nesdev.org

byuu wrote:I am less fine with it being a surface mount connector. Getting that wired up to a breadboard for prototyping will be an absolute nightmare.

Assuming you're specifically looking to put a not-0.1" connector into a 0.1" breadboard without a converter, I fear you're doomed to failure anyway; I have never seen a connector where they're anything but the same pitch inside and out.

byuu wrote: I am less fine with it being a surface mount connector. Getting that wired up to a breadboard for prototyping will be an absolute nightmare.

Surface mount isn't as nasty as it seems. Before last spring term, I had absolutely 0 surface mount experience. After a dozen projects or so this summer, I'm down to 0.5mm pitch without any issues. 2mm pitch is HUGE for smd. I ordered 2 samples of the -50's, so that should work out to 4 connectors (in case I wreck one or two along the way...). Tell you what, if it fits, I'll send you one on a 0.1" breadboard header.

> Assuming you're specifically looking to put a not-0.1" connector into a 0.1" breadboard without a converter, I fear you're doomed to failure anyway

Getting things to breadboards is always fun. For my serial controller cable, I wanted to bypass the controller ports and run the 1.25mm x 11-pin FFC directly into a breadboard, but nobody makes 1.25mm breakout boards. So the idea was to mount an adapter to perfboard, then add an IDC header to it, then use an IDC cable to connect to an IDC breakout board.

In this case, worst case, I'd probably try soldering wire between each pin of a through-hole connector to a through-hole IDC header. Whatever works.

byuu wrote:>In this case, worst case, I'd probably try soldering wire between each pin of a through-hole connector to a through-hole IDC header. Whatever works.

If that SMD socket works, looking at the footprint, I would guess the pins would be large enough to accommodate soldering an IDC cable directly to the pins. You could even bend the pins so they point down instead of out.

If you wanted to bypass the controller port via the FPC cable, here's a connector (again, surface mount...) http://www.kyocera-connector.com/prdct/ ... /6207.html

I have through hole 1.25mm FFC connectors. I decided against it because I figured out how to make my own controller cables using 24-28AWG crimp terminals. They go right onto the SNES controller port terminals. It was a side step to try and get past the horrors of those Retro-Bit cables. Nicer not requiring people to take the SNES apart. Downside is I have never found affordable 24-28AWG crimpers, but I just solder stranded wire onto the terminals and that seems to work just as well.

Fun side note: the CPU actually has three output pins for the controller port. It's just that only one is connected. But if you soldered onto the CPU, you could clock 3 bits at a time via the controller port. Fun, but impractical for plug and play, as is the FFC bypass.

I managed to get the code to compile for the Atmega*U2 series chips, even all the way down to the 8U2 (though I'm not going to bother with it since the 16U2 costs $0.01 more than the 8U2 and has the same footprint and pinout, but twice the flash memory). I'm waiting to hear back on a second opinion on the changes I made to get it to compile, and then I need to double-check my schematic. Then I can trace the board. This might actually work!

Just wondering, but does your USART adapter make any use of the LED on the Teensy++? That's the one pin assignment I'm still unsure of with regards to porting the design over to the 16U2. Everything else matches up great, I'm just not sure on the LED. It's getting annoying enough to track it down that I'm about to just assign it to whichever GPIO fits best on the board, and if you need/want it, I'll go through the code and reassign it there...

Now that I've hunted through the documentation and various schematics, I think I've finalized the board, so I can start tracing. With this, I also have a pretty good idea on price.

ATMega16U2-AU: $4
ISL84544H: $1
USB-B Header: $1
16MHz Crystal: $0.50
Resistors/Caps/LED: $1-2
Reset Switch (optional): $0.50
PCB: $2-5 (varies)
SNES cable/port: varies

So the total cost is ~$10-13 plus the cost of the connector and cable (I'll be ripping the connector off a broken multitap I have sitting around) and maybe an enclosure, if you feel so inclined.

Revised board design:




Also, I should be getting those Samtec connectors in today or tomorrow, so I'll know soon enough whether or not they work for the expansion port.

I hooked a regular LED up to the power rail, which can be disconnected via the DIP switch. It was just there to say "the device is plugged in and receiving power." It isn't necessary at all, may as well keep the cost down.

You may be set on using the SNES connector, but personally I use DB9 male on the board. I also offer a DB9 female passthru so that the same controller port can do both controlling and PC communication.

Anyway, it looks neat. There may be more demand at $13, but it'll still be really small.

> Also, I should be getting those Samtec connectors in today or tomorrow, so I'll know soon enough whether or not they work for the expansion port.

Awesome, can't wait!

As far as the cost goes, that's more a matter of cost to build than anything. I'm not expecting to sell any. Routing traces ended up being far less painful than I expected. I may add a power indicator LED to the USB power line, that wouldn't take much. It wouldn't take much to modify this board for DB9 footprints, either.

I finally got around to opening my Naki Tribal Tap, and was pleasantly surprised to find that the 7-connector cable is actually terminated on the board side by a 0.1" Molex connector. That, and the controller ports are each attached to a small board that breaks out to 0.1" through-hole headers. With this discovery, I'm going to change up the board footprint accordingly, and the really nice part is the cable will be reusable, since I can just put a Molex header (or just 0.1" pins) on the board and leave the cable intact. Personally, I'd say this is even better than DB9. You could plug this directly in to a breadboard with just some straight pins.

I managed to get it all traced only to realize I had flipped one of the headers >.< Oh well, I think this should be good, other than perhaps choosing a different crystal and/or reset button. I may also consider adding another LED so I have one for power, one for monitoring Rx, and one for monitoring Tx. As it is, since it's not really using the LED that's there, if I leave it at 2, one is power, and the other I'll program to indicate Rx/Tx activity both to the same LED. I may play around with the spacing on the LEDs a bit to accommodate through-hole LEDs or surface mount, that way if I put it in an enclosure, I can opt for through holes and bend the leads 90 degrees to shine out the front of the enclosure...

I finally got around to mounting header pins on the controller ports. The pinout on the ports was broken out to 2.50mm , so I had to bend the 2.54mm pins a bit to get them to fit, but I bent them back afterward and they work just fine in a breadboard. I also transplanted the extra pins while I was at it, so all 3 of them have all 7 connections populated. It was surprisingly relatively painless, and I'm really happy with the results. I now have a matching set of cable and ports, all broken out to 0.1" headers, and all with the full pinout

You have male ends there, but the SNES deck is also the male end. So extension cables (male on one end, female on the other) won't connect your SNES to your board :/

This will be good for the passthrough portion, of course.

I'm not using an extension cable for my board, I'm using the cable from a couple posts back. It has the female controller port on one end and a standard Molex connector on the other. My board will have the passthru port mounted permanently (on the left of the board in this image), and a Molex header on the other. That way, I can connect the controller cable, but can also remove it if I need it for a later project (as you well know, 7-connector cables are more than a little difficult to come by). Here, I added the connectors to the board image, maybe that will help you see what I'm going for. Look at the cable I posted above and you should see it plugs right in to the header on the right-hand side of the board, and that's how I'll be connecting to the SNES.