why is everything breaking?

August 11th, 2014

Within the last few weeks I’ve had the following problem with mechanical devices.

  • rear hub on bicycle
  • crankcase gasket on lawn mower
  • 3TB hard drive used for backup for three household computers died
  • coolant leak in 2009 Dodge Caravan
  • and a CD/DVD drive on a PPC Mac mini used for backup and connection to old peripherals is showing signs of imminent failure
  • My son says that I am having problems because all my stuff is old.

    Except for the the coolant leak, I’m getting replacement parts and doing my own repairs/replacments. I’d probably look the coolant leak, except that I just finished replacing the crankcase gasket on the lawn mower engine, and don’t feel like diving into another engine right now.

    I’ve had a lot of issues over the years with the lawn mower which was built in 1994 and would replace it with a new one, except that repairing it is so much less expensive than replacing it. For instance, the crankcase gasket, PTO oil seal, and a welded muffler (which I found was busted when I pulled the engine to repair the crankcase gasket) cost me all of $35 to repair or replace. That wouldn’t even pay the taxes on a new mower.

    When I pulled the crankcase cover off, I was expecting to see an engine on it’s last legs, but the internals looked good. Over the last couple of years, it was using a lot of oil, and I was thinking that the rings were going, but now I think that the oil was just leaking out of the crankcase. Now I think that the engine might last a few more years.

    SCELBI 8B update

    August 11th, 2014

    Hi SCELBI fans,

    Based on scarcity of reports, it may not seem like it, but I am still actively working on the SCELBI 8B card set. It is pretty tedious work, and reporting every new set of tweaks to the layout is not very exciting, so I haven’t been reporting ongoing progress.

    To be honest, I prefer exploring new projects, rather than revisiting old ones. Though the SCELBI 8B is “new”, it leverages so much of the 8H design, that it really seems more like revisiting the 8H, than doing something new.

    At this moment, I am going through final design checks of the SCELBI 8B backplane, memory expansion, and 4K SRAM PCB layouts. I already have quotes and have done DFM checks with my PCB supplier. I think I will probably build up an very basic 8B system without I/O in a temporary chassis in order to get these cards checked out. This is so I can offer the PCBs for sale for those of you who can’t wait for the whole package. Based on what I know at the moment, I might be ordering PCBs in as little as a week or two from now.

    The chassis sheet metal and the EPROM card will come a little later as follow on efforts.

    After that, my next efforts will probably be the SCELBI keyboard and oscilloscope interfaces. Those will be “new” to me, so will be a lot more fun to do than the 8B has been. The O-scope interface poses some interesting challenges, as we don’t have any original software for it. We will have to craft some drivers using only the hardware implementation as a reference.

    The World War 1 Meuse Argonne Campaign compared to the World War II Battle for Normandy

    August 2nd, 2014

    This is a follow up to a previous post about my grandfathers involvement in World War 1.

    World War 1 seems almost forgotten these days. Compare the little remembered Meuse-Argonne Offensive with the Battle for Normandy. Each of these campaigns is considered a major contributor towards ending a tragic war.

  • The Meuse-Argonne Offensive lasted 47 days from September 26,1918 till the end of the war, November 11, 1918.
  • The Battle for Normandy lasted 86 days, from June 6,1944 to August 30, 1944.
  • The Meuse-Argonne Offensive involved 1.2 million American and French troops and about 450,000 German troops.
  • The Battle for Normandy involved 1.45 million Allied troops and about a million German troops
  • During the the Meuse-Argonne offensive, the French and Americans suffered about 187,000 casualties and the Germans about 100,000 casualties.
  • During the battle for Normandy, the Allies suffered about 225,000 casualties and the Germans around 425,000 casualties.
  • It seems a bit of a shame that the loss, pain and suffering of hundreds of thousands of people in World War 1, has nearly been forgotten.

    Eventful bike ride

    July 20th, 2014

    I had some issues during yesterdays bike ride.

  • strange sounds started eminating from the rear end. I stopped to check it out. Since it sounded a bit like the twanging of spokes and just last week I experienced a broken spoke, I decided to tighten the spokes on one side half a turn.
  • after the noise continued, I made a decision to continue the full ride, at a slow pace, instead of cutting it short.
  • a few miles later the noise increased and I got off the bike to check it out. turns out the bearings in the rear hub where completely shot
  • I did have a phone with me, but since it was 7:00 AM and the rest of the household are late risers, I decided to walk home
  • The bearings were so bad that the wheel would not track within the frame even when walking. Intermittant tire rub on chain stays, for the full walk home
  • about 1/3 the way home I slightly twisted my ankle. Started regretting deciding to walk home in cycling shoes
  • another mile and I twisted my ankle again – much worse this time
  • a little ways further on, I decided to remove the cleats from from my cycling shoes to make walking easier
  • cycling shoes slipping going up a steep hill
  • got home around 10, making it a 15 mile “ride” at an average speed of under 4 miles an hour
  • took the rear hub apart and found only a single broken ball in the right side races. The rest of the balls on that side must have disintegrated and fell out
  • the race in the right side cone is bad and an internet search reveals that cones are no longer available. Guess I’ll find a replacement hub. Meanhwile I’ll use my spare set of wheels
  • More Vintage Fonts – The SCELBI Front Panel Logo

    July 17th, 2014

    For a while I’ve had an interest in vintage typefaces. This interest stems from my efforts to reproduce vintage literature and the logos.

    When I made my SCELBI front panels, one of the challenges I had, was to match the font for the SCELBI logo. Here is an image of the overall front panel.

    SCELBI front panels

    SCELBI front panel

    and a close up of the logo.

    Scelbi logo- close up

    Scelbi logo- close up

    Even though this is a variant of the omnipresent Helvetica, finding a digital font to match was harder than you think. After searching through dozens of variations of Helvetica and derivative fonts, in the end, I choose to use the very similar Helvetica Black. In some cases in the past, I have used Adobe Illustrator to manually recreate logos and such, but recreating characters accurately is extremely difficult and it is vastly easier to use off the shelf fonts.

    Scelbi Logo in Helvetica black font

    Scelbi Logo in Helvetica black font

    It’s a pretty close match, but in reality the original logo’s characters are just a bit wider, so it’s not perfect.

    I’ve known for a while that the single largest source of typefaces for printers and advertising people back in those days was Letraset brand rub on transfers. Back in the day, anyone that had anything to do with the printed word, would have had a Letraset reference manual, which contains hundreds of fonts and other visual goodies. Vintage Letraset manuals are available from used book sellers and even on ebay. Recently I picked up a 1981 edition.

    Letraset Reference

    Letraset Reference

    One the first things I did when I got this guide was to determine if I could find an exact match for the SCELBI front panel logo. Here is what I found on the page with bold type Helvetica fonts.

    Letraset Helvetica Bold

    Letraset Helvetica Bold

    If you ask me, plain old Letraset Helvetica Bold is an exact match and was the source for the lettering on the original panels. Now I’m just going to have to figure out the best way to digitize and scale the example letters found in the Letraset guide for my next batch of front panels.

    Keyboard Gizmos

    July 15th, 2014
    parallel keyboard gizmos

    parallel keyboard gizmos

    Here is a popouri of parallel keyboard projects that I have been involved with designing.

    Not shown

  • A simple single chip AVR PS/2 to parallel keyboard adapter that I forgot about when I assembled the items for this picture.
  • At top is:

  • A reproduction Datanetics keyboard
  • At bottom, from left to right…

  • Home etched/prototype PS/2 to parallel keyboard adapter
  • Production PS/2 parallel keyboard adapter configured for Apple 1/Mimeo
  • Production PS/2 parallel keyboard adapter configured for Apple ][. The latest version firmware can also be used with other vintage computers
  • Home made Apple ][ keyboard to Apple 1/Mimeo motherboard dongle with clear switch - based on schematic at Wendel Sander's Apple 1 site
  • Production Vintage Micros Apple ][ keyboard to Apple 1/Mimeo motherboard dongle with clear switch. Similar to home made one
  • Corey Cohen’s parallel keyboard multiplexor. Automatically accepts and switches input from two different parallel keyboards to a single destination motherboard.
  • Not shown are a number of projects that I have started, but not completed (yet)

  • Datanetics replacement using modern components
  • MM5740 replacement using modern micro controller
  • And then there are the projects that never got past investigation stage.

  • ADB bus to parallel adapter
  • PS/2 Apple IIe keyboard adapter
  • It’s really hard to imagine that I have spent so much time mucking with simple parallel ASCII keyboard technology.

    Vintage Micros has Apple 1 motherboard to Apple 2 keyboard dongles available.

    July 13th, 2014

    I helped with the PCB layout for this cute little dongle. It has a clear screen switch, which isn’t normally available on an Apple II keyboard and does the rewiring necessary to connect a stock Apple II keyboard to an Apple 1, Mimeo or other clone. These are not needed for replica 1s, which has the keyboard pinout necessary to work with Apple II keyboards.

    Here is one in action, hooked up between the keyboard and the Mimeo.

    Keyboard Dongle

    Keyboard Dongle

    They are being sold by seller vintagemicros on eBay: ebay listing

    One word of caution – be sure to connect pin 1 of cables to pin 1 on PCBs. If you reverse them, you will probably blow the 7404 on the keyboard’s encoder board.

    One last thing, I connected pin 4 of the keyboard socket to the clear screen input on the Apple 1. Pin 4 is normally not connected on an Apple II keyboard, but if you make the keyboard encoder mod to use the repeat switch as clear screen input as described on Wendell Sander’s site, it will work without any further wiring changes.

    Model Rocket Launch Circuit

    July 5th, 2014

    A few years ago I built a launch controller for Estes type model rockets. This was designed to be powered by a 12 volt battery. I usually use a battery off a riding lawn mower or motorcycle. When I was a kid I used to use a motorcycle battery that had a bad cell to launch model rockets and it worked great.

    complete launch control

    complete launch control

    This controller has two cables that are used to carry power from the battery to the ignitor.

  • A pair of clamps at the end of 10 feet of cable are used to connect to positive and negative terminals of a 12 volt battery.
  • A pair of small alligator clips at the end of 20 feet of cable are used to connect to the rocket engine ignitor
  • launch controller

    launch controller

    This controller has the following features.

  • A 1/4 inch phone jack used as a safety key. With the plug removed from the jack, the power is disconnected from the ignitor circuit.
  • A two color green/red LED – The LED’s red light is lit when power is available from the battery. The LED’s green light is lit when the safety plug is installed and power is available. Since both LEDs are lit when power and safety plug is installed, the resultant light looks yellow.
  • A green LED that lights up when power is available, the safety plug is installed and the ignitor is connected.
  • A red button that will launch the rocket if all LEDs are lit.
  • Now the reason for this post. The last time I used this controller, I did have a rocket launch imediately when I inserted the safety key. The LED that indicates that the ignitor is connected was designed to draw 10 milliamps through the ignitor and that current is flowing when the safety key is inserted. Though I tested that current through a number of ignitors, it must have been enough current to fire the one that went off prematurely. What I have decided to do was to substantially reduce the test current through the ignitor. The easiest way to do this was to use a smaller test current through the ignitor and use a transistor to amplify the current to a level high enough to light the indicator LED.

    Here is modified schematic. By the way, don’t bug me about the LEDs which are shown connected backwards in this drawing.

    launch controller schematics

    launch controller schematics

    With this design the test current through the ignitor can be roughly calculated as:

    12 volts/100000 ohms = 120 micro amps

    I think that this tiny current would be extremely unlikely to launch a rocket inadvertently. I measured the current through the LED as about 30 milliamps. This means the transistor is multipling the current by about 250 times.

    SCELBI 8B update

    July 4th, 2014

    Here is a composite image showing the great progress that has been made on the 4 boards that need to be done for the SCELBI 8B.

    Composite Image of 8B Boards

    Composite Image of 8B Boards

    Still a ton of tweaking and fine tuning needs to happen, but you can almost see the light at the end of the tunnel. I also need to find some time to figure out what I’m going to do for the custom chassis.

    The backplane layout is for all intents and purposes done. The memory expansion card is next closest to complete. I haven’t done much with the PROM card, but I’ll do a complete review of it. The curved traces promise to require a lot of effort to match up to the original. The 4KSRAM cards need a lot of work and since the original has curved traces, it will take an extra effort to finish.

    More on the Sightseeing Sixth Division in World War I

    July 1st, 2014

    In a previous post, I noted how my Grandfather was a member of the Sightseeing Sixth Infantry Division in World War 1. I found a history of the Sixth Infantry in World War 1, written at the end of the war. The history can be downloaded from this site

    Using this history and Google Maps, I created a rough map of where my Grandfather might have travelled. The routes from place to place are not exact, but the general areas where the Sixth travelled up to the end of the war are taken from the history. Travel from the LeHarve to the training facilities around Chateau Villain would have been by the famous 40 men or 8 horse box cars.

    The Sightseeing Sixth Division in France in the Fall of 1918

    The Sightseeing Sixth Division in France in the Fall of 1918

    The other interesting thing that I learned was the extreme lack of transportation that troubled the Sixth Infantry’s movements. Horses were poor cast offs from the French army. Often the men hauled their equipment themselves, instead of relying upon horses or trucks. Once at the front, most travel was on foot. It is interesting that my Grandfather was a truck driver for a unit that apparently had very few trucks. The only mention of type of truck, was a brief mention of usage of some “liberty trucks”.