NTPi

I think posts from here on will be chronological, but I can't give any guarantees.

Anyway, this Raspberry Pi (first generation) was my best friend's and he stuffed it away since it was underwhelming for him (I don't know what he was trying to do with it), so I decided to take it off of his hands to tinker around with. Sometime last year, I decided to turn it into a stratum 1 time server and got it all put together in the latter half of April, with the fine tuning being finished in May.

I started with removing the RCA video and the stereo jacks, since I was going to mount the antenna connection there. I also removed the ribbon cable connector for the camera (marked as S2 on the board), since I didn't need it and it might've been in the way of the GPS breakout board (it might not have, but I didn't want to take chances). I got frustrated with trying to de-solder the parts, so I ended up destroying the jacks to get to make it easier. The ribbon cable connector I ended up using the cutters to break the legs away from the board, almost damaging the capacitor at C1 in the process (I think it made it out alive). I did the work during lunch at work, since work was better equipped (or at least I thought I would've been better off doing it at work).

The newly-modified board in the lower-half of the unmodified case.

Instead of using the SD card protector add-on (that I would've had to buy), I decided to use a micro SD card adaptor that fits entirely within the case, which allows me to make a cover with a flat piece of plastic instead to cover the hole.

View of the micro SD card in the adaptor in the SD card slot of the Raspberry Pi.

I also decided to use some flat-black paint stuff to paint the stems of the light pipes, so that there'd be less bleeding, and once that was dry, I put the case together (without the screws) before setting it aside for the night.

The painted light pipes, holes where the jacks are supposed to be, and the outside view of the adaptor.

I then modified the case by cutting away the plastic between the holes meant for the RCA and stereo jacks to allow for the GPS antenna connector. Drilling an appropriately-sized hole wouldn't have worked well, so it's why I decided against it.

Though it didn't matter, I did spend a small amount of time to make it look decent.

Out of an old project box from Radio Shack (which was for my electronics correspondence course final back in high school, which became the first version of the soundcard switchbox), I cut an appropriately-shaped (and sized) piece to cover the massive hole and to mount the antenna connector to. I also cut a piece of thinner plastic to make up for the extra space.

It was turning out better than I was expecting.

Then, I cut another piece for the SD card slot, using the leftover piece of the panel I cut off of the box (the picture was taken after I cut both pieces from that panel), and also another piece of the thin plastic to fill the remaining space. While I wanted it to be completely flat, I didn't feel like cutting into the box any more than I already had, so I just took the more efficient route instead. The cover is tight enough to where it is fairly difficult to slide, but it wouldn't matter much with the micro USB cable plugged in.

Inside and outside view of the SD slot cover.

Afterwards, I drilled the holes to mount the GPS breakout unit, and the hole for the antenna connector. As you can see, I also cut a hole in the thin plastic (also labelling the orientation so that the hole lines up properly), so that the connector would mount against the thick plastic. I used some wires from motherboard harnesses along with an appropriate plug, which I had to drill a hole to allow it to be plugged into the GPIO header.

The drilled hole, GPS breakout mounted to the case, inside view, and outside views.

I finished it up by securing the antenna connector and testing it out. Eventually I got a hold of some leftover parts from work and redid the wire harness from the GPS to the GPIO so that the wires would be at a right angle from the pins. Unfortunately, the IDC connector that I used was 3-pin (two of them) and not 6-pin, but it didn't matter, since it was going to stay plugged in. While using the crimper, I didn't squeeze/release properly, which misaligns the plug, and ended up smashing the last in the crimp series (red wire). I crimped the connection properly, so it works just fine, but just doesn't look that pretty.

The connectors, close-up of the "bite mark", height of the pins compared to the wire exit, height of the plug compared with the GPIO pins, and the GPS end of the cable with orange wire cut.

Unfortunately, while reassembling the NTPi after taking pictures, I accidentally sandwiched the antenna cable between the spacer and the case, and luckily I caught it before it became a major problem. While it looks bad, I made sure that the cable wasn't shorted, and carefully reattached the GPS unit (the picture above was before disassembly for pictures). I also scraped the solder mask of the GPS a bit when I was using pliers to hold the nut to loosen the screw, but since it was just the ground pad, I wasn't too worried about it.

Top and bottom of where the cable was sandwiched, and the scrape on the GPS.

Well, needless to say, I put the case back together, and got it back up for testing. At some point I put in the coin-cell battery for the RTC before putting the NTPi into service. I played around with the offset amounts some, and at some point just left it alone when I was okay with it.

Inside picture to mimic a previous image taken, better inside shot, NTPi assembled with screws painted.

While the NTPi isn't at the location where I was planning on having it, I find it actually better than the planned location, since the GPS unit gets quite warm, and the magnetic antenna is able to stick to the metal drywall cover where the window is.

It used to be the time server for all of my computers, but since I've switched to Manjaro, it now just serves the time for my old smartphones (which are being used as clocks). I might modify the NTP config file and start using the NTPi again, but, that's for another time...

The case I used is the Cyntech case for those wondering.

Broken Picture Links 3

I've finished fixing all the images that were hosted on Photobucket. I was also going to write a post or two, but I'm pretty tired from work, so I'll think I'll leave that for tomorrow.

Broken Picture Links 2

I decided to fix all the links before writing any new posts, and the uploading process took a bit longer than I thought, so any images that was hosted with Photobucket will be completely broken until I fix them.

I know it's inconvenient, but please bear with me while I take the time to fix all the links. I'm not sure what I'll have to do, but hopefully it'll be as easy as pasting the new links over the old ones.

Broken Picture Links

Recently, as I've previewed some of the images of past posts, I've noticed some are broken (though work fine when you click on them), and I'll be slowly fixing them.

Photobucket is the one causing the problems (might also be that I turned HTTPS on for the blog, it's hard to say), which I used to like using, but with how things are recently, I've stopped using them for images for new posts.

Again, it'll be a slow process, since I have to hunt down the picture on my server before uploading it to the new image hosting site and then update the link within the post. For now, just hold the control key and left click for any images that don't appear.

15-pin D-Sub Testing Breakout Board

This wasn't the first circuit board that I've designed, but it was the first challenging one to do because I wanted to keep it as small as possible and that I had a lot to work around.

The holes for the pins and the mounting clip/hook thing wasn't really hard to do, since the datasheet told me all that I needed to know (even though I kinda ignored the suggested hole diameter a bit).

There was another 15-pin D-Sub tester at work that was made with the same PCB-mounted connector, a length of 15-conductor ribbon cable (originally 20-conductor), and some female molex pins with heat-shrink tubing around each pin. While it did the trick, it was annoying to sort through the pins, even though the main ones were labelled (which I later updated because there was a rare occasion that I needed to find one of the pins that wasn't one of the common ones).

I took the picture after I had discarded the twist-tie that was separating the main pins (rather after I discarded the mess).

Since I had created a couple boards or so before this, I figured this would be the much sane replacement for the mess of cables, as it would make it easier to use and be much more compact. But as I said, it became a challenge to keep the board as small as possible (to keep the board cost down) while trying to route the traces efficiently.

The smallest "standard" trace width in EAGLE is 0.254mm (0.01in, 10mil), which was what I tried to use, but when I thought I was done, the design rule for OSH Park told me that I had spacing issues (the spacing between the 0.254mm traces and the pads for the pins were too close to each other), and I think at that point, my only option was to make the traces as narrow as possible.

OSH Park's minimum trace width is 0.1524mm (6mil), which I used and was able to successfully complete the board. I marked pins 1 and 15, put circles on the main pins (1, 2, 9, 10, 13, and 14), put lines to denote pins 6 through 10, marked pin 15 for the connector (for proper mounting orientation), and then labelled the board on the bottom.

Bottom and top sides of the board without the connector.

I now just plug the cable that I make into the tester, have one test lead in contact with the cable and the other on the corresponding pin number on the board. I haven't used it officially yet, but I have tested it to make sure it works (not like I had already checked the design enough times to make sure I didn't screw up the order). Since OSH Park sends boards out in threes, I've got a couple extra for whatever.

Top, bottom, and front sides of the board with the connector.

Mei-chan Upgrade

Back in October (2015), I kinda vaguely posted about the Blu-ray upgrade that Mei-chan got. I'll just be giving a few extra details that I missed (and also a picture I never posted).

Because I had about 300 DVD disks to copy, I kept the DVD-RW drives connected (albeit not in the same place), since copying 4 disks at the same time is a bit more task effective (though probably not effective time-wise). It's probably the only time I'll ever have four optical drives connected to a computer.

This was after upgrading the assorted hard drives that Mei-chan had to two 1TB hard drives. I'm hoping I posted about that, but if I didn't, there's not much to say about it. The hard drive upgrade had to happen first before this because there wasn't going to be enough SATA connections (never mind the lack of space), as Mei-chan only has six SATA ports.

Just a quick picture I took to remember the moment.

Apps and Battery Life

The article I read is: Closing Apps to Save Your Battery Only Makes Things Worse

I've seen a few articles about this on Google Play Newsstand, and have been ignoring them. Why did I finally give and click on it? Because Ting decided to post it to their Facebook page.

I'm neither going to agree nor disagree with the articles, simply because it depends on how you use your phone. I hardly use my phone, so I get home from work with about 90% battery life. I have a set of countdown timers running (because it's easy to lose track of time when I'm focused), and only really poke around on my phone during lunch.

I've always closed apps because I don't want it using CPU/RAM. Yes, the automatic resource management has gotten better over the years, but every-so-often I'll notice the phone lagging a bit, to find that I left a ton of apps open. This is mainly why I'm not going to agree or disagree with the article, because it's really a case-by-case issue. For the general public that spends almost every free second on their phone, yes it's great advice, but for light users, it might not be the best idea.