Background

The KENBAK-1 was developed by John Blankenbaker (from whose surname it takes its name) in the early 1970s as a teaching tool. The guiding criteria were that it should be low-cost, educational and able to satisfy users with simple programs. It could be serial and slow, which would reduce costs and yet create the desired environment. It would have to handle as many programming concepts as possible. Due to its small size, the native language would be machine language. Above all, it had to be a stored-program computer (Von Neumann machine). The aim was to design a simple computer.

Designed before microprocessors were available, its logic consisted of 132 standard TTL integrated circuits mounted on a single PCB. Serial memory was implemented using 2 MOS shift registers of 1024 bits each, for a total of 2048 bits. Input was made via switches and output was displayed via LEDs. Although the computer had a clock of 1 MHz, it processed only a few hundred instructions per second because it had a serial architecture, so it processed only one bit at a time.

Critical issues

• PCB has to be made starting from photos: high-resolution ones are available but with components mounted, those without components are low-resolution ones. We’ll try to overlap them to get the complete traces.

• There is one component (1404a) quite rare, and 4-5 others of medium rarity.

• Case dimensions have to be deduced from indirect informations, such as photos and proportions compared to other objects whose real dimensions are known.

• Need to reproduce the mechanics.

Realization

The realization of the printed circuit was made possible thanks to the patient work carried out by Stefano Ferilli (printed circuit design) and Gianfranco Mazzarello (production of the gerber file essential for printed circuit production). Subsequently, the team composed of the manager, Michele Perniola, and Stefano himself, moved on to the realization of the prototype. The realization of the front panel was made possible thanks to the electrical components and the metal hexagon made available by Stefano. The realization of the case was made possible thanks to the commitment and competence of Vito Fariello.

Michele Perniola, the Project Manager

Here are some photos that show the work done.

Beginning of the inspection with a microscope

Even the 3D printed keys are ready

On 30.11.2021 the prototype of the front panel was made.

On 23.12.2021, thanks to Vito Fariello, the prototype of the case arrives at the headquarters.

December 30, 2021 update.

Procede l’assemblaggio del prototipo da parte del nostro Michele Perniola/The assembly of the prototype by Michele Perniola is proceeding

Test dei circuiti integrati/Integrated circuit testing

Alcuni risultano buoni con un tester ma con l’altro no/Some components look good with one tester but not with the other

Comunque non sono stati trovati circuiti segnati difettosi da entrambi i tester/However, no circuits marked bad by either tester were found

 

Per cui si è proceduto al montaggio/So the assembly proceeded

Here is the result of the assembly. There are only 4 ICs to go.

January 15, 2022 update.

Nuovo pannello frontale in alluminio fornito da S. Ferilli dopo la foratura/New aluminum front panel supplied by S. Ferilli after drilling

Pannello con led e pulsanti montati/Panel with led and mounted buttons

Prima parte del cablaggio/First part of the wiring process

La squadra al lavoro!/The team at work!

January 24, 2022 update.

Paper labels were attached with transparent magic tape. It’s a temporary solution we can remove without causing any damage to the panel.

The key caps made by Vito Fariello were fitted.

Purtroppo il pulsante sottostante ruota liberamente con zero resistenza/Unfortunately the button underneath spins freely with zero resistance

January 31, 2022 update.

Front panel wiring.

April 13, 2022 update.

First power up.

Manca una pista sul circuito stampato/A lead is missing on the printed circuit board

Alcuni transistor sono errati (2n4403 al posto di 2n4401) e tutti i transistor hanno i piedini “invertiti” rispetto al layout “normale”/Some transistors are bad (2n4403 instead of 2n4401) and all transistors have “reversed” pins from the “normal” layout

I condensatori del multivibratore del clock sono fuori tolleranza e di valore errato (1.5 nF nominali, uno 2 nF e l’altro 2.5 nF)/The clock multivibrator capacitors are out of tolerance and incorrectly valued (nominal 1.5 nF, one is 2 nF and the other is 2.5 nF)

Sostituiti e “girati” i transistor, sostituiti i condensatori/Replaced and “turned over” the transistors, replaced the capacitors

Il multivibratore del clock oscilla correttamente/The clock multivibrator oscillates correctly

Fase di clock T7, T6, T5… corrette/Clock phase T7, T6, T5… are correct

Fase T1 in corto/Phase T1 is shorted

Short circuit hunt on the printed circuit board.

… there is no short circuit. Unfortunately SN74LS54 ICs have been supplied instead of SN7454 – this is one of the few cases where the ’74LS’ series IC is not a valid substitute for the corresponding series ’74’ as they have a completely different pinout 😖 – We had to order correct ICs:  they should arrive in the next few days.

Alimentatore per +5V/Power supply for +5V

Alimentatore -12V/Power supply for -12V

Manual and diagrams ready for debugging

April 20, 2022 update.

“Memories” (shift register 1024 bit SY1404A) have been installed.

An error has been identified on the printed circuit under the integrated n.104. The socket has been removed, the track has been cut and a new jumper has been installed (next to the one previously made to replace the missing track).

April 24, 2022 update.

Unfortunately, the very difficult acquisition of the printed circuit design “didn’t come out well”, especially on the component side where the only good photo available was… with the soldered components.
About 16 errors were identified considering both missing and incorrectly connected tracks.

The tracks were temporarily “fixed”. Some fixes were done by unsoldering sockets to cut bad tracks underneath.

The behavior of the lights/buttons now feels “more natural”.

The next step will be the verification of the test patterns reported in the Kenbak manual.

May 26, 2022 update.

The Kenbak-1 prototype is finally functional. It has been tested with an appropriate program included in the manual, which results in the scrolling of the indicator lights.

June 2, 2022 update.

We move on to painting the case.

Stages of assembly.

The adhesive printed panel arrived.

Final stages of assembly…

Test video of completed prototype behaviour.

Photo of the assembled prototype.

Condividi su ...

Facebook

email

Twitter

Pinterest