Author Archives: Mike

SCELBI Cassette Read and Write Cards Ordered!

Posted on by

It should take a little less than two weeks for them to arrive. During that time I’ll continue work on moving my 8B system from a temporary chassis to one of Corey’s custom chassis.

I started on the read card in December, so these two cards took about 7 months to do. I guess I might be slowing down a little bit. On the other hand, I’ve had a lot of other things in progress at the same time.

Assumming that these two boards work out, this leaves only the keyboard and the 2 oscilloscope cards left to do to complete the entire SCELBI board set. The keyboard and scope analog cards should go pretty fast, but the scope digital card is a fairly complex double wide card, so it will probably take a while.

Apple 1 Registry Updated – One of Which May be Myth Destroying

Posted on by

This update is way past due, I had knowledge of some of the new units since last fall. Three new Apple 1’s are listed. The most interesting is number 1-0052,owned by James Coble. His story seems to shatter the myth that the Byte Shop, not Apple, added the serial numbers on the units in the first batch. There were some issues with that story before, but this information seems to completely shatter it. I have a sent a query out to some experts to see if they can comment on what James has to say. Stay tuned for an update.

Another interesting note is that the unit that was sold on ebay is reported to be going to Germany. The world-wide market for Apple 1 continues to be stronger than the domestic US market. I think the Ford Museum purchase might have been an anomaly.

Lastly, I’m pretty skeptical about the report from the bay area about an Apple 1 being given to a recycler. That unit is not going to get mention in the registry until far more concrete details (like a picture) are revealed.

Old School Engineering

Posted on by

I just viewed Dave Jone’s latest video blog about the Sony Walkman. For some reason, I decided to download the service manual and have a look. One thing that I immediately noticed, was the lettering on the schematic appears to be another example of Leroy Lettering.

This caused me to reflect a little bit about my first year or two of college and my first job in industry. Before discovering computer science, I started out in a mechanical engineering program. The first engineering oriented courses you took, were drafting courses. I had no trouble with perspective and different technical aspects of drawing. However, I really had trouble creating drawings that looked nice, clean and sharp. There was no mention of Leroy or any other mechanical lettering system, so we had to hand letter our work. To this day, I don’t have the hand of an artist, and I think that lack of natural artistic ability held me back in that program. The skill of an artist, developed or natural is something that is apparent in a well done engineering drawing or schematic.

In years gone by, many engineers would spend their workday at a drafting board. I started my first co-op engineering job, in the late 70’s. That first company that I worked for, still had a drafting department for creating PCB layouts, as well as an art department that did the artwork for manuals as well as marketing material. The “uniform” of many of the experienced engineers was a white shirt, dark pants, a dark tie and a pocket protector. Second level managers omitted the pocket protector and added a sport coat. It’s basically the look of the NASA mission control team for the Apollo program.

A few years later, in the early 80s, I remember going for a job interview at IBM’s small system division. That was the first home of the IBM PC which was located in Boca Raton, Florida. Almost all the engineers there still wore that “uniform”. I did see one guy wearing a colorful shirt and jeans. He stuck out like a sore thumb. The IBM employee that was with me at that moment, pointed him out, and said he didn’t really fit in to the culture.

How times have changed.

8B Assembly Manual Mistake Discovered, the hard way

Posted on by

I’ve started working on building up a second set of the “core” SCELBI boards. I call the boards that are shared between the SCELBI 8B and 8H, the “core” boards. These are the front panel, the cpu, the DBB, and the input boards. When I built my prototype 8B, I didn’t bother to make new “core” boards, I just borrowed them from my 8H. Now that I have a new set of the “core” PCBs in stock, I figured that I would make up a second set, so that I could run both the 8H and 8B at the same time.

As part of the building of these new boards, I figured I’d use the 8B manual that I am in the process of digitizing. Yesterday, after completing the front panel board, I discoved that all my LEDs were installed backwards. The manual states:

The anode of the L.E.D. (shorter lead) goes in the top most hole (furthest away from the card connector). The cathode of the L.E.D. (longer lead) goes in the bottom hole.

This is actually the reverse of how LEDs are made – the cathode has the shorter lead, not the anode. I didn’t encounter this issue when building the board for my 8H, because the 8H front panel instructions were written for the old style LEDs in the cans, so I had to figure out the right way to connect them without benefit of the SCELBI manual. You would think that with all the projects that I’ve done, that I would remember how LED’s are constructed and connected. However pinouts are the sorts of details that are easily found in reference material, so I don’t put any effort into memorizing pinouts of anything.

As far as the newly digitized manual goes, one of my intentions is to add content and clarification when it is missing, but maintain the “character” of the original manual. I think, in this case, I will correct the text and add a footnote that explains how the original manual was wrong.

Cassette Write Board Verification in Progress

Posted on by

Verifing the layout against schematics, is an important step in my reproduction process. My usual method of schematic/layout verification is to print a copy of the schematics then mark each trace as I verify the connection with a marker. This time I did it all digitally using photoshop as a replacement for the paper.

Tape Write Schematic Markup

Tape Write Schematic Markup

The blue layer notes omissions, corrections or clarifications to the schematics. Most of these are due to the mistakes made in the initial rev of the board, which were later corrected. The version of the schematics I have access to, does not have those changes incorporated into them.

During this process, I did find a number of traces that were hidden under chips. Omitting these traces is an easy mistake to make. I missed one of these on the memory expansion board. This time, I am reasonably sure that I found them all.

Civil War Controversy: The Relieving of General Warren Before Five Forks

Posted on by

A while back, I created a web page with some of my views on controversial topics of American Civil War history. The first topic talks about my view of the efficiency of Army of Potomac Corps commanders in both 1864 and 1865. One of the most controversial changes was on April 1, 1865, when General Phil Sheridan relieved General Warren of command of the V corps. To this day, the correctness of this decision is debated by civil war buffs.

The website http://www.beyondthecrater.com has a number blog entries covering both sides of the controversy. What is especially interesting is reading the blog entry covering messages between various Union generals on May 30th and April 1st.

My view is that Generals Grant, Meade and Sheridan had every right to relieve Warren. In my view, Grant didn’t trust Warren to do the right thing in a crisis. The messages captured in the Official Records clearly indicate that Generals Grant and Meade were attempting to micro-manage Warren’s actions. If Grant had confidence in Warren, I believe that he would have given Warren general directions to work with Sheridan in order to turn the Confederate left and cut the South Side Railroad. A commanders lack of confidence in a subordinate should be enough to justify removing a person from such an important role.

Whether Warren was a competent corp commander or not, really is quite a different discussion. Even if he was actually the best corps commander in the world, if he didn’t have the confidence of his boss, relieving him of command, was the right thing to do. No commander should be saddled with a subordinate that he can’t trust in a crisis.

Web Page Unique Visitors

Posted on by

I’ve been using the same hosting service since 2009 and just went back and collected yearly stats for “unique visitors”.

Web Page Unique Visitors

Web Page Unique Visitors

The data for 2009 was only for 11 months so I multiplied by 12/11. The data for 2015 is for 5 months so I multiplied by 12/5.

The overall trend is for a general increase in activity, but this year could be slightly down over the past couple. However one real popular blog posting could affect the yearly results in a big way. November of 2014 had a major temporary increase in hits with my report of Steve Wozniak still thinking about improving the Apple II design generating over 26,000 visits for that month alone.

The total number of “unique visitors” over the last 6+ years is 537,704.

Dr. Galfo’s Integer BASIC Compiler

Posted on by

I just put up a web page with documentation and DSK images of Dr Galfo’s Integer BASIC Compiler (IBC). This compiler was used in development of some well known Apple II games. It runs around six times faster than WOZ’s BASIC interpreter, so it’s worth trying out. Compiling an Integer BASIC program is easy. You simply load the program in the standard Apple interpreter and then run the compiler. You will be prompted for a few options and then, after the compiler is run, are given the option of running the program.

New Version of SCELBI OS/X Emulator Released

Posted on by

The release of a new version of my OS/X SCELBI/8008 emulator was motivated to provide support of the Modified Creed Monitor for the 8008. To do this, I added menu options to support to optionally setting (input) and clearing (output) the most significant bit of input and output serial data. For the MCMON, the input menu for set the bit should NOT be set. The output setting doesn’t matter.