1088XEL Atari ITX Motherboard DIY Builders Thread

[+DrVenkman]

 

The female 0.1" pins should be soldered in for the UAV and the X1 sockets. (There maybe one other place, but I am not in front of it right now)

Thanks for the feedback. A bit of clarification if I might - if I intend an NTSC-only system, should I merely solder in X1? Or is there a reason I should socket it with female header pins?

 

And as for the UAV, the BOM includes on machine-pin socket for the equivalent of the 4050 socket used on standard A8 systems. I have not yet soldered that in. I figured I would fit that into place along with the rest of the pins and headers on the main board and the UAV, then solder things together while everything is sitting in place to ensure proper alignment, much like the instructions for the U1MB and Sparkfun FTDI board. Is that the proper approach or is there enough play that I can solder in that machine pin socket first?

[+mytek]

Thanks for the feedback. A bit of clarification if I might - if I intend an NTSC-only system, should I merely solder in X1? Or is there a reason I should socket it with female header pins?

And as for the UAV, the BOM includes on machine-pin socket for the equivalent of the 4050 socket used on standard A8 systems. I have not yet soldered that in. I figured I would fit that into place along with the rest of the pins and headers on the main board and the UAV, then solder things together while everything is sitting in place to ensure proper alignment, much like the instructions for the U1MB and Sparkfun FTDI board. Is that the proper approach or is there enough play that I can solder in that machine pin socket first?

You could solder in X1, and this would be recommended if you really have no intention or need to go PAL in the future. Soldering is always more reliable than a socket in this case.

 

The UAV uses both a 16 pin DIP and a 6 pin SIP machine pin sockets, with the later giving the option of cutting off a piece from a DIP socket if necessary. And since a SIP socket is not self aligning (easy to install tilted to either side), it's best to plug the UAV into the sockets first, insert that into the motherboard as a complete assembly, and solder it in.

 

On the SparkFun board, it does not get socketed, and is instead soldered directly in place using two 9 pin pieces of break-away male header pins. I would also recommend doing that as a complete assembly as well.

[+MacRorie]

Thanks for the feedback. A bit of clarification if I might - if I intend an NTSC-only system, should I merely solder in X1? Or is there a reason I should socket it with female header pins?

 

And as for the UAV, the BOM includes on machine-pin socket for the equivalent of the 4050 socket used on standard A8 systems. I have not yet soldered that in. I figured I would fit that into place along with the rest of the pins and headers on the main board and the UAV, then solder things together while everything is sitting in place to ensure proper alignment, much like the instructions for the U1MB and Sparkfun FTDI board. Is that the proper approach or is there enough play that I can solder in that machine pin socket first?

 

 

My experience is that the UAV has enough play to allow you to solder the pins in first. However, YMMV.

 

As Michael notes, you can solder it in, but that means only an NTSC system. Granted, you may never change that, but if you are going to install the rest of the items in to allow for the possibility of a PAL system, why not leave this bit as a socket for the future? Also, any future changes/upgrades/additions will presume that X1 is socketed.

 

At the end of the day, it is your system!

[+Firedawg]

Good day today with the 1088XEL flashing U1MB with FJC's to the latest beta firmware. I spent a little time mulling around with the 1088XEL to gain some competence in its usage . I've been finishing up on the some of the XEL-CF-][, DB9 cables for the case (which should be arriving soon), Mouse Selects, and XEL Status Panels. So, still a lot going on around the 1088XEL!!!

 

Question: In one of my U1MBs during flashing I saw some garbage text. How can I remove the garbage text from the BASIC Slots 2: on down? It shows up in the BIOS screen under each option (not shown).

 

[flashjazzcat]

Garbage was probably caused by slot descriptions originally being in unexpected locations (I still didn't update UFlash to cope with the very oldest stock BIOS revisions, if indeed that's what was on the chip to begin with). Easily fixed, anyway. Just hit 'N' on each slot, backspace over the crap, and edit the descriptions to suit yourself. CTRL-F to flash them when you're finished.

[+Firedawg]

Garbage was probably caused by slot descriptions originally being in unexpected locations (I still didn't update UFlash to cope with the very oldest stock BIOS revisions, if indeed that's what was on the chip to begin with). Easily fixed, anyway. Just hit 'N' on each slot, backspace over the crap, and edit the descriptions to suit yourself. CTRL-F to flash them when you're finished.

 

Thanks Jon! I'm not sure how it got there but I'm sure I had something to do with it;-)

[flashjazzcat]

Thanks Jon! I'm not sure how it got there but I'm sure I had something to do with it;-)

 

Heh... The slot descriptions were quite itinerant until they settled on a home somewhere in the middle of the code in the second or third revision. I put them in the data area at the very top of the ROM (they have to be relocated during upgrades), and they're in exactly the same place in both the U1MB BIOS and the Incognito BIOS. If only life were always that simple.

[+DrVenkman]

You could solder in X1, and this would be recommended if you really have no intention or need to go PAL in the future. Soldering is always more reliable than a socket in this case.

 

The UAV uses both a 16 pin DIP and a 6 pin SIP machine pin sockets, with the later giving the option of cutting off a piece from a DIP socket if necessary. And since a SIP socket is not self aligning (easy to install titled to either side), it's best to plug the UAV into the sockets first, insert that into the motherboard as a complete assembly, and solder it in.

 

On the SparkFun board, it does not get socketed, and is instead soldered directly in place using two 9 pin pieces of break-away male header pins. I would also recommend doing that as a complete assembly as well.

 

Yep, I remembered as much about the Sparkfun board. No worries there. Just trying to figure out if I should go ahead and solder that UAV socket it or wait for the arrival of the UAV; I will wait. I will probably do the U1MB modification and mod up the tall headers in the next few days, then install the what I can while I wait for the UAV and the new video jack.

 

Another quick question re crystals - since I've already installed the 74xxxx logic chip socket and discrete components you ID'd in the image earlier, I'll probably go ahead and just install both crystals. So in that case, the idea is to socket X1 to replace with a PAL clock crystal if desired, but solder X2 since the PAL colorburst crystal will only be used in a PAL configuration and never need to be replaced. Is that correct? I'm not familiar with PAL systems at all, having never owned one nor disassembled or worked on one.

[+mytek]

 

Yep, I remembered as much about the Sparkfun board. No worries there. Just trying to figure out if I should go ahead and solder that UAV socket it or wait for the arrival of the UAV; I will wait. I will probably do the U1MB modification and mod up the tall headers in the next few days, then install the what I can while I wait for the UAV and the new video jack.

 

Another quick question re crystals - since I've already installed the 74xxxx logic chip socket and discrete components you ID'd in the image earlier, I'll probably go ahead and just install both crystals. So in that case, the idea is to socket X1 to replace with a PAL clock crystal if desired, but solder X2 since the PAL colorburst crystal will only be used in a PAL configuration and never need to be replaced. Is that correct? I'm not familiar with PAL systems at all, having never owned one nor disassembled or worked on one.

 

Yes to everything you said. All PAL components including X2 can be permanently left in place, even if you choose to set this up as NTSC. So in that way you can change to PAL by changing X1, Antic, and GTIA. The NTSC/PAL jumper is for the V-Gate chip, and simply matches it's color switching to the video mode in use.

[+DrVenkman]

I had a few only a little while this morning, so rather than fire up the iron for soldering work I decided on the brutal part-surgery I'd been worried about.

 

Success! Thanks for the tips, Michael!

 

 

I haven't yet soldered them in. I'm going to tackle removing the U1MB 90 degree headers this evening or tomorrow, I think.

[+LarryL]

Tried to add RGB to my 1088XEL by means of SOPHIA.

Board received yesterday so I installed it today.

 

I had to add an additional 40pin socket to increase clearance

 

but now it looks fine:

It tried it with a DIN13 to SCART cable which worked fine with my Atari ST (not the mono variant, but the RGB variant)

But now I do not get a stable picture.

On a standard LCD there is simply no picture (cannot sync), on a Atari SC1425 (SCART input) there is a picture, but not synchronized - it is "running through the screen from bottom to top"

I tried to upload a video, but I get an Error "You aren't permitted to upload this kind of file"

Can it be the the HSYNC/VSYNC signal is not properly working?

What I've checked is, that the pins 9/12 from the DIN13 (J9) are going through to the J27/CSYNC on the board

any ideas?

cheers

Michael

 

[+mytek]

Tried to add RGB to my 1088XEL by means of SOPHIA.

Board received yesterday so I installed it today.

 

I had to add an additional 40pin socket to increase clearance

IMG_0229.jpg

 

but now it looks fine:

IMG_0230.jpg

It tried it with a DIN13 to SCART cable which worked fine with my Atari ST (not the mono variant, but the RGB variant)

But now I do not get a stable picture.

On a standard LCD there is simply no picture (cannot sync), on a Atari SC1425 (SCART input) there is a picture, but not synchronized - it is "running through the screen from bottom to top"

I tried to upload a video, but I get an Error "You aren't permitted to upload this kind of file"

Can it be the the HSYNC/VSYNC signal is not properly working?

What I've checked is, that the pins 9/12 from the DIN13 (J9) are going through to the J27/CSYNC on the board

any ideas?

cheers

Michael

 

 

Thanks for the reminder on looking into this. Yep I saw the same situation when trying to use the GoldStar variant of the SC1224, whereas the JVC works just fine (unfortunately this just recently became apparent). Bob Woolley and I talked about what might be happening, and I think it relates to trying to use CSYNC for both the V and H sync inputs on the monitor (that's how it's presently configured on the DIN-13 output of the 1088XEL). Anyway the thought was that having the HSYNC pulses present at the VSYNC input might be causing a bit of confusion. Two possibilities to fix this situation might be to first simply disconnect the CSYNC signal from the VSYNC output and see if that fixes the problem (some monitors are fine with only having the CSYNC going into HSYNC and nothing more. The other possible solution, is to insert a low-pass filter into the connection between CSYNC and the VSYNC output so as to strip out the HSYNC pulses. The 2nd solution is obviously more attractive, since it should still allow the JVC monitor to function as well. I'll test out these ideas tomorrow and get back to you.

 

Luckiliy the CSYNC to VSYNC trace connection is on the bottom side of the 1088XEL motherboard, so intercepting it will be easy, only requiring a trace cut.

[+mytek]

I had a few only a little while this morning, so rather than fire up the iron for soldering work I decided on the brutal part-surgery I'd been worried about.

 

Success! Thanks for the tips, Michael!

 

IMG_9479.JPG

 

I haven't yet soldered them in. I'm going to tackle removing the U1MB 90 degree headers this evening or tomorrow, I think.

 

Just a reminder. Don't solder those in without first plugging in the U1MB .

[+DrVenkman]

 

Just a reminder. Don't solder those in without first plugging in the U1MB .

 

Yep, I remembered. But I gotta remove those pesky 90 degree headers first. That's for later tonight or tomorrow, I think.

​This afternoon is for a visit to a new (to me) craft brewery taproom here in Nashville with my son. Meeting a long-ago school friend and some others to share a pint or two and enjoy some winter comradeship.

[+LarryL]

 

Yep, I remembered. But I gotta remove those pesky 90 degree headers first. That's for later tonight or tomorrow, I think.

​This afternoon is for a visit to a new (to me) craft brewery taproom here in Nashville with my son. Meeting a long-ago school friend and some others to share a pint or two and enjoy some winter comradeship.

Be careful while desoldering - I nearly spoiled the U1MB by sucking out too much from the board.

 

„Winter comradeship“ sounds pretty cool - I wish we would have some winter here in Germany. Still +8 degrees Celsius

 

Have fun!

[+DrVenkman]

Be careful while desoldering - I nearly spoiled the U1MB by sucking out too much from the board.

 

„Winter comradeship“ sounds pretty cool - I wish we would have some winter here in Germany. Still +8 degrees Celsius

 

Have fun!

 

Thanks, I'll be careful.

 

As for winter, it's been well below freezing all week. It as 6°F (-14.4°C) a few days ago. That's kind of ridiculous cold for this part of the world. It won't get above freezing again until sometime tomorrow, I gather.

[+mytek]

LarryL can you please try this mod to see if it fixes your problem.

 

 

I tested on both a JVC and a GoldStar SC1224 and it worked in either case.

 

- Michael

[+LarryL]

how important is exact capacity?

Closest I have is 22 nF type 223

[+MacRorie]

 

Yep, I remembered. But I gotta remove those pesky 90 degree headers first. That's for later tonight or tomorrow, I think.

​This afternoon is for a visit to a new (to me) craft brewery taproom here in Nashville with my son. Meeting a long-ago school friend and some others to share a pint or two and enjoy some winter comradeship.

 

 

Remember: Clip them off (carefully) and then attempt to desolder them.

[+LarryL]

tried it with the 22 nF - result even worse.

Now picture is also moving from left to right (or vice vera - couldn't figure out)

Even without 2k2 and cap, only trace cut, same result.

 

Maybe it is my cable?

 

I checked pinout and found, that pin 9 (HSYNC) of DIN13 is not connected to SCART cable at all

Pin 12 (VSYNC) is connected to pin 16 of SCART. Pin 16 called RGB mode - which seems a bit strange to me

Pin 2 (composite video / TTL CSync on ST) is connected to SCART pin 20 (Video sync input)

[+mytek]

how important is exact capacity?

Closest I have is 22 nF type 223

It really needs to be what I showed in order for this to properly filter out the HSYNC pulses, and just pass the VSYNC. Since this is an unbuffered fix, having too much capacitance can also begin altering the HSYNC pulses to the point where you'll lose horizontal sync.

 

However with that said, to get the same time constant with a .022uf (22NF) capacitor you will need to change the resistor to 1k. Please give that a try if you have that resistor value on hand. No guarantee, but if it works for you I will test on my end as well.

[+mytek]

tried it with the 22 nF - result even worse.

Now picture is also moving from left to right (or vice vera - couldn't figure out)

Even without 2k2 and cap, only trace cut, same result.

 

Maybe it is my cable?

 

I checked pinout and found, that pin 9 (HSYNC) of DIN13 is not connected to SCART cable at all

Pin 12 (VSYNC) is connected to pin 16 of SCART. Pin 16 called RGB mode - which seems a bit strange to me

Pin 2 (composite video / TTL CSync on ST) is connected to SCART pin 20 (Video sync input)

Wow so now comes the real problem, and that is your cable was meant to hook up to an actual ST computer. Whereas the DIN-13 jack on the 1088XEL was aimed at connecting to an ST monitor. May sound like the same thing, but it's not. The ST computer provided some alternative signals not even used by the SC1224 such as the TTL CSYNC that you mentioned.

 

Since your ST to SCART cable is looking for CSYNC, you could run a jumper from pin 9 to pin 2 to see if that satisfies it. Edit: this would probably be a good addition to a future 1088XEL version board, as well as filtering out the HSYNC pulses going to the VSYNC pin 12. I never even thought about the possibility of an off the shelf ST to SCART cable, but if we can get this to work that would certainly open up the interface capabilities.

[+LarryL]

Wow so now comes the real problem, and that is your cable was meant to hook up to an actual ST computer. Whereas the DIN-13 jack on the 1088XEL was aimed at connecting to an ST monitor. May sound like the same thing, but it's not. The ST computer provided some alternative signals not even used by the SC1224 such as the TTL CSYNC that you mentioned.

 

Since your ST to SCART cable is looking for CSYNC, you could run a jumper from pin 9 to pin 2 to see if that satisfies it.

did not make any change

 

It really needs to be what I showed in order for this to properly filter out the HSYNC pulses, and just pass the VSYNC. Since this is an unbuffered fix, having too much capacitance can also begin altering the HSYNC pulses to the point where you'll lose horizontal sync.

 

However with that said, to get the same time constant with a .022uf (22NF) capacitor you will need to change the resistor to 1k. Please give that a try if you have that resistor value on hand. No guarantee, but if it works for you I will test on my end as well.

 

resistors I have enough - unfortunately no change

 

In order to get rid of the uncertainty of what is what in my SCART cable, I intent to build a fresh cable from DIN13 to the DIN8 of my SC1425 according to following picture

 

but not today - maybe I find some time during the week..

 

thanks for your effort so far and have a nice sunday

[+mytek]

LarryL sorry that didn't work for you.

 

I also noticed that there is a SCART example listed for the VBXE that might be a possibility, since that one only requires CSYNC and the RGB signals. I think it's in the VBXE manual on Lotharek's site. Since those same signals are available on the DIN-13 jack of the 1088XEL (pin 9 being Buffered CSYNC) it should be possible to wire up something that works.

 

Good luck and let me know what you find out.

 

BTW, it's probably best to leave the DB9 connector attached to the original Sophia video output cable, and just crimp on the IDC-10 so that you have two different possible ways to interface Sophia. If I recall correctly that would give you Commodore 1084D RGB monitor compatibility out of the box, which might be easier to adapt to the SC1425.

 

My intentions for the included DIN-13 on the 1088XEL was always aimed at interfacing to the SC1224 series Atari ST series monitors, which apparently I missed the ball with getting the VSYNC aspect correct for all versions. Anyway the resistor/capacitor mode will fix that oversight when using something other than the JVC version. I did include the added possibility of having stereo audio through that connector, which was not originally supported in the SC1224's. However when it comes to other monitors it really is best to utilize Sophia's provided DB9 connector since it has all of the correct sync and RGB signals at their proper voltage levels (.75v). The SC1224 series requires 1v RGB and TTL level sync for the best results, and with Sophia that means making a mod to increase the RGB output voltage to the required 1v level.

[+LarryL]

BTW, it's probably best to leave the DB9 connector attached to the original Sophia video output cable, and just crimp on the IDC-10 so that you have two different possible ways to interface Sophia. If I recall correctly that would give you Commodore 1084D RGB monitor compatibility out of the box, which might be easier to adapt to the SC1425.

The Sophia came from Simius with the small cable to the XEL-RGB-through connector. Maybe because I mentioned that it is for the XEL.

 

But I will also change this to the VGA9 connector

 

I will report results next weekend