Sticky-Shed Tape: Investigating the Role of Backcoating

John K. Chester 2005-06-13

I have often heard speculation that sticky-shed is caused by some sort of interaction between the backcoating and the oxide coating, but I've never seen any proof of that. The experiments described here were designed to show whether such an interaction accelerates the return of stickiness after a tape has been baked. It is of course possible that something has migrated from the backcoating to the oxide coating (or vice-versa) and thus permanently altered its characteristics. If it were possible to go back in time and remove the backcoating from a freshly manufactured roll of tape at the date when my samples were manufactured, the test outcome might be quite different. However, these results are potentially of interest to those who are seeking to preserve existing tapes after they have been baked and transferred.

I trust that in the future I will never see the need to perform these tests on a roll of tape manufactured in this century.

Tape samples used:

1) Ampex 407, two reels, on 7" plastic reel in generic white box.

Transferred from pancake to reel 1983-02-23. Probably purchased a month or two prior to that date. Good library wind. It's always been stored in a human-habitable environment (no attics or garages). It has serious sticky-shed.

2) Scotch 209, on 7" plastic reel in original Scotch box, no plastic bag.

No date. Probably purchased between 1975 and 1985. Good library wind. Same storage history as Ampex 407. No sticky-shed. Very minor shedding of dry oxide. Minor shedding of dry material from backcoating.

3) Scotch 207, on 7" plastic reel in plastic bag in Scotch plastic library box.

No date. Probably purchased between 1974 and 1980. Good library wind. Same storage history as Ampex 407. No sticky-shed. Moderate shedding of dry oxide (noticeable deposits on guides after one play at 15 ips). Minor shedding of dry material from backcoating (slightly worse than 209).

Differences between Scotch and Ampex tape

Q: Can the backcoating be removed with isopropyl alcohol?

A: No for the Scotch samples. Yes for the Ampex 407 -- vigorous rubbing with a Q-tip saturated with isopropyl will remove the backcoating.

Apparently the Scotch and Ampex backcoatings are different.

Q: Can the oxide be removed with isopropyl alcohol?

A: Yes for all samples.

The Scotch oxide comes off quite easily. The Ampex oxide resists a bit longer, but can be removed by vigorous rubbing with a Q-tip saturated with isopropyl.

It appears that on the Ampex tape, a similar binder is used for the oxide coating and the backcoating. On the Scotch tape, the backcoating binder is evidently quite different from the oxide binder.

Test method

I needed a better method of testing shedding than just playing the tape and looking at the tape guides afterward (or observing that the tape was so sticky that the machine couldn't pull it). Here's what I came up with:

1) Remove the headblock from an Ampex 440c.

2) Cut a sheet of copy paper into 1" x 3" strips.

3) Tape a paper strip to a 1-5/8" diameter pill bottle.

4) Hold the pill bottle against the 440C top plate, pressing it into the tape so tape wrap on the paper strip is about 5/16". Mark the top plate with one location for testing the oxide, and another for testing the backing, so I can easily put the bottle in the correct location for subsequent tests.

5) Run tape at 15 ips across the test strip for one minute.

One test strip can have an oxide test on the left side, and a backcoating test on the right side. (Some early tests were done the other way around.)

Someone else using a different type of paper would no doubt get slightly different results which could not be directly compared to mine. As long as I'm using the same type of paper, this test does seem quite repeatable. I think it is good enough for monitoring the behavior of test samples over several weeks.

Test results to date

I baked one reel of Ampex 407 at 127 degrees F for 8 hours. This completely eliminated the sticky-shed. After baking, the 407 had the least shed of any of the samples.

I divided the baked reel of 407 into the following samples:

1A 700', no treatment

1B 200', stored interleaved with clear plastic leader

1C 150', backcoating chemically removed

1D 700', no treatment

1E 150', no treatment

Working with sample 1B is a real pain. I use a second tape machine to take up the clear leader when rewinding the test reel. The second machine then provides gentle tension on the clear leader as it's drawn back onto the takeup reel when the sample is played. This is the only way I can think of to retain the backcoating, and store the tape so that the backcoating is not in contact with the oxide coating.

The 407 was baked on 2005-05-24, and initial tests were done on 2005-05-25. Subsequent tests were done on 2005-06-04 (10 days later) and 2005-06-12 (18 days later). Subsequent tests will be once a week.

I've scanned my test strips. The scans show color & density, but they don't show whether the surface of the deposit on the paper is shiny. Tests with the Scotch tapes leave a deposit which is not very shiny. Some of the deposit is dust, which falls off when I tap lightly on the test strip. Unbaked 407 leaves a shiny deposit on the paper. Sort of looks like little lumps of goo have been spread across the strip in the direction of tape motion. Backcoating sheds slightly more than oxide.

2005-06-12 test results of 407 samples:

Sample 1C (backcoating removed) is still playable. There is a noticeable oxide deposit on the test strip, but much less severe than full sticky-shed.

Note on display of test strip scans: Two images of the test strip are shown side-by-side. The left image is exactly as scanned. Contrast of the right image has been enhanced by 50%, so that faint deposits on the test strip will be more visible.

Sample 1E has serious sticky-shed, almost as bad as prior to baking. It can't be rewound through the normal tape path.

The color of the oxide deposit on the 1E test strip is mostly brown, and the color of the backcoating deposit is mostly black. The color of the oxide deposit on the 1C test strip is somewhere in between the colors of the oxide and backcoating deposits on the 1E test strip. But, there's not supposed to be any backcoating left on the 1C sample. Could it be that there are sticky pieces of backcoating embedded in the oxide?

Sample 1A also has sticky-shed, but not quite as bad as 1E.

1A and 1E are from the same reel. 1E has much less tape on the sample reel. The 1A test was done by rewinding the tail of the sample reel just enough to do the test, so probably the tape pack at that point is not as tight as for the 1E sample. I can't think of any other difference, except of course that both samples come from different points in the original full reel.

For comparison, here are the test strips for the second reel of 407, which was not baked. (This was before I figured out that I could put the oxide and backcoating tests on the same strip of paper.) These tests were done on 2005-05-25.

The sticky-shed on this test is much worse than the current condition of sample 1E.

Sample 1B (stored interleaved with clear leader) is still playable. The oxide deposit on the test strip is about the same density as the 1C test strip, but the color is mostly brown. The backcoating deposit is denser than the current 1E test, but not yet as dense as the original full sticky-shed test shown above.

Preliminary conclusions:

The backcoating on my samples of Ampex 407 does speed the return of sticky-shed. Sample 1C (backcoating removed) and sample 1B (backcoating not removed, but stored with interleaved clear leader) are still playable, but sample 1E is not.

When the backcoating and the oxide coating are not in direct contact with one another, the backcoating deteriorates faster than the oxide coating.