1088XEL Atari ITX Motherboard DIY Builders Thread

[+mytek]

Nah, I still wouldnt have got it.

 

My undergraduate degree is aerospace engineering; we had two required semesters of EE-for-Dummies (EE301 and 302 or something) and that was it. As a friend of mine said, as we were baffled by some obscure concept after yet another obtuse lecture by the bored and frustrated junior professor, Sparks and magic, man. Sparks and magic. Thats all this stuff is.

 

That same year we were taking thermodynamics of jet and rocket engines, aerospace structural design and analysis, control systems design and theory, numerical methods for engineers, linear algebra ... you know, stuff non-wizards can understand!

That's quite an educational background you have .

 

My credentials on paper are far less impressive... never graduated from High School (due to flunking Government because it was sooo incredibly boring), never went to college (for obvious reasons), and my first 3 jobs after High School were working for 3 different gas stations around my home town. But due to a fortunate circumstance, I got hired and trained to build auto-cascade refrigeration systems by a start-up company in a town 30 minutes away (year 1977). Went on to learn every detail about how these systems worked, got promoted into testing, and then eventually I became the R&D manager working on and creating new designs for systems able to achieve temperatures down to -160 C for what's called water vapor cryo-pumping (used as the main pump in vacuum coating systems, a process to create optics, CD's, DVD's, semiconductor devices, and Integrated circuits). I now run my own business as a consultant/designer, creating similar refrigeration system designs for companies that do electronic environmental testing at extreme temperatures. In between all of that I also pursued a hobby in electronics, where I was self taught. And although I dabble a bit in computers, and do a small amount of programming, it is not my strong suit at all. But I get by. Luckily for me, I am very fortunate to befriend people that make up for my deficiencies in other areas, so as a whole, some really cool stuff can be built like the 1088XEL. FlashJazzCat is just one such example of the kind of people I've had the pleasure to work with that truly put the icing on the cake when it came to this project.

[+DrVenkman]

Education is great but most of the best education comes from life, not school. As you’ve obviously shown! My mom never graduated high school, but eventually got her GED and then went to nursing school in middle age. My step-dad was a high school graduate who eventually became a non-degreed mechanical design engineer working for a contractor in Oak Ridge at the Y-12 nuclear weapons facility. It was fun whenever the U.S. Marshals would come by every few years for interviews as part of his Q clearance process. But he never had a college degree in any of it; he’d been a natural artists and tinkerer/builder and knew intuitively how stuff would go together.

 

But all that being said, I remember V=I*R and a few bits like that from going on 30 years ago now, but much more than that has been lost in the mists of my brain.

[flashjazzcat]

Are there specific hooks built in to the new firmware and how many rom slots will the JAZZ GUI take up?

The new firmware (since June 2015) has included support for 1KB plugin modules, and there are hooks for writing to hardware, adding menu items, reacting to menu edits, etc. This approach was devised in order to avoid hard-coding the functionality of the four general purpose IO signal headers, and with the 1088XEL it enabled easy reassignment of the pin previously assigned to Covox, as well as new functionality on signals S0 and S1. I still had to customise the main BIOS code for the 1088XEL, however, since we wanted a special splash-screen and some unique behaviours on cold power-up.

 

As for the GOS: It requires 128K, or 16 ROM banks. This intrudes on the last 64KB originally allocated for SDX; hence provision is made for shrinking the SDX slot and flashing the 192KB ROMs which are now available. For those with no interest in or need for a WIP GOS, one can also enlarge the SDX slot to 320KB, providing ample room for more stuff on CAR:.

[Spaced Cowboy]

[i deleted this. I'm not sure it was on-topic]

Edited January 23, 2018 by Spaced Cowboy

[+Firedawg]

Still waiting on my case which is on the backside of its estimated delivery range, so I'm expecting it any day. I've got several XEL-CF-II completed and just waiting on my cables. Very much enjoying the experience with using this computer.

 

 

[+DrVenkman]

USPS still tells me I'll be getting the last parts for my kit tomorrow, along with parts to build an XEL-CF. I received replacement LEDs today to undo by fubar from early in the build - thanks, Michael. I was originally not going to not bother heating up the iron and working on such a small task tonight but after sitting down for awhile and eating supper, I got antsy.

 

Yellow "+5V" light works.

 

 

And the green "On" light works.

 

 

Whew.

 

[+DrVenkman]

I've been a busy boy today! I received the parts to finish my board and get it into a boot-able state. Hurray! Unfortunately, it initially wouldn't boot. Boo.

 

Well, I wasn't getting video anyway. I'd get a quick flash on the screen, the keyboard lights would flash, the NumLock key worked ... So I disconnected, reflowed some potentially-dodgy solder connections and tried again. Voila! Whew.

 

 

 

I do have one issue that hopefully Michael can shed some light on and offer some tips: I can't seem to get into the Ultimate 1MB setup screen. If I press and hold F12, I can see a flash of the setup menu screen but it's gone in an instant and I drop back to BASIC. Remember, this is the U1MB is stole from my 800XL - before I grabbed it to replace the one I borked up on the header pins, I'd been running Jon's latest released firmware from December 2016.

 

 

Any ideas what's going on and why I can't seem to get to the U1MB Setup screen without dropping out?

 

 

[+Stephen]

One at a time, in 2018, new Atari's are birthed. How insane is this? So glad to be a small part of this.

[+mytek]

First of all congrats on your 1088XEL booting up and displaying video

 

I'm not sitting in front of one of my systems to check this out, but the idea is to simply press and release the F12 key all by itself (don't hold it) and that should take you into Setup. In some of Jon's latest 1088XEL U1MB beta firmware releases you can also press F10 (HELP) during boot-up when you see the prompt to do so, and that should also take you to the setup screen.

 

Edit: without being able to try it for real, I'm not sure what pressing and holding F12 will do. Maybe exactly what you are describing .

[+DrVenkman]

I'm sorry, I think I was imprecise or incorrect in my initial post. Pressing F12 simply seems to reset the machine. Pressing and holding lets me see flashes of the Setup menu, but I can't seem to get it to "stick."

 

I'm done for tonight, but I realized a little while ago that I do see Jon's splash screen for his standard U1MB build upon initial power up. I guess I can try to access the setup menu from there. I can try to enable SDX, then see if I can flash the U1MB firmware to the latest 1088XEL-specific build. Perhaps the longer reset/power-on delay he refers to in the release notes will help?

 

​And that reminds me - I need to find some kind of microUSB-to-USB cable so I can test out the Sparkfun interface.

[+MacRorie]

I'm sorry, I think I was imprecise or incorrect in my initial post. Pressing F12 simply seems to reset the machine. Pressing and holding lets me see flashes of the Setup menu, but I can't seem to get it to "stick."

 

I'm done for tonight, but I realized a little while ago that I do see Jon's splash screen for his standard U1MB build upon initial power up. I guess I can try to access the setup menu from there. I can try to enable SDX, then see if I can flash the U1MB firmware to the latest 1088XEL-specific build. Perhaps the longer reset/power-on delay he refers to in the release notes will help?

 

​And that reminds me - I need to find some kind of microUSB-to-USB cable so I can test out the Sparkfun interface.

 

 

Just out of curiosity, have you flashed the U1M with the updated BIOS or does it still have the old BIOS in it?

[_The Doctor__]

He has the updated FJC standard BIOS on the U1MB, he does not have the 1088XEL FJC U1MB bios yet.

[flashjazzcat]

You could eliminate keyboard problems by taking out the battery for a few seconds to clear the NVRAM and then powering on the machine. It should boot straight into the setup menu.

[+mytek]

It is odd that you are not able to get to the setup screen after power-up has completed. Other than the battery trick Jon just mentioned, you could also try the traditional HELP (F10) + Reset (F8) to get you there. Although the F12 macro has been working reliably through the last 4-5 TK-II PIC firmware versions, and should do so whether or not you run the 1088XEL U1MB BIOS.

 

But eventually you need to flash the 1088XEL BIOS anyway, so may as well do that by any means possible. But I have my doubts that will cure the problem you are seeing.

 

BTW, any charging cable for either an Android phone or tablet should work for the SIO2PC connection, and these are sold for pretty cheap at places like the Dollar Store or Walmart.

[+DrVenkman]

Thanks guys. I’ll explore further tonight after work. The U1MB has Jon’s latest “regular” build from December 2016. And the TK-II PIC chip has Michael’s version 2.2 that he release the other day.

 

The inability to get into setup with F12 persisted across two PS/2 keyboards. But once the machine booted up and dropped into BASIC (which was I guess what I was doing with it before I explanted it from its home in an 800XL) the keys all seem to work. I could access the TK-II settings screen, move the cursor around with the function keys, and as you can see, Altirra BASIC at least shows the correct amount of base memory so things are mostly working at least.

[flashjazzcat]

The most recent photos you've posted of the motherboard highlight a shortage of solder around a lot of passive component legs on the top side of the board. I'm certainly not asserting that that's the cause of the issues you're seeing now, but it's the first thing I'd be looking at if I had a board exhibiting issues. The lack of solder might become a problem if the same thing is going on under IC sockets and connectors which we can't see in the photos. You mentioned you had to reflow a few "dodgy solder connections" to get video working, so if all else fails I think it might be worth going over the back of the board again and trying to flow more solder through the vias. The trick there is not overdoing it, of course, since it's a tightly packed board and it would be easy to create shorts. I know you were struggling with the board sinking heat from the iron, and it looks like some of the vias weren't heated through to the top side.

[+Firedawg]

The most recent photos you've posted of the motherboard highlight a shortage of solder around a lot of passive component legs on the top side of the board. I'm certainly not asserting that that's the cause of the issues you're seeing now, but it's the first thing I'd be looking at if I had a board exhibiting issues. The lack of solder might become a problem if the same thing is going on under IC sockets and connectors which we can't see in the photos. You mentioned you had to reflow a few "dodgy solder connections" to get video working, so if all else fails I think it might be worth going over the back of the board again and trying to flow more solder through the vias. The trick there is not overdoing it, of course, since it's a tightly packed board and it would be easy to create shorts. I know you were struggling with the board sinking heat from the iron, and it looks like some of the vias weren't heated through to the top side.

 

Totally agree with FJC as I have mentioned before while working with Michael to find my cause of no video out was simple I had some solder joints that were questionable and ultimately was causing my issue of no video. Getting the proper iron temperature for the type of solder being used is crucial for making a solid electrical connection from the pad to the component lead. I just started my investigation with the components that were responsible for producing video and worked my way back as all the test points for power were correct.

[+DrVenkman]

Indeed. I plan to reflow just about every joint on the board tonight just in case. Even with my standard 60/40 solder, the thicker board made it just a bit more difficult to flow enough solder all the way through the board. I was also concerned, especially at the beginning, because I had no concept of how durable or heat-tolerant the board would be and I was hesitant to overdo anything.

 

Anyway, it boots and appears to basically run. At this point it seems to be down to troubleshooting and I’m both patient and stubborn. So thanks for the tips and we’ll see how it goes!

[imacav8r]

Indeed. I plan to reflow just about every joint on the board tonight just in case. Even with my standard 60/40 solder, the thicker board made it just a bit more difficult to flow enough solder all the way through the board. I was also concerned, especially at the beginning, because I had no concept of how durable or heat-tolerant the board would be and I was hesitant to overdo anything.

 

Anyway, it boots and appears to basically run. At this point it seems to be down to troubleshooting and I’m both patient and stubborn. So thanks for the tips and we’ll see how it goes!

Are the pin holes not plated through? total flow through of solder should not matter...

[+DrVenkman]

They are but flexing the board, components heating up and such can reveal poor joints. Surrounding the leg with solder through the depth of the hole ensures that’s not an issue. Also, this is a 4-layer board with ground and power planes inside. So ground and power pins have to have holes filled well enough to ensure good stable conductivity from either the trace at the top or bottom of the board to the appropriate depth inside the hole to reach the plane needed (ground or power).

[Mr Robot]

Where did you get your SIO socket?

[+DrVenkman]

Where did you get your SIO socket?

Mine came from MacRorie with my kit. If I had built up the system piecemeal I’d have salvaged one from a dead 1050 I have.

[+mytek]

Indeed. I plan to reflow just about every joint on the board tonight just in case. Even with my standard 60/40 solder, the thicker board made it just a bit more difficult to flow enough solder all the way through the board. I was also concerned, especially at the beginning, because I had no concept of how durable or heat-tolerant the board would be and I was hesitant to overdo anything.

 

Anyway, it boots and appears to basically run. At this point it seems to be down to troubleshooting and I’m both patient and stubborn. So thanks for the tips and we’ll see how it goes!

 

Herb before you go hog wild on the re-soldering, you might want to first practice in one area so as to get the heat and your soldering technique right .

 

Here's the area on your board I was thinking of.

 

 

And here's an example of how the solder joints should look when everything has been soldered correctly.

 

 

In that 2nd photo you are seeing the end result of only soldering from the bottom side of the board, but you can obviously see that the solder flowed all the way up and around the lead on the top side. Note: The amount that flowed up is pretty ideal (you don't want too much, which can short across adjacent pads).

 

Looking at your board it doesn't even look like anything has been soldered (this is not meant as a criticism, but as an observation). Either or both of two things can cause this a) not enough heat was applied, or b) not enough solder was applied. Without seeing the bottom side of your board, I'm guessing not enough heat, which can be caused either by the temperature set too low on the iron, and/or the iron removed too quickly from the pad.

 

So first give it some practice, and then when you are getting similar results, go over the entire board.

 

Good luck

[+DrVenkman]

Thanks, Michael. That’s what I plan to do. Yeah, this board is challenging my soldering technique, that’s for sure. I did most of the solder joints with my iron at 750 F, using 60/40 leaded solder, just like I do for all my work on vintage computers and game consoles. Never had any issue getting the solder to flow through the board until this build.

[AtariGeezer]

Thanks, Michael. That’s what I plan to do. Yeah, this board is challenging my soldering technique, that’s for sure. I did most of the solder joints with my iron at 750 F, using 60/40 leaded solder, just like I do for all my work on vintage computers and game consoles. Never had any issue getting the solder to flow through the board until this build.

What is the diameter of the solder you are using? .062 inch or the tiny stuff. The flux in the smaller diameter solder seems to vaporize on builds like these which doesn't allow the solder to flow easily. The 062 inch solder has more flux and helps the flow through the 4 layer boards.

 

Another Tip: Apply the tip of the soldering iron to the pad first, then feed the solder to the point where the soldering iron tip and solder pad meet, count to three and remove the heat. It should flow all the way though this way...