited July 1, 2021 by Beeblebrox
adding more info

[TGB1718]

2 minutes ago, Beeblebrox said:

Just outta interest do you have an old dot matrix printer hooked up or is the interface for a modern printer?

It's one of these SIO printer interfaces, I built an interface using an Arduino from the Centronics plug to my PC, so can print to

a text screen then do the actual printing from there.

[mdivancic]

1 hour ago, Beeblebrox said:

If this makes no difference I'll do the 74LS0 to 74F08 chip swap and factory rewire and see if I get any joy as Jimbaloid did.

My 800xl was completely unstable with Side2/3 before I did this. I’m guessing it will help.

[Beeblebrox]

It's alive!!!!!

 

Got the SIDE3 cart to work on my 65XE with the U1MB at last!!

 

How, you may well ask???  

 

I simply plugged in the new USB power cable that arrived earlier this afternoon!!  Seems it may have been the amps that were needed. The stock Atari brick I was using is V 7.5va (translates as 1.5a). This new USB to Din7 cable is 5V (3a).

 

Incidentally earlier on this afternoon (before I'd realised the new power cable was sitting in my postbox), I soldered a 14pin socket on the board in anticipation of the new 74F08 chip that is on it's way. I also salvaged another 74LS0 chip from a dead board I have and tried this in the socket just in case. Made no difference at the time - the SIDE3 still wasn't working.

 

So looks like it was the lack of juice going into the Atari that was causing the SIDE3 not to boot/be recognised!

 

When the new F4F08 chip arrives I may just pop it in the socket to try it - but at least now I know what the issue seemed to be. 

 

Time to get to grips with the new functionality that the combined U1MB and SIDE3 has to offer (must try not to get distracted booting up games (just loaded the Space Harrier CAR as a test)).

 

At some stage I'll revist the 130XE 320K Stereo A8 which is still having issues.

 

Thanks to all for any help provided so far - really appreciated.

Edited July 1, 2021 by Beeblebrox
adding more info

[Beeblebrox]

Here's one for ya. Following my post above from Thursday and several hours of gaming and flawless performance with the 65XE U1MB and SIDE3 last night, I powered it off and went to bed feeling fairly contented.

This morning I powered up the ol gal and was greeted with this:

 

 (And yeah, I know the jail bars are awful BTW - I am planning on getting Sophia 2 for this particular A8)

 

Bare in mind that:

 

It was working when I powered it off last night after playing several games and a good session of the Brundles.

 

I'd not touched, unplugged/plugged or modifed anything. (The 4 x gang mains power socket that powers everything is unplugged at night and has surge protection btw. Plus all it has plugged into it is the monitor and the Atari USB power cable).

I'd not installed or flashed anything to do with the U1MB either and again I loaded the Brundles and played it without any issues anyway.

The A8 itself hasn't been moved, dropped, treated like a frisbee, drop kicked or shot out of a cannon!!!

 

Given the graphical element to the glitch my immediate thought was that the GTIA chip was bad. So I popped a spare GTIA chip in. Same glitch.

I also tried replacing the PIA chip. No change.

 

(fyi the GTIA, PIA and 74LS0 are all socketed - and again things have been working well since I socketed them earlier in the week).

 

I then decided just in case, I'd unplug the U1MB and pop the original MMU and OS chips back in. Glitch present, a maroon screen! So it's not the U1MB causing this.

 

With the U1MB still unplugged I tried the Rescue in fractalus cart and again just a blank maroon screen.

 

Incidentally I can't replace the Antic chip as: 1) I don't have a spare chip 2) the chip is soldered to the board and up until now I haven't socketed it.

 

I checked the Retroscaler2x taking the signal to my monitor, and the DIN to S-video cable just in case and even plugged in my other A8 to double check which worked ok so it wasn't anything like that. (I wasn't expecting it to, tbh.)

 

Next I popped the 74LS0  chip in and out of the 14 pin socket. No change, same glitch.

 

I checked soldering, socketing chips (and their pins), and wiring from my earlier exploits - (both sides of the PCB) - all looked ok. (And besides it was working fine as I say before I went to bed for hours).

 

I am at a loss!

 

The replacement 74F08 chip is on it's way as I've planned to so that swap out with the 74LS0 anyways after many have advised it in various topics on AA inc this one.

 

So I guess my question would be:

 

Has anyone seen this glitch before or has any idea as to what would cause such a graphical glitch?

I am guessing it can't be the GTIA or PIA given I replaced both with spares.

 

Could the current (notorious) 74LS0 chip have finally failed or even cause this? (I'll know I guess as and when the 74F08 chip arrives.)

 

It's just odd that nothing has changed overnight to then see such a catastrophic failure.

 

For the mo I've left the U1MB unplugged and the original MMU and OS chips in place. 

 

Thanks in advance for any suggestions. Bit gutted as everything was working well and I'd planned some more nerdy A8 gaming sessions this weekend.  

 

 

 

Edited July 3, 2021 by Beeblebrox