Making progress on the SCELBI – I now have draft layouts of all 5 of the cards needed for a basic 8H system in CAD. Some are much closer to production ready than others, but they are all there. One interesting thing to note is that, so far, I haven’t found a single error in the SCELBI schematics that I’ve used as reference.

SCELBI card stack

I have also ordered enough 256×1 memories to populate 4 1K boards, so that I can build a full 4K 8H system. Turns out that American Microsemiconductor has 6000 MM1101A/A1 National parts with a date code 81 at 3.55 each. They are listed as 3-Input NAND-Function Logic Gate, but I’m convinced that that is a mistake. I hope I’m not wrong. 🙂 Older Intel 1101s will easily cost over $10 each and you’ll need 32 of them for just a single 1K board. Ignoring the price, you may have trouble finding that kind of quanity.

American Microsemiconductor usually has high prices, but I’ve been able to find a couple of fairly good deals for RAM there. By the way, if you decide to order a bunch for yourself, be sure to talk to someone and ask for a quanity discount.

http://store.americanmicrosemiconductor.com/mm1101a-a1n.html

For the SCELBI, I’m going to be selling bare card sets, so you will need to find your own components, which is why I mentioning this opportunity. Since I haven’t built up a system with these parts, I can’t guarantee that they will work in the SCELBI.

By the way, I just recently figured out what the H stood for, in 8H. H stands for hobby. We knew that B stood for business, in the 8B system, but somehow it never dawned on me that H stood for hobby. So it seems that from the beginning that SCELBI had plans to create a range of machines. One of the great questions is what would have happened if Nat Wadsworth hadn’t suffered the heart attacks. Could SCELBI, with their vision and engineering skills been able to evolve and remain competitive?

Lastly, I hope to get together a SCELBI podcast in the next couple of weeks.

To evaluate real world 8008 performance, I have done some performance measurements using SCELBAL in my 8008 emulator. The 8008 running a typical BASIC program, such as “99 bottles of beer on the wall”, executes around 32,000 instructions per second. That is .032 MIPS, for those of you that measure processor performance that way. In contrast, the 6502 is said to execute roughly 300,000 instructions per second. However that is not the whole story, as instruction set influcences performance of real world programs in a significant way. For example, a single memory access on the 8008 will require 3 instructions if the H and L registers are not set up in advance. The 6502 can access any memory location in the entire 64K address range with a single instruction, without any need for setting up registers in advance.

Running a typical instruction mix, the 500 KHZ 8008 processor executes instructions about 10 times slower than the later 1MHZ 6502 used in the Apple ][.

Putting the CPU card aside for a while, I put some effort into the front panel card this weekend. First pass is already done, which is record time for a first pass. I guess all the work on the CPU card translated pretty well into basic layout and library stuff, which I could reuse.

SCELBI Front Panel PCB

Next up will probably be the input card, which has a ton of complexity for a card with only 11 chips. Hopefully I can get a first pass of that card done in two weekends.

Here is a summary of some of the main changes:

My 8008/SCELBI webpage can be found at this link-
http://www.willegal.net/scelbi/the8008andScelbi.html

Click here to see it done: Scelbi Layout in 14 Seconds

Scelbi layout

This small section looks half-decent after the first pass, but there is a lot of tedious tweaking to follow. The rest of the board, isn’t even close to this standard.

The CPU board has been a lot of work. The next SCELBI board I’m going to layout will probably be one of the easiest, which is the front panel board.

Scelbi front panel board

If anyone ever said that these early micros were simple, they were wrong.

This board, alone, has 27 ICs, compared to the Apple 1’s 57 (including 16 DRAM chips on the Apple). However the complexity is really revealed by the number of vias. The SCELBI CPU board has about 240 versus the Apple 1’s 296.

SCELBI CPU Board

For each replica PCB board layout, I go through 4 main phases. Placing the major components is phase 1. Connecting the traces is phase 2. Next phase of this board will be aligning traces with an image of an original board. The last phase is an exhaustive checking phase. This is only the first of at least five boards I need to layout before I can power up a replica SCELBI.