1088XEL Atari ITX Motherboard DIY Builders Thread

[flashjazzcat]

Thanks Herb!

[flashjazzcat]

Unfortunately Eidolon appears to use an IRQ-driven custom SIO loader and will therefore never boot from an IDE-hosted ATR. Sorry!

So I guess this long-winded question is: what does the Hard Disk option in the PBI BIOS do in the system “guts” that would prevent certain titles from working when loaded from an ATR on the FAT partition of the IDE device, when that same ATR works through SIO?

Basically, the reason Eidolon doesn't work is the same reason behind most other failures. If a title uses its own SIO loader, it won't work. This was also commonly the case with anything that used older versions of XBIOS. Although custom SIO loaders were probably great at the time of the game's release, you can see how obnoxious they are now for anyone who wishes to host all their stuff on a hard disk. I would beseech anyone developing software nowadays not to use custom SIO loaders or rely on unresonably low MEMLOs. If you run out of memory, learn assembler and make your code half the size or figure out how to exploit extended memory, which is ubiquitous these days.

Michael mentioned page zero locations, and this is true to an extent, but the PBI BIOS does not rely on the few locations used (in the $3x area) to hold their values between SIO calls. However, if the program being loaded uses the same locations and depends on them holding their values between SIO calls, problems are guaranteed. The locations in question are intended for parallel device use according to the Atari docs, so any software which uses them - again - falls under the 'obnoxious, can't fix' category.

Edited February 3, 2019 by flashjazzcat

[ebiguy]

Hi,

 

I have another (small) issue.

Usually, you have to temporarily connect the first 2 pins at J3 to power on the 1088 XEL.

My issue is that, as soon as I plug the power jack, the 1088 XEL boots without having to connect the first 2 pins.

The left LED at D5 (labelled +5VSB) is on but the right LED at D4 (labelled ON) stays off.

When booted, everything is working fine (U1MB, VBXE...)

After booting, if I temporarily connect the first 2 pins at J3, the right LED at D4 is now ON and the screen is slightly brighter (this luminosity seems the right one).

If I temporarily connect again the first 2 pins to turn the 1088 XEL off, the the right LED at D4 is now OFF and the screen gets slightly darker (still very visible and usable) but the 1088 XEL is still running.

The only way to turn the 1088 XEL off is to unplug the power jack.

I guess the 555 at U4 is working but, anyway, I tried with another one but this did not change anything.

I think this behaviour on my board is new but I am not really sure.

 

Any idea about what's going on ?

[+DrVenkman]

Unfortunately Eidolon appears to use an IRQ-driven custom SIO loader and will therefore never boot from an IDE-hosted ATR. Sorry!

 

Thanks for looking into it. Your explanation was fascinating and helpful to understand what's happening. I don't (currently) have my little RespeQt server connected (a dedicated RPiZero W running RespeQt full time), and my SDrive-Max doesn't want to play very nice with other stuff on the SIO chain until I get around to building once of those little isolation boards MrRobot has designed, so it's been setup as the drive device for my 1200XL instead.

 

But as it happens, it was kind of fun to write out a sector copy of the crack to physical media and then use that to boot the game. Reminded me of old times at my local "crack house", lol. There was a guy we hung around with, mid-30's, probably, married but no kids, much older than most of his teenager computer pals of the day. He had a dedicated phone line for his BBS and hosted TONS of warez on his members-only section. He would give us copies of most anything we wanted if we gave him two disks for every one he gave us. So last night kinda reminded me of all that 30-odd years ago.

[flashjazzcat]

I did nothing more sophisticated than cracking AtariWriter Plus and AMAC to run from a 360K DSDD floppy, but they were fun times indeed. Although the scope of the projects one can manage today has become exponentially larger owing to cross-assembly tools, etc, I have some nostalgia for simpler times.

 

An interesting patch would be replacing the Eidolon loader with some code which simply reads the requisite sectors into RAM using standard OS calls. That would fix the problem completely. There's just a little stub of code which waits for a flag at $80 to be set by the IRQ driven loader, at which point a JMP presumably occurs to the start of the game proper.

[+DrVenkman]

I did nothing more sophisticated than cracking AtariWriter Plus and AMAC to run from a 360K DSDD floppy, but they were fun times indeed. Although the scope of the projects one can manage today has become exponentially larger owing to cross-assembly tools, etc, I have some nostalgia for simpler times.

 

An interesting patch would be replacing the Eidolon loader with some code which simply reads the requisite sectors into RAM using standard OS calls. That would fix the problem completely. There's just a little stub of code which waits for a flag at $80 to be set by the IRQ driven loader, at which point a JMP presumably occurs to the start of the game proper.

 

I wonder if I post this to DjayBee's thread if he or someone else will give that a shot … I would be great if all of these old Lucasfilm/Lucasarts titles could be run off the SIDEloader - I didn't try Ballblazer but Rescue on Fractalus works, which is pretty great.

[flashjazzcat]

Bootable titles are a little more limiting since they obviously need to be launched via an ATR, but this would still be a big improvement for anything which can't be converted into an XEX.

[+MacRorie]

Hi,

 

I have another (small) issue.

Usually, you have to temporarily connect the first 2 pins at J3 to power on the 1088 XEL.

My issue is that, as soon as I plug the power jack, the 1088 XEL boots without having to connect the first 2 pins.

The left LED at D5 (labelled +5VSB) is on but the right LED at D4 (labelled ON) stays off.

When booted, everything is working fine (U1MB, VBXE...)

After booting, if I temporarily connect the first 2 pins at J3, the right LED at D4 is now ON and the screen is slightly brighter (this luminosity seems the right one).

If I temporarily connect again the first 2 pins to turn the 1088 XEL off, the the right LED at D4 is now OFF and the screen gets slightly darker (still very visible and usable) but the 1088 XEL is still running.

The only way to turn the 1088 XEL off is to unplug the power jack.

I guess the 555 at U4 is working but, anyway, I tried with another one but this did not change anything.

I think this behaviour on my board is new but I am not really sure.

 

Any idea about what's going on ?

 

 

 

I think you have a short somewhere in the power on pins. There is probably a more technical way to say that, but here's what I think is going on: there is a bridge on either the pins at J3 that is enough to send power to the machine, but not enough power to fully light the +5 LED. I would look at the underside of the board and check all of the solder points at that jumper. Failing that, perhaps one of the pins at U4 is grounded? Would check that area as well.

 

Perhaps someone else has a thought on this, but that is my guess.

[ebiguy]

I think you have a short somewhere in the power on pins. There is probably a more technical way to say that, but here's what I think is going on: there is a bridge on either the pins at J3 that is enough to send power to the machine, but not enough power to fully light the +5 LED. I would look at the underside of the board and check all of the solder points at that jumper. Failing that, perhaps one of the pins at U4 is grounded? Would check that area as well.

 

Perhaps someone else has a thought on this, but that is my guess.

Thanbk you for your answer.

 

I tripple checked the solder and don't see anything bad... Maybe someone can look at the pictures and find something I missed.

The only thing is that I inverted the green and the yellow LEDs. But this does not really matter.

 

 

 

 

 

[AtariGeezer]

The Power Jack needs more solder on the bottom...

[ebiguy]

The Power Jack needs more solder on the bottom...

Thank you for looking at my pictures.

I just added more solder at the back on the 3 big pins and still the same issue.

[AtariGeezer]

Look over the bottom of the board for connections not soldered. It looks like there are a few with no to very little solder...

[+Stephen]

I would also recommend very thoroughly cleaning off all that flux residue. I've had some strange issues in the past from that.

[ebiguy]

I would also recommend very thoroughly cleaning off all that flux residue. I've had some strange issues in the past from that.

Do you have an advice on how to do that ?

[+mytek]

Ebiguy barring any problems with flux or something not properly soldered, it could be a component mismatch or failure. For instance Q2 is suppose to be a NPN transistor (PN2222), and Q1 is supposed to be a P-channel Mosfet (NDP6020P). If the NPN transistor were accidentally swapped with a PNP type (PN2907), this could explain what you are seeing. It would be very easy to make that mistake since both types are used on the board, and they look nearly identical.

 

And if that solder does not use a No-Clean flux, then as has been suggested, leaving it on could be a possible problem, although I doubt that it would be limited to the ON-OFF power circuit only (thinking frequency shift in the oscillator sections). But that power latching circuit is sensitive, and can respond to finger touch alone, so I wouldn't rule out the flux causing a bridging effect as MacRorie suggested. Normal operation with those pins bridged would be to turn on each time power is applied.

 

Flux can be cleaned off with alcohol and a Q-Tip or toothbrush (please don't use those on yourself afterwards ).

 

Edit: BTW, all the components I can identify in your pictures look to be correct .

[+Stephen]

Do you have an advice on how to do that ?

Depending on the flux used, either isopropyl (rubbing) alcohol, acetone, or near boiling water and dawn dish-washing liquid.

[+mytek]

Depending on the flux used, either isopropyl (rubbing) alcohol, acetone, or near boiling water and dawn dish-washing liquid.

 

Very true. Rosin flux usually responds well with 50/50 ratio of alcohol and acetone, whereas if it's a water soluble type, the water and dish washing soap followed by a clean water rinse would be the ticket as you suggested.

[_The Doctor__]

555 is in not properly seated in it's socket, it appears to be tilted, the pins might even be bent at this point... straighten and re seat... check transistors.. a few components look starved for solder. This board needs a thorough clean up.

Edited February 5, 2019 by _The Doctor__

[ebiguy]

555 is in not properly seated in it's socket, it appears to be tilted, the pins might even be bent at this point... straighten and re seat... check transistors..

I have reseated the 555 and same result.

 

a few components look starved for solder. This board needs a thorough clean up.

About starving for solder, do you see that by looking at the component side or the solder side ?

When I look at the solder side there is almost too much solder. But sometimes, the solder does not come out on the component side and the component seems not soldered.

But I guess this is because I did not put flux on the board before starting to solder. I thought this was not useful as my solder already has flux inside.

 

As I have not enough tools (and knowledge) to check every component, I am thinking about a very easy test to do:

Would it be interesting to remove the 555 and plug the power ? Should I get a board not booting because no power comes to the Atari components ?

I just did the test and my board IS booting without it. Is it the expected behaviour ?

[+mytek]

Absolutely should not boot without the 555 timer. This is really starting to sound like a wrong or damaged component, specifically Q1 or Q2. Did you verify that Q2 is a PN2222?

 

Or some sort of bridging across Q1 which is the electronic power switch.

[ebiguy]

I can not see what Q2 is without desoldering it.

Even if I bend it, I can not read what is written on its back.

What I understand so far is that I probably have to rework this little corner of the board.

Despite Q1 and Q2, do you think I could have a bad capacitor in the same area ?

I am asking because I am thinking about ordering again the few components (Q1, Q2 and all Cxx) of this area to start over after desoldering the old ones.

If I order components, I would like to get all of them at once

[+mytek]

Probably unlikely that a capacitor would have any bearing on this.

 

If you are going to order new parts, go ahead and cut the leads on Q2 the presumed NPN transistor so that you can better read it's markings. This also has merit, in that doing so eliminates what I will call the pre-switch for Q1. So after cutting it out, apply power once again with the 555 timer removed and see if the system stays OFF. If it does, there is a good chance that Q2 was either bad or the wrong part. If the system still comes on without the 555, then remove Q1 and see what happens. If it still comes on without Q1, then there is a serious bridge somewhere (hopefully not in one of the inner layers of the PCB).

[Sleepy]

Hi,

 

I´m going to fit my 1088XEL with VBXE. It´s fitted with sophia DVI already and works fine. Installing the VBXE raises a question:

 

Is the pinout on the pcb V1.1 still wrong? I found this correction:

 

Sleepy

Edited February 10, 2019 by Sleepy

[+mytek]

I´m going to fit my 1088XEL with VBXE. It´s fitted with sophia DVI already and works fine. Installing the VBXE raises a question:

 

Is the pinout on the pcb V1.1 still wrong? I found this correction:

pinout.png

 

 

 

That's incorrect.

 

The problem with development threads is that bad or no longer pertinent information still remains. For all of the latest and greatest up to date info, please visit the 1088XEL page on my website. You'll see a complete top side view of the silk screened side of the board. It's labels are correct. Also somewhere on here there were some good tips on installing a VBXE.

[Sleepy]

Thx. Of course, I know your website. Therefrom I got a lot of very helpful informations. :-)

 

Sometimes, finding too much information causes confusion...

 

Sleepy