Category: making things

Tonight after a profanely busy day at work I did cardboard mockups of my current SDrive NUXX front and rear panel artwork. I’m mostly happy with it, but as my printer seems to stretch things out slightly on the page I have to double-check a few things (toggle switch hole and DIP switch markings, in particular), but I think that what I have is close to the final artwork design.

Here is an image showing the rear panel with SIO connector, reset button, etc. Note that the ugly blue switch caps were used because I don’t like them, so I was willing to throw them away on a prototype. The other switch caps may be seen here. I think I’ll be putting black on the front panel and red on the reset button on the rear. It’d be nice if a dark gray cap was available, but Digi-Key nor Mouser seem to list one.

Leave a Comment

Today I learned a little lesson about programming AVRs^†. After sorting that out I was able to properly use the SDrive NUXX‘s onboard programmer with no problem. After that I thought it should have been working, but the device wouldn’t work. Poking around I found the AVR running, but then found that the SD card slot was only getting about 1V, which is far below the 3.3V it requires.

This problem turned out to be caused by the LE33CZ 3.3V LDOs that I’d ordered from Digi-Key. As can be seen here, when fed 5V the LDOs were putting out just over 1V whose datasheets I had read wrong. I stupidly missed the “from bottom” note on the pinout section and been both fitting and testing the part backwards. (This is another thing to fix in the v1.1 / production run.)

For testing last night I worked around this by feeding the SD card 3.3V from a benchtop supply, everything worked great. The following morning I understood my mistake and fitted the parts properly, and now the benchtop supply is no longer needed. The 2GB SD card I’d picked up for cheap at Micro Center worked great, and I had no problems loading most of the .ATR images which I’d tried.

There’s still a couple problems with the PCB, but I’ll sort those out before the production order. Most notably there is the missing traces in the programmer, then I also screwed up the switch marking silkscreen for the front panel. Whoops!

At least it’s working. Yay!

Leave a Comment

After returning home from a really great vacation Sunday morning (more on that later) and sleeping, I got to work on the SDrive NUXX PCBs which had arrived last week. While the boards themselves look good and went together well, I’ve yet to get things working right. I first had some problems getting the chip programmed, but patching two missing traces (my fault) and powering the target device sorted that out.

Tonight I found that as soon as I set the fuses on the AVR, even after a successful program, the chip can no longer be accessed by the programmer. I believe this is because the programmer is then trying to talk to the chip at too high of a rate, but I’m not completely sure. Also, the option for setting the programmer to the slower mode (using PonyProg2000) is seeming to not work, so I’m not completely sure what’s happening.

Now it’s time to just relax and get ready for tomorrow. That’s another day of work, then maybe some bike riding. After that I can get back to working on this, and hopefully figure it out.

More photos are available on pages 2 and 3 of the SDrive NUXX album.

Leave a Comment

Today I received some samples for a new (to me) type of front panel material which I may use on the SDrive NUXX. It is aluminum which has been colored and anodized in a manner which allows for very reasonably priced markings in black and silver (natural aluminum). As can be seen with the ~5mm text above, this method of marking has far more detail than what one sees in panels from Front Panel Express.

The downside to this new process is that I’ll have to order a fairly substantial number of panels (100) to get the price that I’m wanting, but hopefully they’ll sell.

After seeing how nice the sample piece looked, I began working up new artwork for the panels which takes advantage of this different marking method. Here is one of my first test results, showing a nicely Atari-ish font and what I feel is overall a more appropriate look for something which is to be used with 25 year old hardware. Hopefully tomorrow I’ll be able to get the rear panel roughed in, and somehow (nicely) tie them together visually.

There is one thing I feel really guilty about, though. In my search for a good font for the button / slot / LED labeling I ended up using the font Interstate, which is the same as was used throughout the x0xb0x front panel redesign. I guess it still holds as a good stuff-labeling font.

3 Comments

Well, that’s that. Last night right before bed I sent off the order for 10 of the SDrive NUXX v1.0 PCBs. With any luck they’ll be here around the end of the month. If these work well I’ll begin taking orders

As mentioned last night I had to make the PCB 3mm larger to best use the case but no adjustment to the buttons was needed. Now it’s time to spend my evenings working on things like the documentation site, figuring out how to handle payments (PayPal? Google Checkout?), and doing non-electronicy things like actually keeping my house clean.

Leave a Comment

Today I received the enclosures for the SDrive NUXXes and found that I forgot to take into account the 1.5mm thick plastic bezels. This meant that I had to make the PCB 3mm longer, so that it’s now 123mm x 100mm. Now I just have to finish tweaking how far back from the edge of the PCB the tactile switches sit and it should be time to call this v1.0 and order a prototype run.

Leave a Comment

This is a JAE D03 connector used for Atari SIO, pulled from the Okidata printer controller mentioned earlier. I used this to draw a part footprint for SIO connectors and so far this has worked out well. A couple minor adjustments have been made, but the original drawing is still accurate.

Today I received a few connectors I had ordered from B & C ComputerVisions and while they ended up being a slightly different type (side by side photos here) the footprint will work for these as well. Additionally, I pulled open all of the Atari hardware I have here and found that all the SIO connectors seemed to be one of these two types.

The Japanese-made hardware (1010 and 1027) had the thick-based JAE D03 connectors and the other stuff (1050 from Singapore and 800XL from Hong Kong) had some clone of the thin-based part. I had to adjust the outline slightly because the thin based parts have a somewhat larger footprint, but this wasn’t a big deal as the pads all matched.

I was a bit disappointed to find out that the parts from B & C ComputerVisions are pulls, complete with slightly bent / solder-y pins and dirt on them but they washed up nicely and should be fine to use.

For reference, here’s the four different (two types) of SIO connectors I’d found, taken from this post that I made at AtariAge asking if anyone knows of any different part styles:

JAE D03: In all the Japanese-made Atari hardware I have. Thick base, well made. Held down with self-tapping screws or riveted in place. (Photo)
MX 7514-4-13: Received from B & C ComputerVisions when I ordered two connectors. I expected them to be new, but they are very obviously pulls (dirty parts / slightly bent pins / solder on pins). (Photo)
CO12995 5-82-4 MEC-5: Inside a Hong Kong-made 800XL. Appears identical to the MX 7514-4-13. Riveted in place. (Photo)
UNKNOWN: Connector similar to the previous two, in a Singapore-made 1050. I could read TAIWAN on the back, but while I could ensure it’s not the same as the part in the 800XL I couldn’t make out the rest of the part numbers. Held down to the PCB only by the 15 pins and nothing in the larger holes. (Photo)

Leave a Comment

Here is the first draft of the front panel for the SDrive NUXX. Yes, I’m trying to make it as 80s / cheesey as possible while still looking decent. Today I received an order from Digi-Key and validated the footprints from all the received parts against a print of the PCB, and everything looks good. Here is a photo of the SD slot sitting on a print, which I think looks rather nifty. I also spent some time trying out the LEDs (Image 1 · Image 2) to ensure that their brightness sufficient and even enough.

Here is an image of the rear panel, including the relatively massive cutout for the SIO connector. Hopefully it won’t be much longer before I get my hands on some more SIO connectors, can validate their footprints / sizing, and then order up the prototype PCBs.

Needless to say, I’m getting pretty excited about this project. It’ll be really nice if it works out.

Leave a Comment

Today my non-profit license upgrade for CadSoft EAGLE (to version 5) arrived so I was able to begin laying out the PCB for the SDrive NUXX. The first order of parts should arrive later this week, and hopefully by the end of the weekend I’ll be ready to order the prototype run of boards.

(Yes, the main ground plane is missing from that rendering. The small one there is for the built-in AVR programmer.)

Leave a Comment

While it may not look like much, this is the first completed draft of the schematic for my rebuild / remake of SDrive. This may not look like much, but attached to every symbol is a footprint for every part which I intend to use in the finished device. All that is left is to validate the part footprints against actual items, lay out the board, give it 20 or 30 sanity checks, then order the prototype PCBs. Yay!

(While the name in that document says SDrive: nuxx Edition the name shall be SDrive NUXX.)

Leave a Comment