How we restored our Apple 1 to working condition…

…and then got Steve Wozniak to power it up for the first time!

by Ralph Simpson, History San Jose volunteer

The Apple 1 was first sold in May, 1976 and was the first personal computer as we recognize it today. Before the Apple 1, personal computers used toggle switches for input and lights or printers for output. The Apple 1 brought the innovation of a keyboard for input and a TV screen as output. The original sales price of the Apple 1 was $666.66, which included a motherboard with the chips already installed for an operational system. The TV monitor, keyboard, transformers for power supply, mounting board, cassette tape player and interface board were all provided by the user. Later, the cassette interface board was made available from Apple for an additional $75. 200 Apple 1s were manufactured and 175 were sold. Of these, only 46 were known to still exist and only 6 had been reported as operational. (This list would quickly change in June).

History San Jose is fortunate to possess a rare Apple 1 computer as part of the Perham Collection of Early Electronics. The collector interest in the Apple 1 was always high but seemed to skyrocket after Steve Jobs’ death in October of 2011. The three most recent auctions of a working Apple 1 shows a dramatic rise in prices: November 2010 for $210,000, June 2012 for $374,000 and Nov. 2012 for $640,000. Meanwhile, in October 2012 there were no bidders for a non-working Apple 1 with a starting price of “only” $80,000. Apparently, collectors placed a very high value on whether the Apple 1 was operational. We did not know if the museum had a working machine but now finding the answer to that question became more than an academic exercise.

May 23, 2013
Another working Apple 1 was scheduled to be auctioned on May 25, 2013, which had wide press coverage and speculation on the final auction price. It was in this setting I had asked Jim Reed, History San Jose’s Curator of Library and Archives, for permission to power up the Apple 1. On May 23rd he agreed and two days later the auctioned Apple 1 set a new record price of $671,000.

May 24, 2013
To power up our Apple 1, I wanted to enlist the help of some experts, so I emailed the owner of the Apple 1 registry, Mike Willegal, for his advice. Also, the museum’s Apple 1 was not included in the registry, so I sent some pictures to verify our Apple 1 as the 47th known Apple 1 in the world. Mike recommended two Apple 1 experts, Dr. Wendell Sander and Daniel Kottke. After contacting them by email and phone both agreed to help and suggested we include Allen Baum in our project. We scheduled our first meeting for May 30th.

May 30, 2013

We had our first meeting at History San Jose to examine the Apple 1 and plan for the testing, restoration and press release of the Apple 1. The 3 Apple 1 experts surpassed my wildest dreams, both technically and historically.

Dr. Wendell Sander was Apple employee #16, with 97 patents to his credit and counting. He is one of only 2 original owners of an Apple 1 and his was one of the 6 working examples. He also had an Apple 1 replica, which was invaluable in testing the components of the museum’s Apple 1 and having known working parts to swap with suspected faulty parts.

Allen Baum is a microprocessor architect with 26 patents to his credit and a high school friend of Steve Wozniak. He is also an owner of a non-working Apple 1 he received from his father, who was Apple employee #34. Allen brought his Apple 1 to the museum to do testing and repairs while we worked on the museum’s Apple 1, often using his computer as the guinea pig for our testing.

Daniel Kottke was a college friend of Steve Jobs who was employed by Apple in the summer of 1976 to plug in and test the chips in the Apple 1s, along with Steve Wozniak, Steve Jobs and sister Patty Jobs. He joined Apple full-time in 1977, becoming Apple employee #12.

We set a lofty goal for our project, which was to demonstrate a working Apple 1 by powering it up for the first time in a press event. We needed to come up with a plan to test as much as possible without turning on the entire system, while minimizing any chance of damage to the Apple 1. I certainly did not want to be responsible for frying our Apple 1!

The first thing we did was to examine our Apple 1, noting some of the unique items and any obvious problems. The HSJ Apple 1 was originally owned by Winston Gayler, who later wrote a book on the Apple II circuit design, so he was obviously quite a technically advanced user. His Apple 1 was evidence of his expertise, which included several user modifications. He seemed to be very cautious of heat on the motherboard, adding a homemade strip of metal to the surface of the 4 eproms as a heat sink. He also added a rectifier bridge with a huge heat sink off of the motherboard. His transformer and fuse set up was also unusual and was identified as a possible problem area, since this would be one of the few things that could cause real damage to the Apple 1.

The museum’s Apple 1 did not have the “NTI” logo on the motherboard, which means it was made from the first batch of 100. It had the white ceramic 6502 microprocessor with a date code of 1576, meaning it was manufactured in the 15th week of 1976 or the first week of April, which was coincidentally the same date as the incorporation of Apple Computers. More rare was the white ceramic Peripheral Interface Adapter (PIA) or 6820 chip. There is only one other known Apple 1 with this chip made of white ceramic, and it is owned by Liza Loop. Liza’s Apple 1 was the very first Apple 1 manufactured. Liza was a school teacher who wanted to use the Apple 1 in her classroom, so Steve Wozniak paid for it personally and gave it to her. She still owns this Apple 1, making her the only other original owner of an Apple 1 along with Wendell. The original advertising photo of the Apple 1 also shows a white ceramic PIA chip, but there is no record of what became of this computer. Another unusual feature of this Apple 1 was the two 7400 logic chips, which were military grade rather than commercially available chips. The military version is normally very expensive, but this may have been supplied by the distributor at the time because they were available to fulfill the order from Apple Computers.

So the first important decision to be made was what to do with the unusual power supply. One option was to remove the entire power arrangement and restore the board to the way it was originally shipped. This way we could install a known, working transformer to ensure clean power and minimize the risk of damage to the motherboard or any chips. While this would be the safest approach, our guiding principle was to keep the board as it was originally donated to the museum, including all the user mods. The Apple 1 was a hobbyist machine and most were modified in some fashion, so maintaining that history is important. So we decided to test the motherboard with the user modified power set up but after the removal of all the chips.

Our test plan was to remove all the chips, cleaning each one with contact cleaner. We then would test the power for the correct voltages, which are ą12V and ą5V, without any chips on the motherboard. We would then test the chip individually, using Wendell’s Apple 1 replica. After testing the motherboard as much as possible, we would then test the Radio Shack keyboard, cassette interface and a TV display from correct time frame.

We also discussed which applications we should display for the press event. The Game of Life and Star Trek were selected as two games popular at the time. So we concluded our first meeting without any testing but had a good plan of attack for the testing and power up of the Apple 1.

June 4, 2013
We had our first hands-on work, removing all the chips, placing them on a foam board in the same physical position as on the motherboard and meticulously cleaning the legs of each chip with contact cleaner. Some were coated with a film of corrosion, which is a commonly reported problem with the Apple 1. Usually, removing and re-seating the chips would fix this problem but we took the extra step of thoroughly cleaning each chip. We then tested the power, which looked good, so we powered up the naked motherboard. There were no problems and the right voltages were at the correct contacts, so our fears of problems with the power supply were relieved.

Allen did the same with his Apple 1 and then inserted the top 2 rows of chips which control the display and his display worked immediately. After we repeated his test, our display did not work. Instead of a matrix of at signs “@” on the display, we got a series of alternating white and black stripes. We swapped a few chips which were obvious candidates for this type of failure, using chips from Wendell’s Apple 1 replica, but these did not fix the problem.

After Allen’s Apple 1 displayed correctly it then ran into problems when he installed the remainder of his chips. So we ended our first day of testing with both Apple 1s inoperable and some research and study to do before our next test.

June 6, 2013

This was an exhilarating day of major progress and success. We got Allen’s Apple 1 working today, making it the 7th working Apple 1 on the planet. The debugging effort took several hours of swapping known, working chips until the defective chips were identified. Two memory chips and a 74154 decoder chip were replaced, after which his Apple 1 worked flawlessly.

We also debugged the museum’s Apple 1 display problem and after testing several chips discovered the faulty chip. It was the 9316 chip in position D8 on the board. We replaced it with a correct date code chip (7706) which was donated by Wendell and the display worked. We then tested the rest of the board, but using the microprocessor and PIA chip from a known working replica and that worked correctly also. At this point we felt we could have swapped in all of the original chips and make two Apple 1s operational in one day, but kept to our plan of waiting to power up the museum’s Apple 1 on press announcement day.

We also tested out the cassette interface board, using a Panasonic Slimline model RQ-2102, which was a common cassette player used with the Apple 1 in the late 70s. The cassette tape was notorious for having problems, but we found it worked correctly by setting the volume level just below the highest volume. We then tested the original white ceramic microprocessor chip and PIA chip in Wendell’s replica and that was also successful. Since the legs of these chips are very fragile, we set these aside and did not use them or test them further until we plugged them in for the final power up. At this point we had everything working except for the Radio Shack keyboard, which did not work at all. We elected to call the day a success and come back another day to work on the keyboard.

Since we were close, we also discussed ideas for the press announcement. We decided to have 3 Apple 1s at the event, Allen’s and Wendell’s both work and if the museum’s powered up successfully, we would have 3 of the now 8 working Apple 1s on display. We decided that Wendell would run Star Trek, Allen would run Game of Life and the museum would run an application that simply displayed a rolling set of images, starting with Woz, Steve Jobs, the Apple logo, the Apple II, etc. We also thought this would be significant enough to interest Woz in coming to the event, so Allen called him to make the invitation. Woz immediately accepted and we arranged a date to meet with his schedule. We now had a set date for the announcement, June 18th.

We invited Mike Cassidy from the San Jose Mercury News and Peter Jon Shuler from KQED radio for the press event. Other museum staff and invited guests added up to a guest list of about 40.

June 11, 2013
One week before the press announcement and the keyboard is still not working. We tested the motherboard using a replica keyboard and this worked correctly, so we knew we had a fallback plan if we were not successful with the original keyboard, but the original keyboard with the homemade clear/reset button and pine wood board definitely had the coolness factor. This keyboard had an unusual and extra plug arrangement, with an extra-long keyboard cable. The usual Apple 1 keyboard connected directly into a chip on the motherboard.

After cleaning up this plug and the plug going into the board on the keyboard, we were finally able to make the keyboard work, with 2 exceptions. One problem was that the first key entered from the keyboard was ignored. This was more a nuisance, but the second problem was a real issue. Pressing the “Cntl / M” keys was supposed to create a carriage return, but this did not work. Through trial-and-error we discovered that “Shift / M” worked. So now we had a working keyboard. After some more testing, we found the keyboard worked for a while then would intermittently fail. After re-seating the plugs and once re-seating a chip, we got it working again but felt the problems were too intermittent to rely on for the press announcement. Instead, Wendell decided to fabricate a new connector which would go directly from the keyboard to the chip on the motherboard, bypassing the extra keyboard plug and long cable. This way we could use the original keyboard and have a more reliable connection.

Up to this point we had done our testing with a modern LED monitor, which worked well, but we wanted to use a 9-inch black and white television from the mid 1970s instead. We tested several monitors, including an Apple II monitor, but they all had problems with the vertical hold. This problem only occurred after the screen became full.

June 13, 2013
We tested several monitors but continued to have the same problems. We also exercised the keyboard, cassette tape player and the rolling image application to make sure they still worked properly. Everything seemed to work well except for the display. Even the keyboard worked well with the new plug arrangement. Remember, we still have not connected the microprocessor or PIA chips into the motherboard. We decided to come back for final testing with an Apple III monitor, if that didn’t work we would be forced to use a modern, flat screen LED monitor with this ancient Apple 1.

June 14, 2013
We tested the Apple III monitor, and it seemed to work a little better but still had the vertical hold problem with a full screen. Wendell knew the Apple 1 was susceptible to this type of problem because of a rather weak resistor in the display circuitry. This was a 3K ohm resistor which we piggybacked with another 3.3K ohm resistor, giving it a better chance to work with a variety of monitors. After testing out this new additional resistor, everything worked flawlessly. We tested the rolling screen application and everything worked as planned. At this point we left the Apple 1 set up on the table, not to be moved until the press announcement.

June 18, 2013 – Press announcement and Apple 1 power up by Steve Wozniak

Today is the big day. We arrived early to set up the 3 Apple 1s and test out the applications on Allen’s and Wendell’s machines. We did not test out the museum’s Apple 1 but finally installed the original microprocessor and PIA chips for the first time since we started this testing. One of the guests, Andy Jong, surprised us by bringing his working Apple 1. We did not have time to set it up with a keyboard and monitor, but displayed it on the same table. Then Steve Wozniak and his wife Janet arrived on their Segways, and Woz decided at the last minute to bring along his Apple 1. We now had 5 Apple 1s on display!

Woz agreed to power up the museum’s Apple 1 but wanted to do it jointly with Wendell Sander. First the Apple 1 “@” matrix displayed, showing that the display monitor worked, which included about half the chips on the Apple 1. Next, the carriage return was entered and the Apple 1 prompt was displayed, showing that the microprocessor was working. This was a huge relief. Next, the cassette tape was used to load the rolling image demo, which worked and was displayed for the rest of the event. The other two working Apple 1s were also successful in displaying Star Trek and Game of Life. Watch as Woz and Wendell power up the Apple 1:

The 3 working Apple 1s showed the varied history of these machines. Allen’s machine was completely unmodified and appeared to have never been mounted on a board. This may have been one of the 25 unsold boards that his father acquired while working at Apple. As described earlier, the museum’s Apple 1 was heavily modified by Gayler in the 1970s and early 80s. As Woz later pointed out, the Gayler unit was “personalized,” as it was meant to be. It had remained as is since its donation, however, until we began testing. Wendell’s Apple 1 was continuously used through the years with updates bringing it into the 21st century. It has a flat screen LED display, an iPod interface instead of the cassette tape, and a custom-made clear acrylic case.

This was an historic day. Not only did we have Woz start up the museum’s Apple 1 for the first time in decades, we had 5 Apple 1s on display on one table. At this point our Apple 1 became the 8th working example, which also means we had exactly half the world’s supply of working Apple 1s on display. Ironically, the only Apple 1 not working belonged to Woz. He wisely left it with Wendell to do his magic and restore it to working condition.

Woz and Wendell signed and dated the Apple 1 for the museum and for Allen Baum. Wendell’s Apple 1 was previously signed by Woz, with the inscription, “To Wendell, a most incredible engineer! Apple would not have happened without you!”

Woz spent the next couple of hours at the museum, reminiscing about the Apple 1 and the early days at Apple Computers. Although he had designed many personal computers on paper, he had never built one because of the cost. When he started designing what later became the Apple 1, he did not plan on it being a computer at all. He was designing a terminal to be used with DARPANet, the predecessor to the internet. He wanted an inexpensive way to use a keyboard and TV as the display for use as a computer terminal. Then MOS Technology startled the industry with a $20 microprocessor, the 6502. At the time, the Intel 8080 was priced at $400 if you were an individual buying just one chip. So Woz bought this MOS chip and redesigned his terminal to become a personal computer. He said his motivation was better than any salary or stock option, it was a passion to have his own personal computer.

His original plan was to have a PCB manufacturer build a batch of 100 board at $20 per board. He and Steve Jobs planned to sell them for $40, so the break even point was only 50 boards. They went to the Byte Shop to see if they were interested in selling these boards, but the Byte Shop insisted on selling the boards complete with chips already installed, so Woz set a price of $666.66. The rest, as they say, is history…

Read Mike Cassidy’s Mercury News story about the event.