CONTENTS AT A GLANCE Understanding Alignment Problems
Recognizing the problems Repair vs. replace Tips to reduce floppy-drive problems Spindle speed Track 00 test Radial alignment Azimuth alignment Head step Hysteresis Head width

Using Alignment Tools
Advanced tools

Aligning the Drive
Drive cleaning Clamping

Further Study

Floppy disk drives (Fig. 56-1) are basically electromechanical devices.

Their motors, lead screws, sliders, levers, and linkages are all subject to eventual wear and tear. As a result, a drive can develop problems that are caused by mechanical defects instead of electronic problems. Fortunately, few mechanical problems are fatal to a drive. With the proper software tools, you can test a troublesome drive and often correct problems simply through careful cleaning and alignment. This chapter explains the concepts and procedures for floppy drive testing and alignment.





A Teac FD-235 3.5" floppy drive.

Teac America, Inc.

Understanding Alignment Problems
The causes behind floppy-drive alignment problems will vary somewhat depending on the design of the drive itself, but some common causes can crop up time after time:
s Wear in the drive’s head positioning mechanism can eventually cause the radial align-

ment to drift out of specification.
s Various forms of debris often find their way into the drive’s mechanical parts and will

often accelerate wear in the drive—and in some cases, affect alignment directly by changing the way that the drive’s sensors and mechanisms respond. s Dirt and normal wear might cause the head mechanism to not slide as easily as it should. This puts an excessive load on the small stepper motor used to position the heads. s The read/write head assembly itself can become bent or otherwise damaged from accidental abuse—often not enough to cause a complete failure, but enough to significantly affect the alignment and cause problems when reading a diskette recorded by another drive. s Drive alignment problems might also be revealed by marginal diskettes. If a drive is slightly out of alignment, it might work with a good data diskette, but not with a marginal “cut priced” one.


As a technician, you will need to understand when a floppy drive is showing signs that might be related to alignment errors. In general, you should always respond to a chronic drive error by examining the disk media itself. Slowly spin the disk and observe both sides of the oxide layer. The layer should be smooth and even throughout—like the smooth surface of a quiet pond. If you see any marks or scratches on the disk, you should suspect that either the R/W heads are misaligned or that a significant buildup of oxides are on the R/W



head(s). If you have not already cleaned the R/W heads as part of your regular repair practices, clean them now. If the problem persists, the head assembly is probably severely misaligned and you should replace the drive or re-align it as you see fit. Other classic indicators of alignment trouble are reading and writing errors. Data is checked when it is read from or written to a drive. When you encounter a drive that has difficulty reading diskettes that were written on another PC (or writes diskettes that other PCs have difficulty reading), the drive’s alignment is in serious doubt. Fortunately, alignment software can test the drive and report on its specifications, allowing you to see any unacceptable performance characteristics. If you find that the drive is faulty, you can then decide to replace or re-align the drive at your discretion.

Floppy-drive alignment continues to be a matter of debate. The cost of a floppy-drive alignment package is often higher than the cost of a new drive. When compared with the rising costs of labor and alignment packages, many technicians question the practice of drive alignment when new drives are readily available. True, most casual PC enthusiasts would not choose to align a misbehaving drive. However, testing software has an important place in any toolbox. At the very least, test software can confirm the faulty alignment of a drive and eliminate the guesswork involved in drive replacement. For enthusiasts and technicians who have a volume of drives to service, alignment tools offer a relatively efficient means of recovering drives that might otherwise be discarded. Ultimately, one of a technician’s most vital tools is an open mind—you can repair or replace the drive, depending on what makes the most economic sense in your particular situation.

Now that you’ve seen the most common causes of floppy-drive problems, you can recommend some pro-active steps to avoid or reduce problems in the future:
s Keep the floppy drives clean Not only is it important to keep the drive’s R/W heads

clean, but it is also important to keep dust and debris from accumulating inside the drive’s mechanisms. A static-safe vacuum cleaner can usually remove unwanted dust or the debris can be removed by “blowing down” the drive with a can of electronicsgrade compressed air and a long, thin nozzle. s Do not force disks in or out This is a classic cause of drive problems. Disks that become stuck in the floppy drive (usually because the drive does not release the 3.5" disk’s protective shroud properly) should be removed with the utmost care. s Keep air from circulating through the drive Floppy drives should generally be installed outside of the normal air flow in a PC. This will reduce the amount of dust and debris that normally flows around and through the drive. s Use good-quality diskettes “Bargain” diskettes often use inferior oxides, which tend to rub off and accumulate on the R/W heads. If you encounter an unusual number of failures with “bargain” diskettes, you should clean the R/W heads and switch to a better brand of diskettes.



Using Alignment Tools
Drive alignment is not a new concept. Technicians have tested and aligned floppy drives for years using oscilloscopes and test disks containing precise, specially recorded data patterns. You might already be familiar with the classic “cat’s eye” or “index burst” alignment patterns on oscilloscopes. This kind of manual alignment required you to find the right test point on your particular drive’s PC board, locate the proper adjustment in the drive assembly, and interpret complex (sometimes rather confusing) oscilloscope displays. Traditionally, manual alignment required a substantial investment in an oscilloscope, test disk, and stand-alone drive exerciser equipment to run a drive outside of the computer. Although manual drive-alignment techniques are still used today, they are being largely replaced by automatic alignment techniques. Software developers have created interactive control programs to operate with their specially-recorded data disks. These software tool kits provide all the features necessary to operate a suspected drive through a wide variety of tests while displaying the results numerically or graphically right on a computer monitor (Fig. 56-2). As you make adjustments, you can see real-time results displayed on the monitor. Software-based testing eliminates the need for an oscilloscope and ancillary test equipment. You also do not need to know the specific signal test points for every possible drive. Several popular tool kits are on the market, including FloppyTune by Data Depot, Inc. and DriveProbe (Fig. 56-3) by Accurite Technologies Inc. The contact information for both manufacturers is listed at the end of the chapter.

Although software tools make up a majority of the typical floppy-drive service options, the tools available to serious floppy-drive service technicians do not stop at software. With the proper supplemental test hardware (such as the Drive Probe Advanced Edition from Accurite Technologies, Fig. 56-3), a PC can be turned into a comprehensive floppy-drive



The DriveProbe automatic drive-test display.
Accurite Technologies, Inc.




DriveProbe: the advanced edition.
Technologies, Inc.


test bed that supports all types of standard PC drives, as well as Macintosh drives and many types of floppy-disk duplicator drives.

Aligning the Drive
At this point in the chapter, you are ready to start the testing/alignment software and go to work. Before starting your software, however, you should disable any caching software that will cache your floppy drive(s). Because caching software affects the way in which data is read or written to the floppy disk, caching will adversely affect the measurements produced by the alignment software. To ensure the truest transfer of data to or from the floppy disk, boot the PC from a “clean” boot disk to disable all TSRs or device drivers in the system. Once the alignment software is started, eight major tests gauge the performance of a floppy drive: clamping, spindle speed, track 00, radial alignment, azimuth alignment, head step, hysteresis, and head width. Remember that not all tests have adjustments that can correct the corresponding fault.

Floppy-drive R/W heads are not terribly complex devices, but they do require precision positioning. Heads must contact the disk media to read or write information reliably. As the disk spins, particles from the disk’s magnetic coating wear off and form a deposit on the heads. Accumulations of everyday contaminants, such as dust and cigarette smoke,



also contribute to deposits on the heads. Head deposits present several serious problems. First, deposits act as a wedge—forcing heads away from the disk surface, resulting in lost data and read/write errors, and generally unreliable and intermittent operation. Deposits tend to be more abrasive than the head itself, so dirty heads can generally reduce a disk’s working life. Finally, dirty heads can cause erroneous readings during testing and alignment. Because alignment disks are specially recorded in a very precise fashion, faulty readings will yield erroneous information that can actually cause you to improperly adjust the drive. As a general procedure, clean the drive thoroughly before you test or align it. R/W heads can also be cleaned manually or automatically. The manual method is just as the name implies. Use a high-quality head cleaner on a soft, lint-free, anti-static swab, and scrub both head surfaces by hand. Wet the swab, but do not soak it. You might need to repeat the cleaning with fresh swabs to ensure that all residual deposits are removed. Be certain that all computer power is off before manual cleaning and allow a few minutes for the cleaner to dry completely before restoring power. If you do not have head-cleaning chemicals on-hand, you can use fresh ethyl or isopropyl alcohol. The advantage to manual cleaning is thoroughness—heads can be cleaned very well with no chance of damage from friction. Most software tool kits provide a cleaning disk and software option that allows you to clean the disk automatically. With computer power on and the software tool kit loaded and running, insert the cleaning disk and choose the cleaning option from your software menu. Software will then spin the drive for some period of time—10 to 30 seconds should be adequate, but do not exceed 60 seconds of continuous cleaning. Choose high-quality cleaning disks that are impregnated with a lubricant. Avoid “bargain” off-the-shelf cleaning disks that force you to wet the disk. Wetted cleaning disks are often harsh and prolonged use can actually damage the heads from excessive friction. Once the drive is clean, it can be tested and aligned.

A floppy disk is formatted into individual tracks laid down in concentric circles along the media. Because each track is ideally a perfect circle, it is crucial that the disk rotate evenly in a drive. If the disk is not on-center for any reason, it will not spin evenly. If a disk is not clamped evenly, the eccentricity introduced into the spin might be enough to allow heads to read or write data to adjoining tracks. A clamping test should be performed first, after the drive is cleaned because high eccentricity can adversely effect disk tests. Clamping problems are more pronounced on 5.25" drives, where the soft mylar hub ring is vulnerable to damage from the clamping mechanism. Start your software tool kit from your computer’s hard drive, then insert the alignment disk containing test patterns into the questionable drive. Select a clamping or eccentricity test and allow the test to run a bit. You will probably see a display similar to the one shown in Fig. 56-4. Typical software tool kits can measure eccentricity in terms of microinchesfrom true center. If clamping is off by more than a few hundred microinches, the spindle or clamping mechanisms should be replaced (be sure the diskette itself is not damaged). You can also simply replace the floppy drive. Try reinserting and retesting the disk several times to confirm your results. Repeated failures confirm a faulty spindle system.





Screen display from a DriveProbe eccentricity test. Accurite Technologies, Inc.

Media must be rotated at a fixed rate in order for data to be read or written properly. A drive that is too fast or too slow might be able to read files that it has written at that wrong speed without error, but the disk might not be readable in other drives operating at a normal speed. Files recorded at a normal speed also might not be readable in drives that are too fast or too slow. Such transfer problems between drives is a classic sign of speed trouble (usually signaled as “general disk read/write errors”). Drive speeds should be accurate to within ±1.5%, so a drive running at 300 RPM should be accurate to ±4.5 RPM (295.5 to 304.5 RPM), and a drive running at 360 RPM should be accurate to within ±5.4 RPM (354.6 RPM to 365.4 RPM). After cleaning the R/W heads and testing disk eccentricity, select the spindle speed test from your software menu. The display will probably appear much like the one in Fig. 56-5. Today’s floppy drives rarely drift out of alignment because rotational speed is regulated by feedback from the spindle’s index sensor. The servo circuit is constantly adjusting motor torque to achieve optimal spindle speed. If a self-compensating drive is out of tolerance, excess motor wear, mechanical obstructions, or index-sensor failure is indicated. Check and replace the index sensor or the entire spindle-motor assembly. You can also replace the entire floppy drive outright.

The first track on any floppy disk is the outermost track of side 0, which is track 00. Track 00 is important because it contains the boot record and file-allocation information vital for finding disk files. The particular files saved on a disk can be broken up and spread out all over the disk, but the FAT data must always be in a known location. If the drive cannot find track 00 reliably, the system might not be able to boot from the floppy drive or even use disks. Floppy drives utilize a sensor, such as an optoisolator, to physically determine when the R/W heads are over the outermost track.



Select the track 00 test from your software menu and allow the test to run. A track 00 test measures the difference between the actual location of track 00 versus the point at which the track 00 sensor indicates that track 00 is reached. The difference should be less than ±1.5 mils (one-thousandths of an inch). A larger error might cause the drive to encounter problems reading or writing to the disk. The easiest and quickest fix is to alter the track 00 sensor position. This adjustment usually involves loosening the sensor and moving it until the monitor display indicates an acceptable reading. Remember that you only need to move the sensor a small fraction, so a patient, steady hand is required. The track 00 sensor is almost always located along the head-carriage lead screw. Mark the original position of the sensor with indelible ink so that you can return it to its original position if you get in trouble.

The alignment of a drive’s R/W heads versus the disk is crucial to reliable drive operation because alignment directly affects contact between heads and media. If head contact is not precise, data read or written to the disk might be vulnerable. The radial alignment test measures the head’s actual position versus the precise center of the outer, middle, and inner tracks (as established by ANSI standards). Ideally, R/W heads should be centered perfectly when positioned over any track, but any differences are measured in microinches. A radial alignment error more than several hundred microinches might suggest a head-alignment error. Select the radial alignment test from your software tool kit and allow the test to run. A typical radial alignment test display is illustrated in Fig. 56-6. If you must perform an adjustment, you can start by loosening the slotted screws that secure the stepping motor, and gently rotate the motor to alter lead screw position. As you make adjustments with the test in progress, watch the display for the middle track. When error is minimized on the inner track, secure the stepping motor carefully to keep the assembly from shifting position. Use



Screen display from a DriveProbe motor-speed test. Accurite Technologies, Inc.




Screen display from a DriveProbe radialalignment test. Accurite Technologies, Inc.

extreme caution when adjusting radial head position—you only need to move the head a fraction, so a very steady hand is needed. You should also re-check the track 00 sensor to be sure the sensor position is acceptable. If you are unable to affect radial head alignment, the drive should be replaced.

Not only must the heads be centered perfectly along a disk’s radius, but the heads must also be perfectly perpendicular to the disk plane. If the head azimuth is off by more than a few minutes (1⁄60 of a degree), data integrity can be compromised and disk interchangeability between drives—especially high-density drives—might become unreliable. When the heads are perfectly perpendicular to the disk (at 90 degrees), the azimuth should be 0 minutes. Select the azimuth test from your software tool kit and allow the test to run. Figure 56-7 shows an azimuth-alignment test display. An azimuth-alignment test measures the rotation (or twist) of R/W heads in terms of + or – minutes. A clockwise twist is expressed as a plus (+) number, and a counterclockwise twist is expressed as a negative (–) number. Heads should be perpendicular to within about ±10 minutes. It is important to note that most floppy drives do not allow azimuth adjustments easily. Unless you want to experiment with the adjustment, it is often easiest to replace a severely misaligned drive.

The head step (or index step) test measures the amount of time between a step pulse from the coil-driver circuits and a set of timing mark data recorded on the test disk. In manual oscilloscope adjustments, this would be seen as the “index burst.” Average index time is typically 200 s for 5.25" drives, and 400 s for 3.5" drives. In automatic testing with your software tool kit, you will see time measurements for both heads on the inner and outer tracks (Fig. 56-8). The actual range of acceptable time depends on your particular drive, but variations of ±100 s or more is not unusual.



If the head step timing is off too far, you can adjust timing by moving the index sensor. As with all other drive adjustments, you need only move the sensor a small fraction, so be extremely careful about moving the sensor. A steady hand is very important here. Be sure to secure the sensor when you are done with your timing adjustments.

It is natural for wear and debris in the mechanical head-positioning system to result in some “play”—that is, the head will not wind up in the exact same position moving from outside in, as moving from the inside out. Excessive play, however, will make it difficult to find the correct track reliably. Testing is accomplished by starting the heads at a known track, stepping the heads out to track 00, then stepping back to the starting track. Head position is then measured and recorded. The heads are then stepped in to the innermost track, then back to


Screen display from a DriveProbe azimuthalignment test. Accurite Technologies, Inc.



Screen display from a DriveProbe index-to-data test. Accurite Technologies, Inc.




Screen display from a DriveProbe hystersis test. Accurite Technologies, Inc.

FIGURE 56-10

Screen display from a DriveProbe head-width test. Accurite Technologies, Inc.

the starting track. Head position is measured and recorded again. Under ideal conditions, the head carriage should wind up in precisely the same place (zero hysteresis), but natural play almost guarantees some minor difference. A typical hysteresis test-measurement display is shown in Fig. 56-9. If excessive hysteresis is encountered, the drive should be replaced because it is difficult to determine exactly where the excess play is causes in the drive.

Another test of a drive’s R/W heads is the measurement of their effective width. Effective head widths are 12 or 13 mils for 5.25" double-density drives, 5 or 6 mils for 5.25" highdensity drives, and 4 or 5 mils for all 3.5" drives. As you run the head-width test with your software tool kit, you will see effective width displayed on the monitor (Fig. 56-10). As



R/W heads wear down, their effective width increases. If the effective width is too low, the heads might be contaminated with oxide buildup. When small head widths are detected, try cleaning the drive again to remove any remaining contaminates. If the width reading remains too small (or measures too large), the heads or head carriage might be damaged. You can replace the R/W head assembly, but often the best course is simply to replace the drive outright.

Further Study
That’s all for Chapter 56. Be sure to review the glossary and chapter questions on the accompanying CD. If you have access to the Internet, take a look at some of these floppydrive maintenance resources: Accurite Technologies: http://www.accurite.com Data Depot: http://www.datadepo.com NEC: http://www.nec.com