Look! Over there! A pile of bad news!

I’ll get straight to the point here: There’s no silver lining to this post. There’s a lot of bad, and a lot of scrambling to make the best of a very bad situation.

To start it all off: The screens on the V3 badges did not work. More accurately, they do work but they have a different driver chip in them. It is a problem that is fixable via code, but I am not a programmer in anything more than the most basic things. I can take existing libraries and combine them, but interfacing with hardware on the level of a display chip is well beyond me. For completeness’ sake, the driver is an HX8347D – In SPI mode/pinout. I have had a hard time finding any libraries for the HX8347D in such a configuration, and then there’s another hurdle – None of the libraries I’ve found work on the ESP8266. Fixing this issue in, at this point, less than three days is just not going to happen.

In addition to that, the V3s have a handful of other problems. There’s something screwy between the ATTiny and the ESP8266, which is causing the ESP8266 to not boot properly. This IS a code issue, as putting the ATTiny into a bare code state with only specific pins set does allow the ESP to boot up. The CP2102 also doesn’t connect to the ESP module – I’m not sure what the issue is as I haven’t poked into it at all, but the module connects to USB, but it doesn’t connect to the ESP. And the final issue is that the adhesive used for the battery packs is just not good enough. The pressure from the wires is enough to eventually separate the battery pack from the board.

On the plus side: The voltage regulators work – Both of them. I had also included a method for bypassing the CP2102 when prototyping, so that works to upload code. It’s annoying because an external adapter is needed, but it works. The ATTiny also works largely as intended from what I can tell (Although I haven’t tested it much).

So where do I go from here, with a pile of largely unusable DigiBadges?

I’ll likely simply do nothing with the pile. Replacing the screens is costly and time-consuming, resources better spent making sure this issue is resolved for future versions. I may put them up on the shop with a notification about the issues. Doing this will take some time, however, so I’ll need something in the meantime.

That something is going to be a modified V2 badge. It won’t decode JPEGs, but it’s a known device. It works. I’ll have to do some modifications for components that the PCB house can source and place, but that should be fairly simple. It should also be a bit cheaper than the V3 design, since it has much fewer components. Depending on cost, I may keep it around as a “Light” version of the badge.

Additionally, because of the no-longer-linear progression of badges, I’m not going to call them the V2+, V3 New, or whatever. I’m not entirely sure what I WILL call them at this point, but trying to keep generation numbers would just become a headache.

As far as BronyCon goes, I’ll still be there, at booth #33 in the Artist’s Alley. I’ll be showing off the V3 badge I’ve made work by swapping the screen on, and will be more than happy to answer any questions you may have.

As always, feedback, input, suggestions and the like can be voiced on Discord, on Twitter, or on Facebook.

V2 Case Almost Finished!

Hello! It’s Alabaster with a long-overdue update.

I’m so close to done. There’s not much left for me to do. It’s taken me so long for so many reasons, and I’d like to talk about that a little before I mention my progress.

At BronyCon, I mentioned to those who asked that the case would be done mid to late August, and that never happened. I wanted to figure out a way to make the case nicer than the last one, and was having a hard time finding a way to do that. I’m not an expert designer, everything I know is self-taught for the most part. I tend to find myself in a position where I’m just not skilled enough to design the things I think of, and I try and work around that. On top of that, there’s the fact that 3D Printing isn’t always exact. I could design two parts that fit together perfectly, and have them not fit when printed. It’s a constant cycle of printing things, and making adjustments to things by a tenth of a millimeter.

The problem I had with the V2 case was the way the lid would attach to the case. Andon and I threw a ton of ideas at each other, but I was having a hard time replicating them in a way that my printer could keep up with. In the end, I’m fairly happy with what I was able to do, and I’d like to hope that everyone else will be too.

Here is what the case looks like in it’s entirety, right now. Mind you, I haven’t smoothed out all the edges, and I haven’t finalized anything yet.

Front view of the assembled design file

Front view of the assembled design file

Rear view of the assembled design file

Rear view of the assembled design file

I have a few photos of the printed and assembled case. Taking these pictures, I noticed a bunch of little errors that I may have missed otherwise. I’ve made note to change about 15 things, and I’ll go do those as soon as this gets posted.

View post on imgur.com

View post on imgur.com

View post on imgur.com

In the end, you can see I was able to successfully make the lid better. I made a real, working hinge! Inside the hinge is a 1mm diameter metal rod keeping it all together. It’s pretty simple to look at, but I had to reprint it several times before it all fit correctly more than 60% of the time. The only thing I have left to do for the case is make a latch on the front. I have a few ideas on how I want to do that, and that shouldn’t take me too long to figure out. More than likely I’ll end up using the same idea as the v1, except just a tiny bit in the front or the two front corners. We’ll see what works best while not looking like garbage! I really wanted to bring in some external parts for the latch, as well as the hinge, but that’s not going to happen, sadly.

As always, feel free to reach out to me on Discord, or at my email alabaster@matchfire.net

 

Alabaster on the V2 Case

Hey guys, been a good while since I posted, and there’s a kinda good reason (not so much) for that. I’ve been having rampant computer problems that have almost barred me entirely from using my Desktop computer (the one I do all my work on). Turns out, the memory leak I was trying to hunt down and kill was in more than just the three programs I had removed from my desktop about a month ago, but in Windows itself.

Long story short, the problem is fixed now. And I need to catch up. Badly… I’ve been putting of saying anything until I had the problem sorted out, and now that that time is now, here we go!

SO! I will be scrapping the half-finished design I had been working on in order to favor a new one. That being said, now that I’m not scraping along at 98% memory usage all day, I have the ability to both work faster, and show my work to you guys. As some of you may have already seen, I made a post on Twitter asking for votes on the type of media you would like to see (either a livestream, or a timelapse). HERE is a link to the tweet, so you may go vote. Now, keep in mind, if I do a livestream, I will also be posting that on the YouTube, if a bit later than I would post the timelapse.

 

Anyhow, yeah. Like Andon said at the end of his last post, I will be working to make it so that the V2 case is compatable with the V2+, V2.5, V3, whatever we call it. That way, you guys don’t have to wait months again for me to get my act together. I promised people at BronyCon that I would have the case design done by mid to late July, and I failed to deliver. And for that I am sorry.

My new “release date” is early to early-mid August.

 

As usual, if you have any questions, I can be reached on Twitter @MysterAla or through my email alabaster@matchfire.net

It’s a bird! It’s a plane! It’s the Super Badge!

Greetings, everyone! Sorry I’ve been so quiet lately. After BronyCon I had some family issues that took up a large amount of the week after that. Since then, I’ve been mostly relaxing and playing large amounts of Fallout. But, I haven’t been entirely lazy. I’ve been mulling about doing a “Super” DigiBadge – A larger one with many more features, but at a higher, potentially significantly higher, price point. After BronyCon, Purple Tinker – The founder of BronyCon and a, well, tinker, hopped on our Discord Chat (Yes, that’s a link – Join us!) and asked about… a larger DigiBadge with more features. The conversation that ensued was glorious, with loads of ideas bouncing back and forth.

The biggest issue with the screens for the Super are that they require a VERY large number of pins. Put one on the 328 and you have just about nothing left. The obvious solution is to find a screen that uses fewer pins. We didn’t go for the obvious solution, because while price is less of an issue, it’s still a huge motivator. Instead, we went for the possibly-insane but workable solution of two microcontrollers on the same board. An ATMega328 will drive the screen, and will be programmed to accept commands to display things. Essentially, it’ll be the GPU of the system.

The real brains of the project will be an ATMega32u4, the same that is found in the Arduino Pro Micro. In addition to having 3 additional Digital pins over the 328 (Bringing the total to 16), the 32u4 also allows for six of those digital pins to be used as analog pins, bringing the usable number of analog pins to 11. That’s not even the biggest feature. That belongs to the fact that the 32u4 is USB-Native. What does this mean? You won’t need a special programming chip to interface with the device, making it significantly easier to program.

After we figured out how the project would work, we started figuring out what sort of other goodies we could throw in there. Below is the current list of features that I’m planning on putting on the Super, but be aware that it will likely change before the everything is said and done:

  • 2.8″ TFT LCD screen
    • Touchscreen, too!
  • MicroSD card slot
  • ATMega328 for GPU
    • FTDI Connection will be available for reprogramming
  • ATMega32u4 for CPU
    • USB-Native support for reprogramming the ATMega32u4
    • All unused pins for both microcontrollers will be broken out
  • 2500 mAh 3.7v LiPo Battery
    • Including charging circuit
    • Will charge if Super is plugged in but off.
  • 3.3v Regulator
    • No more screen dimming as the battery fades
  • USB Micro port for programming and charging.
    • Charge on the go with your phone’s charger!
  • 1Mb on-board flash storage for saving settings
    • Set a “Favorite” badge to default to.
    • Start to a “Favorite” image with an Art Card
    • Set and display your name!
  • 3-Channel Multiplexer/Demultiplexer
    • Both chips can share the SPI bus!

That last bit is a late addition to the board. Most of the Arduino-to-Arduino communications solutions out there are one-way. I wanted to have the 32u4 have access to the SD card, but the 328, being the GPU, would also need access to it. I’d have to code in a way to get information from the SD card, into one ATMega, and then to the other ATMega. This seemed a little excessive. The solution is a Multiplexer/Demultiplexer or MUX chip. This chip allows easy switching for three channels of communication (Clock, MOSI, MISO) between two sources to one destination – Or from one source to two destinations. And, it can all be controlled via the 328, so the 32u4 doesn’t even lose any pins, aside from those used for communicating with the 328 and the SPI pins. But those were going to be used anyway. The 32u4 would simply send a command to the 328 saying it wants the SD card, and the 328 would set the MUX appropriately and then use one of its own pins as the CS pin.

We had explored some other options for the boards, such as Bluetooth or WiFi capabilities – But then we run afoul of FCC regulations. Currently, Matchfire boards fall under the “Subassembly” category, making them exempt from FCC certification. If the FCC were to tell us that they didn’t comply with the Subassembly category, certifying the boards with the FCC would cost somewhere between $1000-$2000, which is a lot but not terrible. That all changes when you start throwing around wireless transmission. If you’re building your own wireless device, you have to get it certified. There’s a few caveats to that, but one of those is that you can’t make more than five boards, and another is that you can’t advertise them for sale. Certifying such a device runs somewhere between $10,000 to $25,000. Or more. It’s expensive, and Matchfire does not make enough money to pay for that sort of thing. They’re a little more lenient when you’re using a pre-built wireless transmitter, but those are a bit more expensive and outside of the price range of what we’d like for default inclusion. I plan on designing around users being able to include one of those, but it would have to be purchased separately.

Currently, I’m guessing, and only guessing, that the purchase price for these will be somewhere between $35 and $40.

Post-Con Breakdown

Greetings, everyone!

So, after spending the weekend at BronyCon and having a blast and running out of things to sell, I’ve had the chance to run the numbers. Lots and lots of numbers. I have some good news, better news, and plenty of thoughts along the way.

The most common question I received from other vendors and even a few people was “Did you break even?” – A question whose answer is not as easy as I would like. The short answer is “No.” But that’s with some of the most basic math. I’m not going to go into specifics, but that’s counting the costs for the Pendants, which weren’t available for sale at BronyCon. It’s not exactly fair to count something you didn’t sell, but it’s also not particularly easy to split out the costs. A lot of the components for the V2 were shared by the Pendant, and splitting the price isn’t exactly the easiest thing to do. Still, the best quick math that I’ve done says we just about broke even when you don’t take into account the Pendants. But you know what? I’m not too concerned with breaking even. I honestly didn’t expect to break even, as the production for the weekend was ambitious.

That said, it really helped a lot. As I mentioned above, I ran out of things to sell. That doesn’t mean I sold everything I brought. A number of the Version 2 badges suffered from something I missed – I expected the battery packs to be a bit tighter to the board than they were. This led to quite a few breaking a solder joint, which made them unstable as moving the battery pack would and did cause the power to cut, turning the device off. This is a simple fix, as it just involves re-soldering the pins and securing the loose end of the battery pack with hot glue or similar. I have a whole bag of V2s that broke in this manner that I need to repair, but thankfully I made quite a large amount and was able to sell somewhere around 80 of the badges. I haven’t gone through and done a full inventory of what I have left, so exact numbers will have to wait.

There were a few super-minor issues with the programming that will be addressed, and I’ll go through those here. First, when an SD card is inserted, but the card load fails, the badge will continually attempt to reload the SD card. This only came up with some V2 badges that had issuess accessing the SD card slot, but it would also crop up if a bad SD card was inserted into the device. The second “issue” is that, even with an SD Card loaded, the V2 will restart into badge mode. This will be changed into slideshow mode.

As far as money is concerned, it was essentially a wash. As far as everything else goes? Well, the response I received from people was fantastic. People loved the idea, and I ran out of sellable items. Going by sales numbers, if I hadn’t run into the issues I had, it’s entirely possible we’d have broke even, even after adding the additional cost of the pendant – Products that weren’t even available to sell.

And, for a final bit of good news, Matchfire has been accepted to vend at Nightmare Nights Dallas! I’ll have V2 digibadges there, I’ll have Pendants there, and maybe a new design or two. As I have the core products that I need (V2s and Pendants), I’ll be able to fund some experiments into a V2 “Super” – A larger DigiBadge that has more capabilities. I’m looking into various options for these badges, but at the moment I have my eye on a screen that’s both larger and a touchscreen.

That’s it for now! It was absolutely wonderful to be able to meet with everyone and the response I received was absolutely fantastic. Thank you all for a wonderful weekend, and I’ll keep you posted on updates and information about future products and appearances.

A most shocking update

Greetings, ladies and gentlemen! I’ve got a great big nice update for you today.

Over the last week or so, the parts orders that I’ve placed have trickled in and I’ve accumulated quite the pile of boxes. Today, I finally got the “Big One” – The final circuit boards! So, I immediately set about assembling the first “Production” DigiBadge V2 (I’d have assembled a pendant, but there’s one package due in later this week I still need). Used the solder stencil, put it in the reflow oven, hand-soldered the screen and battery on. Drop some batteries in, flip the switch…

Nothing. Batteries get warm. Out comes the multimeter! Check the resistance between ground and supply. There’s a short. Spot two of them on either side of the microcontroller. A dab with the solder wick fixes these real quick. Put the batteries back in, hit the switch.

Nothing. Except wait, the microcontroller’s getting hot! That’s typically indicative of backwards voltage. Check things, double check things, everything looks right. On a hunch, I look at the battery pack, and go grab my previous V2 prototype. Surprise! The battery packs are not identical – The ones I had been using had switched pins compared to the ones from the bulk source.

So I sigh in exasperation, take the batteries out, put them in backwards, and switch the thing back on. I didn’t expect much, as reverse polarity is typically the easiest way to fry a chip, but the badge fired right up and sprung to life, working absolutely splendidly. This issue will be addressed in the upcoming V2.1 DigiBadge, but if you get a DigiBadge at BronyCon or until I run out of stock on the V2, you’ll have to remember to put your batteries in backwards. Additionally, as this is a part issue with the battery holder – Something shared between the DigiBadge and the Pendant – this will mean that pendant batteries must also be placed backwards. As with the DigiBadge V2.1, this will be addressed in a Pendant V1.1.

But there is good news! The issue with the SD card resetting the chip seems to not be present with this specific badge. I’m not going to promise it’ll work with your V2, as I’m not 100% sure WHY it’s working properly, but it’s allowing me to debug and address the issues with the code. I’ll be keeping my eye on all of the standard-production V2s and seeing if it may have been an issue arising from me hand-soldering parts that weren’t designed to be hand-soldered. If it does rear up again, I’ve got an idea for how to address it in the V2.1.

If you haven’t yet, you can pre-order the V2 and the Pendant over in our shop!