Digitally Archiving Sound Recordings and Moving Images at Stanford Libraries

The Stanford Media Preservation Lab (SMPL) serves to preserve and enhance access to original sound and moving image collection materials held by Stanford University Libraries. The lab’s core operations are centered around creating a high-quality copy of the original content in digital formats that are easily accessed by researchers and others, and that enable ongoing, long-term management of the content for future users. We work closely with curators, archivists, conservators and other staff responsible for the care and management of special collections and archival materials.

We also provide a variety of consultation services to the Stanford community to assist in the assessment and ongoing management of audio and moving image collections in analog and digital form.

SMPL is one of several digitization labs managed by Digital Library Systems and Services. SMPL was founded in 2008, and developed with the consultation services of AudioVisual Preservation Solutions.

SMPL blog

EIAJ Refurbishment Project: Cleaning and Relubricating our Sony AV-3650

Although much of our time at SMPL is spent digitizing and working with library collections, part of our work involves seeking out legacy equipment that can be refurbished and installed in our labs for use in our reformatting work. In 2011, we were fortunate to find a working ½” EIAJ machine for sale. Knowing that it would need some work before it could be used, it lay tucked away until last month when we received funding to start work on restoration.

Before committing time to this particular Sony AV-3560, we made sure that its capstan and reel motors were operational and that the video and audio heads were intact and functional. The tape path received a thorough cleaning with isopropanol before we threaded a piece of raw ½” tape stock as a test. Thankfully, the tape passed through the transport without stopping or creating any visible damage to the surface or edges of the media. The mechanical controls were stiff and slow to respond, but the function lever was operational and the machine would engage playback, fast-forward and rewind.

With confirmation of the basic mechanical function, running motors and a clean tape path, we were curious to see what kind of AV output the machine would show in its current state. Using an already digitized piece of ½” media from one of the library’s collections, we had our first real test of the machine. Although there were problems associated with tension, alignment and dihedral adjustment, both video and audio playback was achieved.


The Sony AV-3650 we’re refurbishing was likely built in 1970. As expected, it’s not easy to get a date of manufacture on equipment of this vintage, but the relatively low serial number on the machine indicates that it’s probably from an early production run.


In our first meeting with the project engineer, Ken Zin, we laid out a plan. Ken has a vast experience restoring videotape machines of all kinds, and he’s completed overhaul jobs on this particular model on numerous occasions. He was quick to point out that because the machine hadn’t been used in many years, the slow response was likely due to the old, dried up grease and dirt that had accumulated on the parts that engage the tape transport. Cleaning and relubricating the machine would be the first step in our process.


We identified some tools that we’d want to have handy before starting work, including our standard cross-head and flat-head drivers, hex drivers, hex keys, pliers and wire snips.


We needed acetone to remove the grime from the deck assembly and metal parts of the transport. A set of broad nylon and brass wire brushes, as well as smaller acid brushes would prove useful where there was heavy buildup.



Ken started disassembly by removing the base cabinet from the machine. The plastic vent grille taped to the bottom of the case was cracked and will eventually be repaired or replaced.


With the cabinet removed, the tenuous nature of the machine’s playback system is exposed. The main belt wraps around a pulley on the drum motor (the large, black plastic hub) and drives the video head pulley (the small silver hub), making for a delicate timing relationship between the speed of the drum motor and the movement of the video head. The tape counter belt can be seen stretching above in the background.


While removing the pinch roller, we noticed an interesting anomaly. The markings on the rubber indicate that this roller was made for use on a PAL standard machine as opposed to an NTSC standard machine. The 50 Hz roller has a slightly smaller diameter than the 60 Hz roller, and this difference may adversely affect playback of NTSC source tapes. Since we know that this machine’s electronics are built for use with NTSC standard recordings, this AV-3650 appears to have the wrong pinch roller installed. Luckily, we have a 60 Hz pinch roller salvaged from another machine. The top of the Oilite bearing holds a sponge material that functions as a reservoir for lubricant, but it has disintegrated and will need to be replaced. The hardened pinch roller will need to be rebuilt with new rubber before it can be used on our machine.



Removing the reel panel that covers the top of the machine exposes some of the mechanics of the tape transport. The reel tables, the idler tires and push rods are visible from this perspective.


This exploded view taken from the schematic show some of the parts that need to be removed in order to clean the base of the deck assembly, as well as some key mechanical parts. Harsh solvents ruin plastic and rubber components, so we cleaned them with distilled water and a solution made for use with rubber.


In the images below, shredded rubber is visible on the metal base of the transport. This is likely due to the idler tires grating against the textured metal motor pulley during tape rewind. This mess will be cleaned up using our acetone and brushes since it's very gummy and difficult to remove. Eventually, the alignment of the tires will be adjusted to make sure they function properly.


After removing the reel tables and rubber idlers, we disconnected the felt brake band that attaches to the supply reel. It is an integral part of the system that maintains tape tension during playback and record. Although discolored, the felt isn’t as worn as we had expected. If we decide that it needs to be replaced after testing our rebuilt machine, we will have to find one in good condition on another AV-3650, or attempt to rebuild the felt cloth strip from newly sourced materials.


Finding these barely existing parts is often cited as one of the major challenges of working on refurbishment projects. Some of the key components of this machine, such as the belts and the reel brakes, were manufactured with very specific properties. Used parts can sometimes be found, but they are often just as worn as the parts needing replacement.


We covered plastic parts with tape so that the acetone wouldn’t damage them. These pieces were easier to cover than remove, since we weren’t going to be using solvent near the VU meters, counter, edit/record buttons or plastic handles. We moved the machine outside for better ventilation and to keep the solvent from pooling on the floor of the lab.


After removing the shredded rubber with our broad nylon and metal brushes, we worked the acetone onto the joints of the arm and lever assemblies. The body was tilted so that the excess acetone would run down into paper towels that we stuffed around the frame. We removed old grease and dirt from push rods, cleaned the idler pulley, the cams, and all the metal surfaces above the base of the deck.


We spent about an hour cleaning the machine before we were ready to reinstall the parts we had removed and start relubrication. Ken supplied some silicone based grease as well as instrument oil, and we started application using wooden handled swabs.





Ken reattached the reel tables and idlers:


Small metal clips on top of the idlers hold them in place:




We tilted the deck upwards to oil the joint on the pinch lever under the plate that covers the capstan assembly:


After about four hours of work, the machine was cleaned, reassembled, and many parts were lubricated.


Before calling it quits for the day, we made sure our machine was still operational with our test tape and planned for the next stage of work. The video and audio connectors attached to the machine were outdated, and required adapters for cabling to modern equipment. We decided our next step would involve installing some new connectors.

Check back soon for more progress!


SMPL Commences with Grant Funded Refurbishment of 1/2" Videotape Recorder

In 1970, the emergence of a standardized 1/2" open reel video format changed the media landscape for documentarians, visual artists, educational institutions and home video enthusiasts. EIAJ-1 became the first widely used non-broadcast video format, and its prevalence has led to an abundance of inaccessible content held by many libraries and institutions.


In August of this year, the Stanford Media Preservation Lab was awarded special funding through The Payson J. Treat Fund for Library Program Development and Research. Over the next several months, the video lab will engage Ken Zin, a Bay Area video engineer and tape restoration specialist for the NASA Lunar Orbiter Recovery Project, in an effort to restore and optimize a Sony AV-3650 video tape machine and other necessary components for use in digitizing video materials from Stanford University Library collections, including the R. Buckminster Fuller Papers, the Philip G. Zimbardo Papers and many others. The addition of EIAJ-1 playback capability will enable SMPL to handle this nearly extinct video format in-house, reducing the cost of outsourcing and providing on-demand access to previously unavailable content. Along with digitizing some of the libraries’ most at-risk collections, the project will promote acquisition of new collections containing ½” video recordings.


We’ll be documenting our progress throughout the project in this blog. Stay tuned for updates!

Syndicate content
« Back