Slave recording deck with LAP Sidewinder (the sidewinder is in blue). The newly recorded tape streaming past the recording head was routed through a couple of spring-loaded feed arms. The hubs (which would later be inserted into the 8-track cartridge shell) were stacked face-to-face onto a drive rod (40-50 at a time) in the sidewinder. So how did these machines work?

Ed could not recall a specific number of plays that an 8-track was designed for, but agreed that it wasn't very many. He also volunteered that music was throwaway in those days. After all, (according to the music execs) who would be listening to today's "pop" star six months from now, so who cared if the media expired in the same time period? The same cheap-quest was the reason for the changes in tape, changes in case design, crappier and crappier foam pads, and ever-cheaper sleeve designs. He said that Ampex in those days was an extremely chaotic place to work and that it was not all that great a place to be a lot of the time. My recollections were more positive, but then I wasn't involved as closely in the politics and bureaucracy as he was.

OK, back to the production: The 8-track tape finishing process was almost entirely different from the other two media types and considerably more complex than either one. Instead of being re-reeled at the recorders, they were wound directly onto the cartridge hubs. Highly specialized devices called "LAP Sidewinders" were used for this. They were manufactured by Livingston Audio Products. The newly recorded tape streaming past the recording head was routed through a couple of spring-loaded feed arms. The hubs were stacked face-to-face onto a drive rod (40-50 at a time) and a clamp collar was clamped in place with a setscrew. The resulting stack was inserted horizontally into the machine (with the shaft facing front-to-rear).

A number of problems arose with this process that had to be solved. First, the winding speed of the spindle (rpm) had to vary widely from the start of the winding process to the end as the hub filled. Second, the tension of the tape as it was being wound was extremely critical. Too much and the tape would bind and wrinkle, too little and it would unwind from the cartridge during playback. But any true 8-track enthusiast already knows this, right? Thirdly, the linear speed of the streaming tape could not vary at all because the recording speed was constant. Fourth, the rotating stack of hubs could not be allowed to stop at any time in the process because the tape would go slack. Fifth, the hub stack had to have enough clamping pressure so that the hubs would not slip in the drive shaft. Sixth, the winding feed arm had to be positioned fairly precisely over the side of each hub during winding (hence the name "Sidewinder"). Seventh, all of this is happening at about 37.5 inches per second (10x) which means that as the tape is being wound onto the hub at its smallest diameter the hub needs to be rotating at about 800 rpm. Finally, during changeovers from one hub to the next, there would have to be an abrupt change in the drive rod speed to maintain tension, tape speed, and so forth, AND there would be a short loop of tape that needed to be pulled out rapidly (but not too far) as the tape looped over the edge of the previous hub. As you can imagine, I spent a lot of time working on these machines.

So how did these machines work? The operators would lift the floating arm almost straight up, load a shaft full of hubs, slide the arm to the front of the machine, and then thread the tape. Lowering the arm locked it into the detents on the ratchet rack and aligned the arm roughly to the middle of the hub's center (as viewed along the shaft axis). The tape would then be threaded through the floating arm (including the twist), and attached to the first hub by adhesive tape or by tightly winding a few turns (riskier but quicker). The tensioning arm would be manually set to the center of its travel by winding on the hubs and blank tape spool.

Assuming the master playback machine was already running, the operator would push the start button on the recorder. The recorder would begin feeding tape very rapidly and the tensioning arm would begin to move toward the front of the machine as the tape began to stream through. The servo motor would then accelerate briskly and the hub would begin to wind. When the pilot tone was sensed, the carriage solenoid would kick the floating arm up and move the carriage back to the next hub position. This happened extremely fast. The extra tape that was pulled from the tensioning roller and arm when the arm was kicked up would loop over the edge of the first hub and then start to wind around the root of the next hub. Since the tensioning arm had now been pulled toward the rear, the servo motor would immediately slow down and the tape feed would be allowed to catch up. This process continued until all the hubs were wound. If the feed tape ran out during the process, the tensioning arm would swing fully to the front and a switch would shut the recorder and Sidewinder down.

The whole mechanism was cleverly designed to have just enough slack loops and vertical play in the floating arm to accommodate these momentary disturbances during startups, stops, and hub switches. I never spoke directly with anyone at Livingston (I'm pretty sure they were all asleep when I was working), so if any of you are reading this, my compliments to you. Very slick design work. If I got a detail or two wrong, forgive me please, it has been a long time.

OK, that's how it was supposed to work. Now here's what happened when things went wrong. As mentioned, if anything happened to the feedback loop on the tensioner arm, at about 38 inches per second tape speed, things went wrong very quickly resulting in stretched tape, overly tight windings, or piles and piles of tape streaming onto the top of the rapidly rotating hub stack, the floor, the next recorder in line, and so forth. This happened often enough to keep me fairly busy repairing things. If the tensioning arm was misadjusted the floating arm wouldn't ride at the right height causing winding tension problems, or dragging the arm's tip on the tape as it was being laid on the hub, or allowing the tape to wind so loosely that when the hub switch happened a huge loop of tape was pulled out causing a big mess as it snagged on the floating arm. And on and on. If the servo gain was not adjusted properly all the same things could happen, or the hub stack speed would oscillate causing uneven packing and crinkled tape. The hubs could slip on the drive rods, again causing a whole host of problems including great sags of tape dropping to the floor. In other words, someone had to watch these machines virtually all the time. Many of these occurrences would generate "interesting" noises that would definitely get your attention.

The first hub loads of tape were always scrapped (if the operators weren't too busy, they unwound them and reused the hubs, otherwise they went into the dumpster). This was because there was no way to tell where in the playback loop the recording had actually started. It was difficult or impossible to synchronously start the recording with the playback because of logistics and because of the "whoop" you'd get on the finished product as the tape came up to speed. So the first one was always trashed.

Back to Page 3 | To Page 5

Back to 8-Track Heaven