That's right, folks! Another installment of Boxman coming your way!

I've programmed Boxman on Scratch twice, Python, Javascript, C++, Commodore Basic, TI-Basic 4 times, C for the CE, Blender, and Unity. So I thought to myself, "What if I made Boxman with no programming at all?"

So that's what's happening! I have designed 2 circuits in Eagle, a logic board and a video board, made entirely out of 74-series logic, 555 timers, and 3 parallel EEPROM chips. I'm to the point where I'm pretty sure it's done being designed (after over a month), but I want a second set of eyes looking at it to make sure nothing I've done is super dumb. The boards once they are made are going to be about 8" x 9" big, and the two will be stacked on top of each other with standoffs and connected with a 60-pin ribbon cable (much like old arcade machine boards. I took some inspiration from the Super Pacman board I found when making the arcade cabinet version of Boxman.) So, here are the links! I'm excited how far this project has come, and I'm excited to hopefully finish it when I get enough money to buy all the parts. (I plan on buying the components first, bread-boarding the major sub-circuits as laid out in the schematic before I buy the boards 'cause I think ahead )

Logic Board:


Video Board:

I was reading through the beginning of your post, and that was the immediate question I was about to start typing. I'm glad you'll be breadboarding it first. Don't forget small capacitors across your power lines with all that TTL logic. I can't wait to see how this goes! If it's successful, I feel like you should ship it to me for Maker Faire next year, so we can show calculator Boxman alongside hardware Boxman.

That would be awesome! But that's pretty soon, isn't it? I'm concerned with how quickly this could get done in relation to the fast approaching maker faire. I'm not sure I can get it done in the next couple weeks. It will definitely be done before 2019 maker faire, however, and I'd be happy to send it to you for that!

EDIT:: haha, upon reading your post more carefully, I realize that what I recommended is exactly what you said. Also, you'll be happy to find that there are indeed 2 10uF caps across the input to the voltage regulator and ground, and the output from the voltage regulator to ground. I'm learninnggg!!

I think I'm going to order the parts either tomorrow or thursday. I hope the lack of replies means that my logic is sound but something tells me that nobody had time to actually look at it which is ok, I've been staring at this schematic for over a month and at this point at least in my mind it will work fine. So I have the digikey carts put together, and I'll keep this updated with progress on breadboarding the major subsystems. Again, I would reaaally appreciate it if anyone could quickly scan my schematic if they have time to make sure there are no stupid mistakes in connecting the chips and things. Thank you!

Necro-update:

I came back after a few weeks of not looking at this schematic to find I had tied two buses together which I was supposed to be subtracting.... yeah. So I believe I fixed that problem and actually managed to do it while reducing the chip count! This was all in the section of the schematic labeled "Boxman Movement & Restriction Logic" if you care to take a look. Anyway, I still need to breadboard a few things to make sure certain chips behave like I think they do before I order the boards. But that shouldn't take long at all seeing as I already have all the parts I should need. I am hoping to get this project done or super close to done during this long winter break. I'll post updates here as they happen.

This is exciting!!

I finally pulled the trigger on ordering the boards! I'm in San Antonio for the Valero Alamo Bowl at the moment (heck of a bus ride from Iowa), but by the time I get back, I hope all my parts will be there. I also ordered some polycarbonate sheets to protect the front and back from any small children when it hopefully gets displayed at the World Maker Faire. Here are some renderings of the boards (they will be stacked on top of each other with standoffs, hence the holes) and before you ask about the batteries, it has the ability to be run off of an external 9-12V power supply.

Finally! I have gotten the boards, done the 6+ hours of soldering (which is still not completely done as you'll see in the pictures, but I estimate 99% done) and as of now, confirmed that the video board does work after a few bodges. These were things that were just wrong in the schematic that I breadboarded correctly but the schematic didn't have certain nets named correctly. The worst of these problems was a net named GND when it was just supposed to be its own thing. I overlooked it because there was no label on the net in the schematic. Anyway, enough talk, here are pictures!! (click to enlarge)




You'll notice a few missing chips and that the voltage regulators are missing on each board. I just haven't gotten them yet in the mail. The polycarbonate sheets for the front, back, and "screen" should come tomorrow which will be fantastic! And once I get all the rest of the chips I should be able to test the logic board and get it all working as well. I already know of one small error: The autorouter in EAGLE missed one signal which happens to be pin 14 (GND) on the Box ROM chip. A simple jumper will fix that.

Long term plan for the maker faire:
I plan to get this boxman working all the way, updating the schematics as I make the bodges IRL, then order a whole new set of parts and boards (in blue, because boxman ) which will be the one I send to Kerm for the world maker faire this year. That won't happen for a while, I don't think. But for now I'll keep posting updates about the state this project is in as they happen. All for now!

So, I'm at a point in trying to fix the logic board where I can't find problems anymore. I got the video board completely working which I don't know if I've mentioned here. But as you can see in the video, it still doesn't work... I can slow down the screen pulsing by adjusting the DIP switches for the box fall speed. The "Box Fell" status LED is constantly lit up which explains the perpetual sound coming from the video board. Other than that, I'm still trying to get it to do something anything different in hopes to learn something. The score works though! One thing I haven't thought of until recently is that I haven't verified that my ROM burning set up that I built on a piece of perfboard is actually burning 100% of the addresses. From what I can tell, based on the score counter, it does seem to work, but tomorrow I'm going to pull out a ROM chip and spot check addresses on a breadboard. That's pretty much all for now. Here's a video and a picture of one of the bodges I've done so far. Sorry for the video quality, I had to record this on the front facing camera on my phone.

Ok, I went to the local university electrical and computer engineering building to use an oscilloscope to probe signals to see if they were what they should be and I definitely learned a few things.

1. First, and perhaps most detrimental, the clock circuit is constantly putting 2 pins high instead of just one. This I presume is because I tied the latch and the clock pin together on the shift register and the latch pin activates before the bits are shifted. This means that my 8 input nand gate is still going to be low even though there is a 1 in the shift register's memory because the 1 is not latched out yet.
PROPOSED FIX: I will need to find a way to slightly delay the latch pin so that it latches after the bits are shifted. Please, if you have any idea how to delay a signal (preferably without using a 555 timer) let me know.

2. Another problem is that the 555 timer which is supposed to be triggering the box pulse in the Box Fall Timer circuit is normally high pulsing low, and furthermore, the low pulse lasts for 3 screen refreshes.
PROPOSED FIX: I will need to use an RC circuit and an inverter (probably made out of a transistor and resistor) to get that signal to be a very short one shot pulse high.

3. A third problem is that the pins coming off of the 8 bit latch in the Randomizer circuit are floating high when the latch chip has its enable pin (active low) high. This puts the output pins in a high impedance state, it doesn't sink them low.
PROPOSED FIX: I'll need to add 6 pull down resistors on those signals.

That's all I'm going to do for now. Once I make all these fixes, I'll go back to campus to use the oscilloscope again to find more potential problems that may surface once those things are working correctly. All for now!

Please, if you have better ideas for fixes, or better ideas to implement these fixes, post them here. Thank you!

EDIT:: I'm pretty sure this will work, but I would like confirmation: Adding a capacitor to ground, say 1nF, will delay a signal by a small amount because it will need to charge up the capacitor? Or will I also need a resistor to create an RC circuit here? IIRC, RC delay circuits are a thing...

I have fixed the first problem using a resistor and a capacitor on both tiers of the clock circuit, and have verified that the clock is indeed working correctly now. This picture is of the clock and latch pins on the oscilloscope. The latch pin is the one that is being delayed by the RC circuit:



I have also developed a circuit that takes in a clock pulse, in this case from a 555 timer, and turns it into very short one shot pulses lasting only nanoseconds. I will build this circuit in the coming days and implement it into boxman once it's done. This circuit is below. The green is the output, and the orange is the input.


And finally, I have fixed problem number 3 by adding 6 pull down resistors to the offending chip in the randomizer circuit.

I have found another problem, however, and this problem is very serious in terms of the circuit behaving as it should, but the fix is as simple as inverting a signal.

4. The output enable line on IC24, the chip that allows the correct octal latch for driving the screen to take in inputs, is being driven backwards. I missed the fact that this chip's enable line was also active low
PROPOSED FIX: I will use a transistor and a pull up resistor as an inverter to invert this signal.

I'll keep updates coming as they happen. I hope I can get this project to the point of even remotely correct functioning soon.


EDIT:: I have an updated schematic in EAGLE with the changes I have made. Here it is, for those who would like to look (click to enlarge) :