back to main turd

Sticky Shed Syndrome and Tape Baking
(click for larger image)

I've been doing transfers from old sessions for various clients and invariably came across the phenomena known as Sticky Shed Syndrome.

So as I was dealing with transfers I eventually had to modify a small convection oven to do exactly 54C as per the Ampex patent; and yep, Ampex knew of this and filed/got a patent on the process:

https://patents.google.com/patent/US5236790A/en

The issue I had was that most store-bought ovens don't go below 150F so I needed to modify the oven to hit 54C/128F. Since I design this stuff, it was easy for me - I really don't recommend it as a DIY unless you're a seasoned hardware engineer. But for those that are, it's simply a matter of placing a resistor across the NTC thermistor input of oven's microcontroller A2D input. Using a calibrated meter/thermocouple, I nailed 128F within a degree or two. Also put indicators across the coils to see when the tiny uC was actually firing the coils.

So as I was doing this, I recently also acquired another microscope for my hot air solder tool that has 1280x1024 resolution for dumping into a PC. Figuring I'd at least try to see if there was a visible difference between tapes that were giving me hell and those that weren't. I did some before and after shots as well as a photo of what appeared to be a tape that held up well with no special storage - PEM 468. As an example here's a song that was dumped from a safety master on 468 - no bake required - http://www.ajawamnet.com/ajawam /sdollar.mp3 . BTW -This was quickly remastered using Wavelab 6.0 with TC Powercore Denoiser, an MD3 and Voxengo Elephant. You can hear the hiss on the one channel as the TC grabs it.

This reel also had a section of Ampex 456 spliced in at the tail. As soon as it hit this, it became apparent that it was suffering from SSS after about 2 seconds - the typical squeal. When the tension was relaxed from the supply and take up reels, the tape just stuck to the heads with the tale tell line of melted binder apparent on the tape where it stopped. So this was the candidate I used for the photos.

So what we have here is:
- A reel of 1/4" tape that had not been played in over 20 years
- A reel of tape wound tails out with a library pack
- A reel of tape that had both PEM 468 and Ampex 456 both around mid 1980's vintage
- The Ampex section was at the tail, leadered with Scotch plastic leader; about 5 minutes of tape at 15 IPS
- A reel of tape that was not stored in any special vault; exposed to Pittsburgh (for 8 years), New Hampshire (3 years), and Northern Virginia (10 years) climates.

The reel was rewound using custom bypass rollers I had Terry at Terry's Rubber Rollers
( http://www.terrysrubberrollers.com/ ) lathe for me.
The machine is a BR-20T with the torque re-adjusted for minimal back tension.
Since this machine has idler arms with rollers, the entire fast rewind was via rollers.

A video of the tape path bypass can be found here
A video of a 1/2" TSR8 tape path bypass can be found here

Here's the photos - Click on the photo for a 1280x1024 image
PLEASE NOTE: these samples were clipped from the reel and baked flat on the
flange along with the parent reel, oxide up, about 4" in length.

	PEM 468 with no baking This is the Agfa PEM 468 which shows significant wear. The majority of the reel was this tape, which transferred easily with no baking. Typical dusty shed (slight due to the wear of the tape).
	Ampex 456 - Pre-bake This is from the head of the section, right after the leader.
	Ampex 456 - Post-bake This is the same section of Ampex 456 after a 3 hour bake at 54C.

Many sites that have audio forums mention it, as well as have a lot of misinformation on the issue. A few sites as well are fronted by people that are experts in this - below is a list of links to various ones that have good info on the problem as well as offer services for tape restoration:

Richard Hess:
http://richardhess.com/notes/formats/magnetic-media/magnetic-tapes/analog-audio/degrading-tapes/
http://www.richardhess.com/tape/

Sonicraft:
http://www.sonicraft.com

Some more info here:
http://www.tangible-technology.com/tape/baking1.html

A bit from Dr Ric Bradshaw, former head of IBM Materials Lab, in a web discussion with Richard Hess, Steve Puntolillo (Soniccraft), and myself that started because of these photos:

"This note contains a great deal of very useful information which could really help you all unravel what is really going on...IF you could actually get some chemical and mechanical analysis of these tapes before and after baking to separate/prove the possible hypotheses that may be at work in these various tape samples.

The Ampex tapes were made using Estane 5701 I am pretty sure...possible a blend of 5701 and 5707....both contain a polyester made from butanediol and adipic acid..the primary difference in the length of the nominal polyester chain....hydrolysis of the polyester-polyurethane cleaves the intial long chain polymerat the polyester repeating units(120,000 Mw or about 340 repeating units long). When the polyester segments of the polyester-polyurethane(PU) is broken down by acid catalysts, water and heat, the coating becomes soft and increasingly cotnains a liquid residue of small polyester fragments. These can pool on the surface and then scrub during tape motion on typical laminated heads used in audio recording such that when the tape stops and reverese direction, the tape slaps down onto the collected adhesive material and picks it up as a "line" or smear of sticky material. We were able to confirm this using high resolution mass spectrometry.

I was trying to get a die made to cut samples for the dynamic mechanical thermal analyzer (DMTA) from the 1/4" tapes Richard sent me before I left IBM..it was never completed by the model shop. I sent Richard's note with the link to Wayne's photos on to Dr Dylan Boday at the IBM Materials Lab in Tucson in the hope that he might be interested in doing some analyses on these tapes samples.

What you need to know is what happens to the tapes during baking....chemically and mechanically. I think I know what is going on, but my direct experience was with several binder systems used to make digital 1/2" tapes in the 70's and early 80's...and limited experience with the later versions of Ampex instrumentation and audio tapes (used by NASA). It would be a relatively straight forward analysis which Benoit was
doing in France a few years ago..but never go the effort fully completed for these tapes.

What you see visually is surface embossments which show up much better with a scanning electron microscope (also available at the IBM Materials lab)..some of the tapes which had observable "dusting" on the surface are sitll probably degraded but the dust is not the degaded polyester fragments but the granular ureas from the polyurethane degradation. The polyester fragments are buried within the coating matrix probably due to not being baked or having been run below the glass transition temperature (below 20C) so that the caoting is mechanically worn producing the dusting rather than sticky surface conditions. I note what appears to be considerable scratching of those tape surfaces consistent with that type or wear induced damage. The tough components of the binder system are the polyesters..they are rubbery even when degraded somewhat...but the urethane fragments left behind are not..they are called the "hard segments" for that reason..they impart strength but are not elastic. These materials are high melting and would act like grit at the head to tape interface...leading to digital drop outs and SNR degradation, but no squeal or detectable noise in an analogue detection chnanel with the track widths used. The effect of baking at temperatures about 45C is to soften the coating. These tapes have a glass transition temperature detectable by dynamic mechanical analysis (DMTA) typically with an onset of 30C with the viscoelastic transition point (tan delta peak max) close to 50-52 C hence your experimental observation of needing to bake the tapes at a temperature aboue 50C to see the cahnge in the surface condition.

What happens to the wraps of tape on the spool during baking is thermal mechanical creep to relieve the compressive and tension stresses placed in the spool of tape during winding. The surface of the tape coating softens and the topography of a worn tape flows back to a smoother surface due to the compressive forces of the over wraps in the spooled tape during backing. If you back sections of the tape as flat specimens in the oven..not wound against the backside of the overwrap..you would actually see material move to the surface...and probably very little change in surface topography...but change in gloss or reflectivity. I believe your observable staining has a similar mechanistic explanation.

As Richard and I discussed during the creation of his paper, baking does not return the polymer to it's intial conditon...just can't happen this way..and once the tape is degraded it is a losing battle to try to "reconstitute it". You can alter it so that it can play, but it's initial durability is gone and it should be copied to a new tape or converted to a digital format."

Microscopy of a Newer Batch of Ampex 456

As I was dumping tapes from other studios, I noticed that even some newer batches of 456 were still exhibiting some degree of stickiness, though not to the extent of the above sample. I noted that none squealed, but near the head and the tail they displayed signs of what Richard Hess's ARSC article mentions as SBS - Soft Binder Syndrome.

At the sections right after the discloration near the head and tail (up to approx 2 minutes in), these tapes when stopped would cause a slight line of melted material, with little deposit on the heads or fixed guides - ocassionally sticking only to the playback head and last fixed guide when the tension was removed from the reels in stop. As the tapes progressed towards the middle of the reel, they became less sticky and played or fast motioned over the fixed lifters with no issues. In fact, even the heads and tails of the tape seemed to tolerate fast shuttles over the fixed lifters without sticking or buildup of residue on the lifters themselves.

On some of the reels, a discoloration similar to that shown on the previous sample above was present. Below are more microscope images I grabbed.

Click for higher resolution - Note that the pre bake backing shots were shot at a lower res
(dumbass me - forgot to reset the resolution in the software).
PLEASE NOTE: these samples were clipped from the reel and baked flat on the
flange along with the parent reel, oxide up, about 4" in length.

	PRE BAKE	POST 4 HOUR BAKE
This is section 1 which was at the very tail of the tape
This is the next section, which shows less discoloration
This is from about 1 minute into the tape where it was stopped after playing at the playback head As the tape progressed less debris accumulated to the point that the middle of the tape showed no signs of this and played well. See Below for 500X photos of this
This is the backing from section 1 as per request from Richard Hess
This is the backing from section 2.

Some 500x closeups of the post-bake "stop" mark - you can download a zip file of the 1280x1024 jpgs here
A Zipfile of two TIFs of the crystal areas can be found here

Here's some shots at 500X of 1990 vintage 1/2" prebake