Hi Ed,
Yes I do plan to cut those pins off of the headers, and grind down what's left in necessary, so they won't dig into the board.
There's also 2 pins on the PS2, 1 pin on the power connector, and 1 pin on the 1/4" stereo connector missing (where the K6 connector happens to now be). I was going to drill 4 holes on the corners, but I am unable to make space to clear the power or ground planes.

Edit: I've found out how to do this, so I can drill corner holes for mounting.

EDIT: No longer using the 1/4" stereo connector.

Hi, EE
I was looking at the image that you link to in your signature ("65O16 Development Board layout"), and there actually seem to be quite a lot of opportunities to eliminate holes. Should I supply details? I'm guessing that won't be necessary, and that the image doesn't reflect your latest work -- is that right? Just trying to help...

-- Jeff

Dr Jefyll wrote:
... Should I supply details? I'm guessing that won't be necessary, and that the image doesn't reflect your latest work -- is that right? Just trying to help...
-- Jeff

Sure Jeff, I am open to learn new techniques, but I most likely won't be implementing them in this design as right now I'm in the final error checking stage. I could use your suggestions/techniques for the next memory expansion/IO board...
I update the layout pics every new post I make, but sometimes it takes awhile for a new updated pic to show. So right now the pic you see should be the latest.

Some of the changes are very simple. Pins 1, 2, 3, 4, 5, 6 & 7 of connector H1 (on the left) connect to traces on the top layer (red). These traces immediately -- and unnecessarily -- go through vias to the bottom layer (green). Instead, the bottom layer traces could extend all the way back to H1. The same applies for six more traces at the other end of H1. So it seems that a minor change can eliminate 13 holes... unless I've overlooked something??

Sorry I didn't speak up sooner. I understand if you don't wish to backtrack at this point.

-- Jeff

Yes, you're right. Thanks for noticing that!

I'll be able to update it tonight...
________________________________________________________
Ok, updated. Thanks to Dr. Jeff for those many pins being freed up, I was able to add a few more power vias from the main power connector to the power planes, and a reference ground signal pin for each of the headers. And still have 8 pins free. Also, I made the ground closer to the POR and pushbuttons.

Parts list is next. Then I'll make a schematic to double check pin assignments against the current board layout.
Then I order parts. A whole $75 worth of parts for 1 board, pretty cheap for a Spartan 6 board...
Once recieved, I'll check tolerances of various parts with a digital micrometer and compare the values to the layout to make sure pads will be aligned.
Then I order boards, and will receive them in 2 days.

Found a couple more via's unnecessarily placed near U2, the FPGA PROM, which saved me a few more holes. I'm near the point of completion, so any holes saved I think should be devoted towards grounding H1,H2 and H3.

The original Parts list posted on this thread has now been added to my sig. and updated.

Finished the main schematic. Forced me to see a couple miswiring problems that I've corrected.
Two more schematics need to be completed.
One with the bypass caps and power pin assignments.
One with expansion connector info.

ElEctric_EyE , my Deer friend .


First of all I'm a big fan of your work .
As far as I can see you have add 330 ohm resistors on your keyboard interface , is that enough ?
Maybe you should pull the up ? pc keyboard has open collector/open drain outputs .

Good luck .

bound wrote:
...Maybe you should pull the up ? pc keyboard has open collector/open drain outputs .

Good luck .

Hi bound! Long time no see. I think you're right. All schematics I see involving a PS2 interface have some sort of pull-up. I added 2 4.7K pullups for the data and clk lines. Although I did not need any kind of resistors in my Spartan 3 SoC PS2 keyboard interface, probably I got lucky. I must get this right the first time, since this is a production board and not wirewrap like the S3SoC...

For those who are following developments daily, you may have noticed I have made many, many changes to the layout over this past week. I took full advantage of free time that has been made available to me due to a change in jobs. Hopefully the change will be for the better, it was my voluntary change, so we shall see...

Anyway yesterday, I thought I was done. I was tightening up all the traces. I was nearing completion when I started to see a large gap in the bottom left of the board. This started making me think back. I had known I wanted to add an 8-pin SOIC SPI 8Mbit Flash. But I needed some interface direct to a PC besides the JTAG (to program the Spartan 6 core and the core ROM). I am trying to bridge a gap between this project and the hard work done by 2 parties (Bitwise and Teamtempest) on this thread here on 6502.org. The ROM is easy enough to program the necessary data using JTAG, although it is limited to a max of ~32Kx16. But how to transfer other data/user programs into this project? This is leading to a C++ software interface project which I have no experience, and also have no desire to learn. Hopefully someone will post some knowledge on this aspect...

But I found a relatively new chip made by Microchip that is a USB to UART. Very cheap at $2 each. Even more importantly, Microchip is supplying the drivers as well for it to work under XP, also Linux I believe. So I have added the SPI Flash and the USB to UART IC's, and also a mini USB connector to the board. Worst case I'll have at least some kind of communication (I hope).

This is it, no more will fit!

I do still need to finish the Bypass Cap/Power connection schematic. But my new job starts tomorrow, so progress will be slow in this final chapter of construction for the next week. Hopefully next week I can order boards and get something to work!

The parts list and Board layout are up to date. The schematic and header pinout still need updating...

ElEctric_EyE wrote:
...This is it, no more will fit!...

I am continuing to optimize the board connections on my spare time... This is a project that is going to be a total success or a partial failure. Worst case it will be a partial success. I'm not too worried about admitting failure as I do have a very keen attention to detail.
Anyway, we will all see soon. I will have to explain the details later after I order and receive the boards and take some pics...

Just now ordered the boards with fingers crossed!

It's very difficult to wait... I've cut out a 3.8" x 2.5" piece from an index card to get a feel for the real size of the main board, to size up mounting holes, connector sizes etc. I placed it up against the monitor and the correct sizing is below...
Those 0603 cap's and resistors are small. I am thinking of investing in one of those lighted magnifying lamps to work on this, and a pair of fine tipped tweezers. Maybe ebay has one cheap...

Most of the "mag lamps" I've seen have magnification from 2x to 4x. Not too good. So I got a 5x & 8x set of "clip on" magnifiers made for glasses. They're called OcuLens from Carson optical. Ordered a new set today from ebay for cheap...

Before I spend more money and buy parts for this board, it must pass my QC. Most people on a certain expressPCB yahoo forum are very pleased with expressPCB service & support. So I went with them. I actually am expecting one or 2 errors on my part, which is why I bought the eye lenses, in order to spot errors before ordering & soldering parts.

Are you getting the boards from EPCB?


There's no such thing as a good, cheap lighted magnifier. In any case, lighted magnifiers are rarely more than 5-diopter. You need to be in the 15-20 diopter range to comfortably work with 0603 parts. I use machinist's magnifiers, which fit over regular eyeglasses and produce adequate magnification. A jeweler's loupe also works.

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!

BigDumbDinosaur wrote:
Are you getting the boards from EPCB?...

Yes... Do you see a problem BDD?

No. I saw your comment about "I've cut out a 3.8" x 2.5" piece from an index card to get a feel for the real size of the main board" and was going to say you can print an exact size copy of your PCB from within EPCB's software. No cardboard cutting needed. You can also export the general board layout to a DXF file that can be loaded into most CAD programs.

Incidentally, if it were me I wouldn't have filled planes on the board. Filled planes increase stray capacitance and in some cases may distort signals, limiting the maximum speed at which the circuit can run. Since it appears you are using EPCB's miniboard-pro service, it might be worth considering the four-layer version. That's what I used for my POC unit. Four-layer allows denser component placement and as the inner layers are power and ground, crosstalk between layers is effectively eliminated. Something to consider...

_________________
x86?  We ain't got no x86.  We don't NEED no stinking x86!