This is a short video dealing with the design process of a Keyboard Matrix for 8-Bit Computers. I talk about some of the hardware design principals as well as software considerations. More of a VLOG style post than anything. But worth a watch if you’re into this sorta thing.
Here is the review of the JLCPCB Fab House that has been promised. Short post since it is late. I’ll likely comment more later.
Now that the boards are verified to work as intended, I have added the boards to the shop. There is also a zip file available for download on the product page that includes a Bill of Materials, documentation for building the board, as well as the required software to use this board. Remember, you must have a DOS computer with a parallel port in order to use this programmer.
The PCBs are available here.
I was recently pointed to the Fab House JLCPCB, which is another cheap PCB fabricating service out of China. Seeing as they have some really good deals on their website I decided to order a batch of PCBs for review.
Typically I order my PCBs from Elecrow (a service that deserves its own review, I really love this company). But I think a side by side comparison would be beneficial. And as such, I will be doing this as soon as the boards come in. They’re currently being finished and will be shipped via the 2-Day DHL service. But, it would appear that they have made a mistake, as they started over in the production, they have a time stamped log of the boards as they’re being produced. It will be interesting to note the quality of the boards versus the Elecrow boards.
I had hoped to have the boards by this weekend. But, it looks as though I will have to wait until next week. As soon as I get the boards in the mail I will get pictures, and start comparing the two services.
The problem that I have been having with PLDs in the recent past may be close to being resolved. The last few times that I’ve ordered GAL22V10s from any source my programmer, a TL866 (aka: MiniPro), has been unable to program them. Considering that my G80-S Computer requires this PLD for it’s decoding logic, and that many users do not have a means to program these chips, this creates quite a problem. So I have set out on a quest to see if the problem with the over 120 GAL22V10s that I have have in my possession is truly a problem with the ICs, or if it is an issue with the programmer. I suspect the second is the cause.
So, I have ordered a batch of boards for the ‘ELM Simple GAL Programmer‘ and have built one to test. Above is the board. The person who designed the board used a Male to Male 24-Pin Serial Cable for some reason, rather than a proper Parallel Cable. I do not know why. But Since I do not own a 24-Pin Male to Male cable, I’m waiting on one to arrive. The rest of the board is done. I’ve already applied voltage and test the VCC and the 15v lines, and adjusted them to their proper voltages.
Assuming this works for programming the 22v10s that I’ve had so many problems with, I will likely redesign this board so that it is a little more proper, and then offer it in the RetroDepot store. However, I will likely sell off the extra boards that I ordered once I confirm they do in fact work. Until then, no sense in selling PCBs that I’m unsure of.
Thanks to Travis Hall, another bug has been discovered. This time concerning the ‘CTRL-O’ functionality. The good news, it wasn’t my fault. Unfortunately neither Dr. Wang or Roger Rauskolb had the foresight for the commonality of SRAM over DRAM. Nor SRAM’s uncanny ability to start up with garbage in memory. So, we had to add a single line of code into the INIT routine so that an initial value is written to the OCSW switch. The code, as it should be written is below:
LD (OCSW),A ;SET INITIAL VALUE FOR SWITCH
LD D,03H ;3X LINE FEEDS
You can patch your code if you’d like, but I will be uploading a patched version once I get a chance to sit down and update the website to include a downloads page. Until then, plug this code in, and recompile.
Please continue to report any bugs that you may find. Future users will appreciate it.
My copy of Lee’s 6502 Badge is nearing completion. Lacking only a 2mhz resonator and a slide switch that would fit the board, I have been forced to wait on parts from Mouser.
I have, however, made some substitutions. Probably the most visible is the choice of EPROM. I have a 27C512 socketed. This is because one can use that as a replacement for the 27C256 provided he places the code in the top half of memory. Second item of note would be the connector that has been added to make a detachable battery pack. I thought it would be a nice addition. Third, I didn’t have the proper resistor arrays, so I used one that was close, and the other I substituted with upright resistors. There is nothing more retro than that… And the last substitute was to make odd sized capacitors work, and replacing the 560nF cap with a 100nF cap. I doubt there will be an issue here. Per my previous conversations with Lee I am under the impression that many times he chooses a part because it’s what he has ‘on hand’, and it works. The alternative was to go with a 560nF Tantalum. It would work, but I didn’t want anything that volitile on my chest (I have seen many Tantalum capacitors go ‘boom’ in my history with older computers).
But, a long story short I only lack those two components and burning the ROM in order to have a working example of Lee’s 6502 computer.
But I have another problem. I’ve looked high and low, but have not found my TTL Serial to USB adapter. But, what I do have sitting on my work bench are a bunch of CH340G RS232 to USB ICs as well as plenty of 12mhz oscillators. Ok, so that’s something. I also have some MAX3232CPE RS232 to TTL ICs. So I have options, and should be able to wire something up. I’ve never tried my hand at making my own USB to TTL adapter. But it could be neat. Especially since I have my own 3D Printer. I could make it look decent.
I’ll post an update as soon as it is finished.
Thanks to Lee Hart, Retro Depot now has a couple of the boards for his and Daryl’s 6502 VCF Badge.
Since they just arrived in the mail there hasn’t been time to complete them. But that should be expected sooner rather than later. I believe the parts box even has a spare LED display left over from the z80 Membership Card that was built about two years ago.
In either case, as soon as it is completed you can expect a review. And the next time Retro Depot makes it to a Festival, I’ll have a nifty 6502 Name Badge / Conputer with me.
Maybe I should build a portable Terminal so that I can program on the go. Now that’s an idea….
The early computer and navigation systems used by NASA are extremely interesting, even in the modern day. The ways that they computed, stored data, and were interacted with are quite different than what we typically think of when we reference a computer in today’s modern age.
However, Francois Rautenbach has been made aware of the discovery of Rope Memory. And has worked quite hard on a method for dumping the programmed contents thereof. The video is below. And a followup video he posted sheds a little more light on the subject. But my question I have to ask is, is the data still valid? From what I can gather from my examination of the video, the module is made up of magnetic core storage, similar to early RAM. And as such, likely lost it’s data a very long time ago.