Side projects and fun

USED SOME NEW-OLD STOCK CHIPS WITH SIGNIFICANT PIN TARNISH  One of my objectives building the EDUC-8 was to leverage the large supply of chips I had, resulting in a low cost for the project. Quite a few of the chips I have had developed major tarnish on the pins. They tested fine on the Retro Chip Tester Pro so I decided to put them into the PCBs. I had some incorrect behavior and once I began to scope lots of signals, I found some that were not correct as generated by tarnished chips. The chip tested good beforehand, so it has to be the solder joint itself.  DISCOVERED THAT THEY WOULD NOT SOLDER RELIABLY I had attempted to overcome the problem by flooding them with flux and heating them longer than usual. Some of the joints looked a bit suspect but I stuck to my guns because I still have the low-cost objective in mind. What occurred in the solder joint was great flow and surface tension to the pad of the board but a very thin layer of oxide insulating the pin itself from the ...

LOADING AN ADDRESS OF 00000000 LED TO 01000000 IN THE PC REGISTER The Program Counter (PC) is implemented by two shift registers which use serial load to clock in incremented or calculated addresses as well as parallel load to set the PC to a specific value.  The Load Address switch engages the clear function of the two shift registers when it is depressed and then does a parallel load from the switch register (the toggle switches on the front panel) as it is released. Holding down Load Address, I could see that the PC was all zero, and my toggle switches were all set to zero, but when I released Load Address the address show was 01000000 instead of all zero. UNIQUE ASPECTS OF ADDRESSES DUE TO EVOLUTION OF THE EDUC-8 From the time that the Educ-8 was first conceived and written about in Electronics Australia until the series of construction articles were published, the designer expanded the machine from its original limit of 32 words up to 128 words, with provision for 256 using ye...

VERIFYING THE DECODER CARD PRODUCES THE RIGHT OUTPUT I tacked wires on the motherboard pins for all the outputs from the decoder and verified that every instruction is properly decoded. There are eight major instructions (AND, TAD, DCA, IOT, OPR, ISZ, JMP and JMS). The OPR (Operate) instruction has subtypes which are decoded and I validated that these were correctly produced for all the combinations of five bits in the OPR instruction.  ASIDE ON ODD OCTAL NOTATION All the documentation for this machine, based as it was on the PDP-8, uses octal rather than hexadecimal. With eight bits, not an even multiple of 3, one has to decide how to express the numbers. With most octal machines, one presumes there is a high bit of 0 which lets us write octal ranging from 000 to 377 to cover all eight bits.  Not so with the EDUC-8. Its 'down under' notation presumes a bit of 0 inserted between the third and fourth bits. Thus if we have binary 10110111 it is written as 527, distinct from the ...

NORMAL BEHAVIOR OF THE DECODER BOARD The board should emit a signal for each of the eight instruction types but only during the Execute phase of instruction execution. The top three bits of the instruction are the instruction code and a 74LS259 chip is used to decode this to produce one and only one output. However, instead of latching and demuxing, the chip is used in a different way. At the beginning of the fetch cycle, at time T0.5, it activates the clear pin which puts zero in all eight output latches. The three instruction bits are fed to the three selection inputs of the chip, which pick one of the eight output latches.  Near the end of the fetch cycle, at time T22.5, it activates the enable pin which will transfer the value on the D input pin into the selected latch (out of eight latches). The D pin is wired to logic high, thus this turns on the bit in the selected latch. The three high bits of the instruction should be stable thus the chip is selecting just one latch. The l...

GOT SLIGHTLY WRONG CABLES BUT TURNED THEM INTO PROPER PARTS I received a link to AliExpress items to finish the EDUC-8, which I expected were premade cables, ribbon cables with connectors on both end, in the 2, 3 and 5 pin widths. What I received were half cables - connectors with a short ribbon cable with bare wire ends. They were also quite short, so even if soldered together they wouldn't have had the 10" or so length to fit in the computer. These cables are all additions that were added to the original EDUC-8 design in order to increase the memory size. The first couple of articles in Electronics Australia where the project was described had only a 32 byte memory, but by the third article where construction began, they had expanded it to 128 bytes with wiring in place for a future update to 256 bytes.  This was due to the rapid decrease in memory cost and increase in memory chip size. At the time of the articles starting in August 1974, memory cost was very high and thus t...

WIRING THE EIGHT TOGGLE SWITCHES TO THE MEMORY BUS FOR TESTING The memory card is not yet installed, but by wiring the toggle switches to the memory bus I can set up the data as if it came in from a memory read. That is the way that the author/designer of the EDUC-8 recommended testing the functionality.  FIRST ISSUE - METAL LOCKWASHER AND THE GROUNDED SIGNAL The connectors that the cards plug into are screwed down on the motherboard. I used some metal screws, nuts and lockwashers. When I began testing with the Timing, Decoder and Accumulator cards plugged in, I found that switch 4 was not working - there was always a ground as the state of the memory bus in that position.  When I pulled the boards, the short to ground was still there. I pulled the suspect connector off the motherboard after removing the hardware, and the short was gone. I soldered down the connector, but just the pin that had the problem, so that I could keep checking to be sure the issue didn't come back. Wh...

SOLDERING IN COMPONENTS TO COMPLETE THE EDUC-8 COMPUTER The main connectors are a mix of four 36x2 pin and eight 18x2 pin connectors, to cover six plug-in cards onto the backplane. That is 576 pins that must be soldered to the backplane. I have about 144 left to solder after I ended the workday.  I also installed ten SPDT on-on switches and five SPDT momentary on switches, plus another SPDT on-on switch of a different design. Each of those has three terminals to solder. Only 48 solder joints, small potatoes compared to the connector soldering I  did.  I have some prewired cables that will support some board to board signals that are part of the design. No work needed to get these ready. 

LAST CHIPS ARRIVED AND WERE INSTALLED ON THE BOARDS The boards are waiting on the decoupling 47nf capacitors otherwise all components are installed other than connectors and switches. The connectors, cables and switches are coming from China as I bought them on AliExpress. The estimates for arrival are June 14, June 17, June 20 and July 1 for various items in the order.  INITIAL TESTING OF FRONT PANEL AND TIMING BOARD I was able to test all the LED driver circuits on the front panel, switching them on or off by grounding the input pins on the edge connectors. Most of the switch circuits are not debounced thus there is almost nothing to them. I verified that, once the switch is installed and operates properly, the outputs are changing appropriately.  A few of the switches are debounced, using this type of circuit: Two one-shot gates are used to convert certain switches to a single short pulse per activation, using a circuit like this: I didn't test this either. It is possible t...

SHORT A FEW INTEGRATED CIRCUITS AND CAPACITORS I was missing a 74LS154 decoder (4 input to 16 output), a couple of 74LS07 buffers to drive LEDs, and a smattering of 74LS00 (quad NAND) and 74LS04 (hex inverter) parts. They are arriving from Digikey except the nearly obsolete 74LS154 which is coming from eBay.  The design uses plenty of 47nf decoupling capacitors on the PCBs and I didn't have a large enough supply. They will be arriving by this weekend and get installed.  CONNECTORS, CABLES AND SWITCHES ON ORDER The front panel uses fifteen paddle switches, five of them momentary contact type. In addition, there are plenty of cables running between boards and to the rear panel where the peripheral device connectors are placed. I ordered these from AliExpress and hope they won't take too long to arrive from China. Similarly, the PCB sockets that fit on the motherboard were ordered from AliExpress.  BOARDS PUT TOGETHER EXCEPT FOR THE MISSING PARTS MENTIONED ABOVE As you can s...

VERY EARLY DESIGN PUBLISHED IN ELECTRONICS AUSTRALIA MAGAZINE This computer is a stripped down version of a DEC PDP-8 minicomputer, comprised of about 100 integrated circuits (and no microprocessors). It is bit serial, like the PDP 8/S which itself was bit serial. At the time in 1974, the parts for an EDUC-8 would cost roughly $300 assuming that the reader made their own PCBs and bent sheet metal for the case.  Radio-Electronics magazine in the US published their design for a Mark 8 one month prior to the first article for the EDUC8. The Mark 8, however, was built around an Intel 8080 microprocessor. While the Mark 8 did not gain much attention, it did inspire the publication of the Altair 8800 in Popular Electronics in January 1975, kicking off the personal computer era.  NEW PCB DESIGNS AND UPDATES BY ENTHUSIASTS STIRRED MY INTEREST Gwyllym Suter has created KiCad files for the boards and ordered a batch of them. I bought one copy from him as I believe my parts cabinets have...