Cassettes: Focus On Shell Mechanics (Sept. 1981)

Home | Audio Magazine | Stereo Review magazine | Good Sound | Troubleshooting

Departments | Features | ADs | Equipment | Music/Recordings | History


At least once a year, Audio tests a number of the latest cassette tape formulations. The effort is concentrated on electrical performance, such things as frequency responses and maximum record levels. It has been a regular practice to check flutter and skew effects, but we have not taken a detailed look at the mechanical inter-relationships that are involved. This article delves into such questions as: Is there such a thing as a low-flutter cassette? Does it make a difference which deck is used? Do all tapes skew? What should I look for in a cassette if I want good mechanical performance? Are some cassettes more reliable than others?

Fig. 1--TDK's new Metal Alloy audio cassette comes housed in a Reference Standard mechanism/shell. (Photo, courtesy TDK.)

Fig. 2--Disassembled Reference Standard mechanism. (Photo, courtesy TDK.)

Even with just a cursory examination in the course of buying tapes, you may have noticed that there is a great deal of difference in the way that the various manufacturers package the product.

This can be more important than it might at first seem. To minimize the possibility of getting harmful dust on the tape itself, the package should be sealed in some sort of plastic wrap. The little tab seal used by a few manufacturers will assure you perhaps that the tape has not been used by someone else, but the openings in the typical cassette box can allow a lot of dust to enter. After opening the pack age (a pull tab is a worthwhile convenience), follow good practice in keeping the cassette in the box when not actually being played. Do not store any tapes in dusty environments. If you must do so, try some of your own plastic protection, or stick to cassettes with better quality boxes with tighter closures. The new Memorex boxes provide better sealing from dust than the older standard-design ones do.

If you have purchased a number of different tape formulations, you will have learned that manufacturers do not agree on what makes the best label. We're not going to dig deeply into this question, but there are a few things to keep in mind. First of all, is there enough space to write in the necessary identification? Some labels are so small, you might need to use some sort of code or abbreviations. What can you write on the label with? Best of all are those which will accept almost anything, but some require a ball-point pen, making changes very difficult. Some cassettes have extra press-on labels, which is a definite help. Take a look at the outside card while you're at it: Some are good, and some have failings similar to those discussed for labels.

Now that we've talked about the wrappings and labels, it's time to take a closer look at the cassette shell itself.

Figure 1 shows the assembled TDK MA-R cassette, the most sophisticated design currently available. It is also relatively expensive, of course, so it is not surprising that most other assemblies are less impressive. The great majority of the premium cassettes sold have plastic half shells which are held together with screws. There are just a few that are sonic welded together, and their manufacturers feel that sonic welding can do just as good a job as screws. It is true that screws must be torqued correctly to hold firmly, but not so tight as to introduce unwanted stresses. It is also true that most of the really cheap cassettes are sonic welded, and that the cassette shells which are most nicely finished are all held together with screws. It is possible, of course, for a manufacturer to use more than one quality of shell and mechanism in its product line. TDK actually uses four, the Reference Standard Mechanism for MA-R tapes, the Laboratory Standard Mechanism for MA and SA-X tapes, the Super Precision Mechanism for SA, OD and AD tapes, and the Precision Mechanism for D tapes. We can't say that this is just the way to do it, but there is a great deal of sense in using a higher quality mechanical assembly to go with higher quality formulations.

To aid in the discussion of design requirements for the assembly, let's examine Fig. 2 which shows a disassembled Reference Standard Mechanism. In the center is the die-cast metal frame which must, and does, provide rigidity and stability, accurate outside dimensions and good finish, good support and accurate location for other components to be mounted in the frame, and parallel sides for mounting the cover plates. The cover plates must be flat with some rigidity and accurately dimensioned. The assembly of these two components has to provide the basic inside space for tape storage and guidance. There are two static-free slip sheets for low-friction restraint of tape wander during play or wind. Now, a cassette shell that is made out of plastic should meet the same basic criteria, ac curate dimensions, rigidity, stability, etc.

Carefully examine the cassettes you are using for surface smoothness, traces of plastic flash, resistance to bending or twisting, etc. Don't try to find their stress limit, but you may be able to weed out some questionable tapes.

When the tape pack is set into the shell, it is threaded over guide pins at each end and then over guide rollers, passing in front of additional guide pins, the tape pressure pad, and the mu-metal magnetic shield. For good mechanical performance, there are criteria that the tape pack itself must meet. The width must be constant, the slitting must not cause any deformation of the edges, the cutting must be a perfectly straight line, with no skew introduced, and the leader must meet the same criteria. As a fast check of the cassettes you have, look at the surface of the tape at the edge of the cassette. It should be perfectly flat and smooth. Any cupping or rippling might cause a range of tape problems, drop outs, changing levels, high flutter, and its getting wrapped around the capstan.

It comes down to this: To get the most out of the tape, it must remain in intimate contact with the heads--easy if its own surface is flat (except as shaped to the head), impossible if its own surface looks like a badly misaligned tire.

Fig. 3--How tape skewing causes alignment errors. A, basic mechanical interface. B, record head adjusted for same angle with tape as play head. Note skewing tape. C, results from turning cassette over without readjusting the record head. The effects are greatly exaggerated for clarity.

The tape must be fastened to the hubs, and yet the hubs must not intro duce any bumps as the tape winds on. The hubs must also be accurately round and concentric with the hub-drive opening. All the stationary guide posts must be perpendicular to the shell-frame reference plane. Their surfaces must be long-wearing and must not damage the tape in any fashion. The guide-roller surfaces must be smooth (TDK states "seam less"), concentric with the bearing pin (stainless steel preferred) and with flanges that guide and control tape wander without causing any edge damage. The magnetic shield should be accurately positioned behind the pressure pad. There are different approaches to the design of the pad and its support. Do check to see that the pad surface is flat to the tape and not misplaced crosswise. It is impossible to check out the internal construction of a typical plastic-shell cassette, unless you take it apart. Per haps if all that is not worthwhile, you would want to check what the manufacturer claims he has done to make his cassette a good one mechanically. If the edge of the cassette mates poorly and is not straight and smooth across, you should be suspicious of any claims for internal excellence. The key words are smooth, flat, perpendicular, parallel, round, concentric, rigid, stable and accurate.

With a few artistic liberties, Fig. 3 shows some of the cassette/recorder alignment relationships and how tape skew affects head alignment. In "A" the guide pins and rollers are shown to be exactly perpendicular to a rigid shell. The molded-in guide pins near the center of the cassette must also meet this criterion. The drive capstans and the record and play heads are shown as perfectly perpendicular to the recorder sup port assembly, which supports and positions the cassette shell exactly. This is our ideal of course, and if the tape ran perfectly straight, maybe it could really happen. First of all, recognize that if the cassette is not positioned accurately relative to the recorder components, some of the accuracy in the cassette itself is defeated. In other words, when you insert a cassette, make certain that it seats firmly into position. If there seems to be a favored position for best performance, use that all of the time.

Fig. 4--Record/ playback flutter with Akai GX-F90 deck. Top, TDK MA-R cassette; bottom, BASF Studio I cassette. (Scales: Vert., 0.05 %/div.; hor., 1 S/ div.)

Fig. 5--Record/playback flutter with Technics RS-9900/US deck. Top. TDK MA-R cassette; bottom, ferric cassette. (Scales: Vert., 0.05%/div.; hor., 1 S/ div.)

Fig. 6--Flutter spectrum with Akai GX F90 deck and TDK MA-R cassette. Scales: Analyzer bandwidth, 1Hz; 'scope. vert. 10 Hz/div.. hor.. 10 dB/div.)

Now let's take a look at how tape skewing affects head alignment and record/playback performance. In "B" of Fig. 3, the play head is shown to be perpendicular to the support plate with the tape curving across it. The curvature of the tape and the resultant angles are greatly exaggerated to facilitate demonstrating what happens. The lines on the tape, are radial lines of the curve and perpendicular to the edges. If we adjust the record head to get it in best alignment with the play head, we get the result shown in "B." Next we flip over this cassette ("C") to see how things would work out, without readjusting the record head. For the play head, the radial lines are just put at the same angle on the other side of vertical. With the record head, however, the angular error is twice what it was before any adjustment.

These simple figures tell us a number of things about what is desirable and what to expect. Most desirable are cassettes which have no skew and which will seat accurately in the recorder. With such tapes, alignment with both heads will remain correct, including when the cassette is turned over. Maybe this sounds unlikely, but, in fact, we have been able to report on a number of tapes that consistently do not have detrimental skewing from sample to sample and from side to side, including both C-60s and C-90s. Obviously, the problem of poor responses from skewing is more severe with separated record and play back heads. Note, however, that skewing can cause some loss in response even in combination record/playback heads. This is particularly true if you are going to play a tape on another deck than the one used for recording.

We know that smooth tape motion is essential for low flutter and that rough mechanical motion would result in high flutter, but how much does it have to do with the deck? The first part of the investigation utilized an Akai GX-F90 which had shown low flutter, the Nakamichi 582 which had average flutter, and the older-design Technics RS-9900/US.

Figure 4 has plots of the record/play back flutter with the Akai deck using a TDK MA-R (top) and a medium-price ferric (bottom). The straight lines are the reference zeros, the vertical scale is 0.05% wtd. pk. per division, and in creasing flutter is in the negative direction. The 'scope traces show a number of peaks not indicated on the meter, but there is no doubt about the much lower flutter with the MA-R (0.03% meter) compared to the ferric (0.08% meter).

When we tried the same cassettes in the Technics RS-9900/US (Fig. 5), the flutter with the MA-R was higher, the flutter with the ferric was lower, and they were pretty much the same with this deck.

Fig. 7--Flutter spectrum with Akai GX F90 deck and a ferric cassette different from the one used to generate the bottom curve in Fig. 5. (Scales: As in Fig. 6.)

Fig. 8--Flutter vs. tape and deck. Top. mediocre cassette with Nakamichi 582; 2nd, poor cassette with 582; 3rd, mediocre cassette with Aiwa AD-3600; 4th, poor cassette with AD-3600, and bottom, poor cassette with Aiwa AD-M700. (Scales: vert., 0.1% wtd. pk./ div., hor 1 S/div.)

Fig. 9--Record/playback 10-kHz phase error and jitter between channels. (Scale: Hor., 30 deg. /city.)

Fig. 10--The two Loran cassettes shown here were exposed to high temperatures in environmental test chambers along with four leading cassette tapes, labels removed. (Photo, courtesy Loranger.)

We went back to the Akai deck and plotted (Fig. 6) the MA-R flutter spectrum for 50 Hz each side of our test tone with a 1-Hz analyzer bandwidth. There are sidebands at ±4 Hz, 23 dB down, and at ±24 Hz, 35 dB down. Figure 7 shows the results with the same deck and another ferric cassette, which had shown twice as much flutter on the meter as the MA-R. Note that the side-bands at ±25 Hz are up to-21 dB, and that there is a lot of energy at a higher level at other points. Similar checks with the Nakamichi 582 showed spectra with reduced discrete sideband levels but with considerable "random" energy close to the test-tone carrier. The meter reading of 0.09% wtd. pk. was indicative of the total level of these many flutter components, even though the energy was not concentrated in a couple discrete frequencies.

Subsequent to taking the above data, two Aiwa decks were obtained which had low flutter with many cassettes. A search was made to find cassettes that had mediocre to poor flutter performance. The results from some of these tests are shown in Fig. 8. Please cote that the vertical scale is 0.1% per division, as compared to 0.05% per division in the earlier figures. As before, the straight-line traces are the reference zeros, and increasing flutter is downward.

The two topmost sets were run with the Nakamichi 582. The so-called mediocre cassette showed no values greater than 0.08% wtd. pk., and an excellent 0.05% was typical. The poor cassette was really that with relatively frequent readings to almost 0.2%, and a few close to 0.3%! The next two sets were made with the recently introduced Aiwa AD-3600. Preliminary tests had shown very low flutter with many cassettes. The first run with the "mediocre" sample showed most readings below 0.06%, with around 0.04% or less very common. These are certainly excellent figures, but on to the challenge of the cassette that had per formed so poorly just before in the other deck. These results (next to the bottom of Fig. 8) were quite unexpected, but the plotted figures were really quite typical--few peaks over 0.06%, with most meter indications less than 0.04% wtd. pk.! This low-flutter result left space for a run with the same cassette in the Aiwa AD-M700 deck, which had also shown well-controlled flutter. There is a notice able increase, compared to the AD-3600 deck, but the 0.08% maximums are still quite acceptable and much lower than the Nakamichi 582 results. In case there is any question, let it be stated here that exactly the same section of tape was used for each recorder and re checks were made of all the results.

In general, if you need very low flutter, you must have a good performing deck as well as a good cassette. A cassette cannot force a deck to have low flutter, but many decks are definitely sensitive to the characteristics of the cassette.

The Aiwa decks used in the tests reported above were the best seen to date in giving low flutter, regardless of the cassette used. Under a number of conditions TDK MA-R appeared to be the lowest flutter cassette. Some of the other good-performing tapes were Ampex EDR, GMI and GMII. BASF Professional II and III, Fuji FX-I. Maxell UD-XL I-S and UD-XL II-S, Memorex MRX-1, Osawa Cr, Realistic Supertape Chrome, Sony SHF, and TDK MA and SA-X. These conclusions must be considered tentative be cause of the limited, relatively short-time testing and because of the proven influence of the deck.

The last of the tape/recorder inter face effects to be discussed is record/ playback phase jitter. If the tape motion were perfect across the head, without waving or vibrating, there would be no shifting in time between channel A compared to channel B. Figure 9 shows the output of both channels of a recorder with a 10-kHz test tone, and with the scope locked to "A" (top). The relative phase jitter of "B" causes the trace to move back and forth on the screen, as shown in this timed exposure. The sweep speed was adjusted for 30 degrees per division, and we can see that the total jitter is about 40 degrees, which is fairly good for a cassette deck. A misalignment of about 40 degrees was purposely left in, evidenced by the displacement of the average position of the "B" trace. This angular discrepancy of the 10-kHz tone indicates an 11-uS time difference. Actual jitter and alignment errors can be much greater than that shown. The conclusion drawn after a series of tests with a selection of cassettes and a number of decks was this: Phase jitter is primarily determined by the deck, but the cassette has some influence on the exact results. The deck with the smallest distance between the record and playback gaps is most likely to have the least jitter. Recognize that such jitter will exist in any subsequent playback. It is also a fact that jitter and skewing can cause fairly high level losses at the higher frequencies when a tape is played back on another recorder, especially when there are head alignment errors.

At the time of this writing, Loranger has just introduced a line of cassettes which have shells made out of Lexan.

Among other things, the manufacturer claims that these shells are much more stable with elevated temperatures, such as might be found in car tape players.

Figure 10 does show very noticeable damage to the non-Lexan shells, so I subjected C-60 spares to oven temperatures of 120° to 160° for one hour, a temperature period that might well be found on a car dash. I found that the cassettes most sensitive to heat distort quickly. Only one-third to the end, though some shells were a bit distorted. The Loran Lexan shell showed the least effect, and it should be noted that the Maxell shells showed very little warp. This certainly is a valid area for investigation, and I will try to gather more information for possible publication later or.


In past years, there were a fair percentage of cassettes that would not survive very many fast winds, and occasion al ones that would not even play through one time without jamming. These types of failures are now much less frequent, and there are fewer cases of various types of sounds, squeaking, moaning, chattering, etc. There are still tapes, however, that are very noisy on one deck and most quiet on another. In general, it appeared that the cassettes with the smoothest winds (quietest) had the lower flutter and the least likelihood of jamming. There were a number of exceptions, and only very lengthy testing would prove whether there is much of a correlation. Good guidance to the tape pack with slip sheets did reduce over the-pack failures. Finally--and once again--the total cassette system performance depends upon the mechanical and electrical characteristics of both the tape and the deck.

(Source: Audio magazine, Sept. 1981)

Also see:

Open Reel Recorders (The Mechanism Of Magnetic Tape Erasure; Focus On Head Demagnetization) (April 1981)

Performance of High Energy in Magnetic Materials in Audio Cassette Recording Tapes (Sept. 1978)

All That Data: Tape Deck Frequency Response and Headroom (Jan. 1981)


Top of Page   All Related Articles    Home

Updated: Tuesday, 2019-05-07 17:28 PST