THE SMITH-KETTLEWELL TECHNICAL FILE

A Quarterly Publication in Braille Talking Book and IBM Diskette Editions from The Rehabilitation Engineering Center The Smith-Kettlewell Eye Research Institute
Bill Gerrey, Editor

Supported, in part, by The Smith-Kettlewell Eye Research Institute and National Institute on Disability and Rehabilitation Research

Braille Edition Produced by Clovernook Printing House Cincinnati, Ohio

Talking Book and IBM Diskette Editions Produced by The Smith-Kettlewell Eye Research Institute San Francisco, California

1992-93 SUBSCRIPTIONS

Make checks payable to: The Smith-Kettlewell Eye Research Institute
2232 Webster Street
San Francisco, California 94115

Please address all correspondence to: Bill Gerrey, Editor at the above address or call: (415) 561-1619

TABLE OF CONTENTS
1992-93, VOLUME 13, NUMBER 4

DIGITAL MAGNITUDE COMPARATORS

A REVIEW OF THE MICRONTA 22-182 DIGITAL MULTIMETER WITH COMPUTER INTERFACE

FOLDING CANES AND MODIFICATIONS THEREOF

DIGITAL MAGNITUDE COMPARATORS

Abstract--

These chips allow two binary numbers to be compared. Outputs indicate whether one is less than, equal to, or greater than the other. Singly, these chips compare 4-bit numbers; however, they can be cascaded to accommodate large numbers. Applications would include comparing a counter with the output of an A-to-D converter. Suppose you run a stepper motor from the clock of the counter; you could get to stop, or change direction, as directed by the A-to-D converter's output.

The Motorola MC14585BCP

Two 4-bit "words," "A" and "B," are fed to four inputs. As they define them, these inputs are: A0, A1, A2, A3, and B0, B1, B2 B3. (The editor prefers to think of these inputs as A1, A2, A4, A8, etc., but the notation is arbitrary.)

There are three outputs: "A less than B," "A equals B," and "A greater than B." The appropriate output goes high for its given condition.

There are three inputs used for cascading: "A less than B," "A equals B," and "A greater than B." For some obscure reason, on the least-significant chip, these are to be tied "low," "high," and "low," respectively. On the other hand, "less than," "equal to," and "greater than" inputs of each higher-order chip go directly to those corresponding outputs of its less-significant neighbor. (If only one chip is used, these inputs are still to be committed with "equals" being high and the others low.)

As you might expect, the outputs of the highest-order chip are the ones you pay attention to for the ultimate comparison.

[The truth table describing the operation of those cascading inputs suggest that the "greater than" one is a dead pin. If you think about it, the chip could learn anything it needs to know by responding to states of the first two. A frequent entry in the truth table for the "greater than" input is "don't care." Another source of literature, "The CMOS Pocket Guide," says to tie the "greater than" input high, as in low high high. One of us ought to leave it open some time and see what smokes--or doesn't smoke.]

Specifications for the MC14585BCP

These are CMOS devices. Their supply voltage can be between 3V and 18V. In quiescence, and without a load, the current drain is minuscule--0.005uA typical, 5uA maximum. (As devoted students of the article "Inside Gates" will remember, SKTF, Summer 1981, current drain of CMOS logic goes way up in operation, since complementary FET's simultaneously conduct while changing state.)

Their outputs can drive two low-power TTL inputs, or one low- power Schottky input. (This means sinking 1.6mA.)

As far as speed is concerned, they give a figure called "turn- on/turn-off delay time." This is the time measured from when an input line is half-way up and the time when the output state is half-way up. It varies considerably with power supply voltage--at 5V, the maximum time is 860 nanoseconds (ns). At 10V, this maximum is 180ns; at 15V it is 130ns. They show typical delays of half these values, but the prudent engineer would plan on a delay of a microsecond at 5 volts.

Pin Assignments for the Motorola MC14585BCP (16-Pin Plastic Package)

The CD4063

Once an RCA number, then made by Harris (and who knows in six months), this is a very similar device to the above. They are not pin-for-pin compatible, however.

As with the Motorola chip, the least-significant device should have its "cascading inputs" as follows: A less than B low, A equals B high, and A greater than B low. (This input arrangement holds true if a single chip is used.) In a cascade of chips, these inputs go to the corresponding outputs of the less- significant neighbor.

The sense of the 4-bit inputs is positive logic--high equals 1. The outputs: A less than B, A equals B, and A greater than B, go high when their condition is met.

Pin Assignments for the CD4063 (16 pins)

A REVIEW OF THE MICRONTA 22-182 DIGITAL MULTIMETER WITH COMPUTER INTERFACE

by Jay Williams

ABSTRACT--

As you may know, Radio Shack no longer sells a multimeter equipped with internal digitized speech. It has been replaced by the Micronta 22-182, which instead provides an RS232C serial port and associated PC-compatible software. In addition to voltage, current and resistance, this meter can measure capacitance and frequency. It also includes a simple continuity tester, a logic level indicator and a transistor beta checker.

Introduction

At $130, this is a good buy. However, if you do a lot of work with electronics, this meter should simply be one more tool, not your "be-all and end-all." For the blind builder there is no substitute for the traditional analog meter with auditory output when measuring dynamic parameters.

Readings can be displayed in several modes. First, all quantitative measurements can be displayed relative to a reference of your choice. Second, the meter can be set to register maximum and minimum readings, and third, it can be set to hold a reading indefinitely. Because you can interface the meter with a computer, you can save a sequence of readings to a file.

In this article I will discuss these features in detail and how the meter can be accessed by your computer. And, because no such device is perfect, I will point out the few but relevant flaws you will encounter when using it.

Thank you, Susan Fowle, for writing some very useful programs that render the meter more user-friendly. Tom Fowle also deserves gratitude for his advice to a neophyte writer on computers such as I.

Location of Essentials--

Since this meter will not function without its battery and fuse, it makes sense to ascertain their whereabouts. Orient the meter so that its control panel faces downward and the end containing the display points away from you. Notice that the bottom cover has two rubber "feet" that extend nearly the width of the cover. One foot is right at the bottom end and the other is about two-thirds of the length distant. In fact, it defines the beginning of the display section.

The door for the battery compartment -- which accepts the 22- 182's single 9-volt battery -- comprises the bottom end of the cover. The Phillips screw that secures the door is a finger's width up from the center of the rubber foot.

In order to gain access to the fuse, remove the four Phillips screws that secure the meter's bottom cover. The battery door is actually part of this cover so you need not remove it separately for this operation. Two of the screws will be found in the extreme corners at the bottom end of the cover. The other two are placed at each end of the rubber foot that defines the beginning of the display section. When the cover is removed, you will find the fuse holder adjacent to the battery and just before the flexible protective cover for the circuitry.

All of these screws have unthreaded portions in the middle of the shanks which prevent their falling out of the holes.

When removing this cover, be careful not to jerk the wires of the small loudspeaker which is force-fitted into a recess in the cover. You can remove this speaker quite easily with your fingers. This loudspeaker emits the various "beeps" that indicate the completion and status of many functions.

The bottom cover also contains a stand that can be extended by grasping it from within the rectangular recess just beyond the battery door.

Control Panel Layout--

Orient the meter so the display points away from you. Below the display and proceeding from left to right are three rectangular rubber buttons.

The first button stands by itself and is the "power on/off" button. Toward the right side of the panel is a pair of similar buttons. The first one toggles through four states: normal reading, data-hold, maximum, and minimum. The button to its right toggles between "normal" reading and "relative" reading. These are discussed in detail later.

Below these buttons is a large, 30-position selector control. At its top-left are two rectangular holes, one above the other. These are the holes into which the leads of a capacitor can be inserted for measurement. At the selector's lower-right is a circular object. This is a socket into which a transistor can be inserted for checking.

Below the capacitor socket is another pushbutton labeled "comm." When toggled on, the meter sends data continuously through the RS232 port. When toggled off, further readings are issued only when you press a key on your computer's keyboard.

At the bottom of the panel are four holes going from left to right. These holes accept the connectors for the test probes. The probes are conventional "pencil-like" extensions with pointed tips. Their connections to the meter are hollow cylindrical plugs. The holes in the meter that accept these plugs have a protrusion that must engage the corresponding center hole in the plug. While these connections greatly reduce the likelihood of contact with your skin, they require frequent attention since they do not make connection until they hit the bottom of the hole. This is the first thing to check if you get no response.

You will use the two right-most holes for most measurements. The farthest hole to the right is the "hot" terminal, the one to its left is "common." To the left of "common" are two holes (BOTH POSITIVE INPUT TERMINALS) for measuring current. First comes one for measuring up to mA200. This terminal has a 2-amp fuse. The left-most terminal is unfused and is for measuring up to 20 amperes. The manufacturer recommends a maximum "on-time" for such measurements of 30 seconds followed by fifteen minutes rest.

The connection for the RS232 port is on the right side of the meter. If you feel down the side, directly to the right of the "hot" terminal, there is what appears to be a lengthwise "scratch" in the plastic, just above the seam. This is actually a row of seven small holes. The 7-pin connector on the supplied cable plugs into this "socket," so that the cable extends toward the top of the meter. The plug can be inserted in the opposite direction, but this will not harm the meter since its innards are optically isolated from the computer's serial port. The reverse connection of this plug results in an "I/O error" message.

A Few Flaws--

Unfortunately, the selector knob has no "STOP" that would prevent continuous rotation. Although you can determine the setting of this control from the information on your computer screen, we recommend that you apply physical markings to indicate at least the highest ranges of voltage and current. You will probably want to mark other positions you use often.

The sockets that are used for capacitors and transistors are poorly designed, so you will have to use some ingenuity in order to ensure reliable connections between their contacts and the leads of the component to be measured.

Although we managed to obtain a reading for transistor beta, it took much "fiddling." We suggest that, since this meter gives you no more significant information than you can get by checking it with a continuity tester (see Continuity Tester Uses, SKTF, Fall, 1982), ignore this feature of the meter.

Meter Functions and Specifications

Turn the meter on by pressing the "power" button. The meter will emit a long and a short beep if it "powers up." Pressing the button again turns off the meter. It utters a very short beep when it "powers down." It turns off automatically if no readings are taken for about ten minutes. Be sure that the test probes are plugged into the two right-most sockets.

Since there is no "end stop" on the function selector, the most logical point of orientation is the "continuity" function. Rotate the selector so that the slotted end of its pointer is at the "nine o'clock" position, between the two sockets for capacitor leads. The continuity position is four clicks clockwise. You can confirm that you have selected this function by touching the probes together and waiting for the steady tone. Depending on the mental gymnastics the meter's performing, the tone may not appear for a second or two. The tone will not appear if the resistance between the probes is 30 ohms or more.

Resistance--

The next six notches going clockwise are the following resistance ranges: 200 ohms, 2 kilohms, 20 kilohms, 200 kilohms, 2 megohms, and 20 megohms. No audible indication of "overload" is presented in the resistance and "logic low" functions. In all other modes the meter emits a continuous tone when an overload condition is reached.

The display and your computer will show a message such as, ". ol kohm." The "PERIOD" can precede the "ol" immediately, or separated from it by a space, or be placed between the two letters. The PERIOD indicates the degree of resolution to which the meter is set within the function being monitored.

D.c. Voltage--

Proceeding clockwise, d.c. voltage is next. Its six ranges follow in this order: 20mV, 200mV, 2 volts, 20 volts, 200 volts, and 1000 volts. Positive and negative readings can be taken.

A.c. Voltage--

The a.c. voltage ranges are next and proceed in reverse order: 750 volts, 200 volts, 20 volts, 2 volts, 200mV, and 20mV. The manufacturer recommends that you not use this meter in circuits whose common is more than 500 volts above "earth ground." They also pointedly discourage the metering of three-phase a.c. circuits with this meter. They further state that if, however, you insist on doing so, do your calculations very carefully.

A.c. and D.c. Current--

Continuing clockwise are the a.c. current, then d.c. current ranges as follows. a.c.: 20mA, 200mA, and 2 amps. D.c.: 2 amps, 200mA, and 20mA.

For measurements up to 200mA, plug the hot test probe into the hole to the left of "common." For measurements between 200mA and twenty amperes, insert the hot test probe into the hole at the far left. As stated earlier -- but worth repeating -- this jack is not fused, and the manufacturer recommends a maximum "on" time of 30 seconds to be followed by 15 minutes of "off" time.

The high-current jack is functional only when the selector is set to the highest current ranges. The jack for the "hot" probe that you use for all other measurements is rendered ineffective in the "current" modes.

Transistor Beta--

Then comes the position for checking transistor beta, designated as h, with the subscript FE. As mentioned previously, taking a measurement in this mode is really more trouble than it's worth. The socket is the culprit. The holes that accept the leads are too large, and there are a lot of them--eight, to be precise. There is a separate semicircle of four holes each for NPN and PNP transistors. The manual does not make it clear as to how to use these four holes. It further states that bipolar transistors, not of the power variety, are the only ones for which the meter is designed to check the gain.

Logic Levels--

The next position on the selector is for determining logic levels. The display reads "ready" when this mode is first selected. Connect the test probes to the minus and plus power supply voltages of the circuit containing the logic levels to be tested. Then, press the "rel" button and start your search with the hot probe. If the point in question is at 70% of the power supply's voltage or greater, the display reads "high" and a continuous tone is emitted. If the point is 30% of this voltage or less, the reading is "lo" and no audible indication is given. If the voltage is between these two references, the word "float" appears.

Capacitance--

Clockwise from the logic mode is a three-range capacitance section. The progression is 2000pF, 2000nF, and 20uF. An overflow condition emits a continuous tone.

Frequency--

Lastly, there is a two-range frequency meter. The first range extends to 20kHz, and the second, is 200kHz. I have gotten a reliable reading with a signal as low as 20 millivolts at 1kHz. The manual advises that you not measure a frequency whose voltage exceeds 250 volts RMS. Here, also, a tone is emitted when the frequency range is exceeded.

Data Presentation Modes--

Once you have selected a function, you have some choices as to how data is presented. Fortunately, the meter defaults to the conventional mode when it "powers up." The two buttons below the right-hand end of the display accomplish most of these. The left-most button toggles through four modes:

Pressing it once initiates "data-hold"; this causes the meter to retain the most recent reading. Pressing it a second time causes the meter to display the minimum reading observed. Pressing the button again displays the maximum reading. The meter updates both maximum and minimum readings when monitoring either parameter, so when you change from displaying the minimum to displaying maximum, the latter has already been held in memory. These data are erased with any further change of function or when the meter's turned off. Pressing it the fourth time makes it give current readings, which is the default.

Selecting the "Relative" Function--

The right-most button toggles between "normal" and "relative." This function is used when determining logic levels and in situations where you wish to set a numerical value as a "zero" reference, such as a particular voltage. For example, you may want to observe the fluctuations of your household a.c. voltage during peak periods. You can start monitoring it at 5:00 p.m., press the "rel" button which causes the meter to read that voltage as "0." From there on, the meter reads the "difference"; you can take readings and the display will indicate their fluctuations either side of zero. Pressing the button again returns the meter to "normal."

The "Comm" Function--

The rectangular button located between the capacitor sockets and the panel that contains the test probe sockets toggles this function on and off. When active, it sends data continuously through the meter's serial port. When inactive, you must update the reading on your computer with a command from the keyboard. This function is useful if you choose to read the meter with your telecommunications software. It should not be used with either the software that comes with the meter or the software we have developed because the accumulated readings will clog the buffers quickly. These programs can be set to perform the same function.

Reading the Meter via your Computer

In addition to the test probes and the cable that connects the meter to the serial port, the meter comes with a 3.5 inch disk that contains three programs and a "readme.txt" file. The programs are: Dmm.exe, Metdemo.exe, and Metdemo.bas.

The program called "dmm.exe" displays the readings from the meter using graphics and is not accessible. On a computer with a monochrome display, it merely prompts you to hit "enter" and then displays an "illegal function" message and prompts you to "press any key to return to the system."

The programs called "metdemo.exe" and "metdemo.bas" are source code and compiled versions of the same program. They will send continuously updated readings to the screen. The ".bas" file is written in interpreted basic such as the common "gwbasic" which comes with many versions of "dos". To run such programs, make sure that a program such as "gwbasic.exe" is in your search path. Then type "gwbasic" followed by the name of the program as, in this instance, "metdemo.bas."

In addition, Susan Fowle has created some programs in Basic that tailor the readings for access by speech and Braille. These programs are public domain, so you may share them, but they should not be sold. We have included them on the disk along with this article. A brief description of each program follows.

In order to run programs with the .bas extension, type "gwbasic" followed by the name of the program and follow the prompts. Brtdmm.bas takes one reading with each keystroke. The cursor remains on the same line as the reading, making this easy to use with either a Braille display or a speech program. With speech, you will want to set a window for this line.

Focdmm.bas was designed to run with the Covox "Speech Thing" synthesizer, but it can be easily modified to run with a speech synthesizer that is driven as a DOS device, (that is, a stand- alone unit hooked to a serial or parallel port). By "digital focus" we mean that the program can be set to present selected portions of the reading to such a speech synthesizer, even though the entire reading appears on the screen.

Rptdmm.bas is also designed to work well with a Braille display. It continuously accesses the meter and is thus akin to reading the meter with the "comm" function. Pressing any key stops the program.

Finally, Stdmm.bas is another program designed to work with the Speech Thing, but will work with other synthesizers. It has an option that allows each character to be spoken separately.

Specifications for Writing your Own Program--

The following communication parameters are required: 1200 baud, 7 bit ASCII; no parity; 2 stop bits.

Data Format--

Each reading consists of fourteen bytes numbered in the hexidecimal standard 1 through 9, and A through E. In the following two examples, a byte that is occupied by a blank space will be shown as an equals sign (for sign). Note that byte E is always a carriage return ().

Command Structure--

  1. The meter must receive the D command to activate data transmission,
  2. the computer must give the meter the C command to clear its memory, and
  3. the computer must give the meter the M command (memory call) to transmit data from the meter's memory to the software).

Cable Formats--

Since the demons of disorder dictated the arrangement of pins in the DB9 plug so that they bear no discernible relationship to those in a DB25, we present them here in Table I. Table II shows the relationship of the 5-pin plug that enters the meter to the DB9 socket on the other end of the cable.

Table I. DB9 to DB25 pins:
Table II. 5-Pin Plug to DB9 Socket

(with the socket facing away from you); pin 1 is in the top left corner. On the 5-pin plug, I will call pin 1 the pin nearest the cable.):

As the serial port on this meter is optically isolated from the rest of the meter's works, the computer must supply power on the two handshaking lines on pins 4 and 7 of the Db9 connector. In the provided basic programs, note the lines which read: OPEN "COM1:1200,N,7,2,RS,CS,DS,CD"

Those used to "dos mode" setups for serial ports will see the addition of settings beyond the 2 indicating 2 stop bits. These set up the handshaking lines, 1 high and 1 low, to provide power for the meter's serial port. Since most communications programs do not allow for these settings, you may have trouble using the meter with such. If you need to power the meter from another source -- that is, you have a serial port which can't be set up to provide power with those handshaking lines -- a 9-volt battery in series with a 470-ohm resistor connected with the plus on the meter's pin 4 and the negative on pin 7 might do the trick. Don't forget the resistor for current limiting to avoid popping the optical isolators in the meter. Failure to provide this voltage will result in no apparent damage, but won't allow the meter to see its port.

FOLDING CANES AND MODIFICATIONS THEREOF

by Bill Gerrey

[At one time, I intended to call this paper "Modifying Commercially Available Folding Canes for Use by the Blind." However, they have steadily improved, thanks to our involvement in their design, so that title would be unnecessarily snide today.]

Introduction

You would never know it from conversations with me or by my works, but I'm an opinionated guy. If you can get me to talking on a subject--and one word might do it--I have set ideas as to how things should be. On the subject of available folding canes:

I hate rigidly affixed (not rubber-mounted) tips that stick in cracks and give you seven extra navels before you get home. I hate plastic tips that don't provide a good sound source for echoes--not making good tapping sounds--so you needlessly walk into trees, and you miss doorways that you are looking for. And, oh, I suppose, a gooey rubber handle is nice when you're walking, but the fact that it increases the size of the folded instrument by 50% makes it bust your pockets; isn't the purpose of a folding cane to get out of your way?

The result is, when people see my cane, they invariably ask, "What earthly kind is that?"

Telescoping Canes

There have been various brands of these over the years. One, made of light aluminum, was so flimsy that it would break under the force of a hard rain. Others used chucks for each section, making a cane that was durable, but whose collapsed length was no more convenient than the extended instrument. Finally, fiber- composition canes whose sections are held together by static friction have become popular. (I confess to having worked on such a one.)

There is a friction-fit cane on the market which is reported to be moderately durable. Two other advantages are that it is very light-weight, and that it has a rubber-mounted metal tip that makes good sounds for echoes and which rarely sticks in cracks.

From my experiments, I have concluded that all friction-fit joints have a mode of failure about which I am duty-bound to warn you:

Between solids, "static friction" ("sticktion") is greater than "dynamic friction." Users of canes held extended by static friction are advised to twist the ends as the cane is pulled taut; dynamic friction will be in play until the operator stops twisting, and when motion stops, it will take more force to dislodge the sections than was applied in "setting" them.

This sounds just dreamy until one experiences conditions where sticktion is "broken" and a joint shifts slightly. Once dynamic friction is the only force holding the joint together, it is likely to collapse.

Where are those trouble spots that make your telescoping cane go flaccid on you? Places where vibration causes slight movement in a joint. What sort of terrain causes vibration? Particularly asphalt.

Therefore, a friction-fit cane is most likely to collapse in the middle of a street--quod erat demonstrondum. As trucks bear down on you and you are trying to find the safety island, you must pull the cane taut and try again knowing that this risk will be no less probable a second time.

As a second spare when you're traveling, however, you might consider adding the NFB carbon-fiber telescoping cane to your collection.

Multisection Canes Held Rigid by Elastic

Various companies, Mahler, Hicor, WCIB, and foreign-made models exist which have four or more sections that are pulled together by an elastic cord when the cane is allowed to fall from your hand. My favorite of these is the AFB "SuperFold." (You may have to order this quickly, since the Foundation itself has spread the rumor that they are going out of business in the area of "aids and appliances.")

All such canes are fitted with my pet peeves: they have rigid tips (mostly nylon), and their bulbous handles make you look like you are carrying a collapsible weapon. [Bay Area police forcefully subdued an innocent blind bus passenger who extended his automatic weapon in a confident way; it sure hit the news out here.]

For durability, most of these models are the same diameter over the entire length. In my opinion, durability requires this. However, the resultant instrument feels much heavier than tapered solid or telescoping canes; their center of gravity is lower.

When I purchase folding canes (usually four at a time), I buy a few extra plastic tips. These can be turned on a lathe to create a 3/4-inch-long peg of 9/32-inch diameter. This peg accommodates a press-on rubber-mounted steel tip intended for canes of the Rain-Shine umbrella company (my favorite solid cane, by the way).

In order to modify the handle, it is necessary to take the cane apart. Such canes are made in various ways, so specifics cannot be given here. Suffice it to say that in most cases, the elastic should be temporarily removed.

Taking a knife to that rubber defense device on the top piece, I strip down to the bare aluminum (giving that delectable morsel to the neighbor's dog). My choice is to replace the rubber with heat-shrinkable tubing; you need some insulation there in cold weather.

You have two choices of handle styles. One is to fit the top end with a "crutch tip" from the hardware store and provide a loop an inch or so from the top. The other is to allow a loop of the main cord to emerge through the top, as many already do. The latter arrangement is simpler, but it is not my favorite.

[By the way, the purpose of such a loop is to hold the sections in a bundle when the cane is collapsed. Never!, ever!, think of this loop as a "wrist strap," as it is sometimes advertised. If your cane gets in a jam, such as getting trapped under a wheel of a vehicle backing up, you've got to be able to instantly disengage yourself from the cane and move to safety. Any suggestion that a cane has a wrist strap is irresponsible thinking.]

The rubber handle may have been the mechanism for holding the top end of the elastic cord; this is true for the Superfold, which uses the small hole in the end of the rubber to trap a knot in the cord. Often, the top section is just a piece of tubing, perhaps flared at its lower end to mate with the next section. To trap the main elastic, I borrow a trick from the old AFB "Aluminoid Cane" as follows:

Find, or make, a washer whose diameter is a close, but loose, fit in the tubing. The hole in the washer should be large enough to accommodate the cord. Perhaps halfway down the tube, make five or six deep dimples around it with a center punch. A knot in the cord will keep it from pulling through the washer; the dimples will trap the washer at that point in the tube.

In order for the washer to seat properly on the dimples, they must be on the circumference of the precise cross section; you don't want the washer to cock at an angle such that it could flip on edge and pass through the restriction created by the dimples. For this reason, wrap tape around the tubing to serve as a guide for the center punch.

Now you can chose your loop configuration. If the elastic has a loop which, before, emanated from the top, you can position the retaining knot such that the loop still does so. I prefer to use smaller elastic cord to make a loop 2 inches down from the top.

To accomplish this, I drill the diameter of the tube at this 2- inch distance. Eventually, I poke the two ends of my small elastic into the tube from either side, fish the ends out the top, tie the ends together in an "over-hand knot," and pull the loop so that the knot disappears into the tube. (The main cord has to have been installed first, and the shrink tubing must also be applied first.)

Once the top section has been dimpled and drilled, heat- shrinkable tubing can be installed. Most American canes have an outside diameter of one-half inch; the appropriate shrink-tubing size is 5/8- or 3/4-inch. (See "Using Heat-Shrinkable Tubing," SKTF, Spring 1983.) This tubing can be gotten in various colors; it is my inclination to stay away from black or gun-metal blue, as I don't want a beating like that other blind guy got.

The tubing should reach the bottom of this section; however, its position at the top depends on your style of loop. For example, for the arrangement where a separate loop is positioned two inches down, you will want bare aluminum for the top three- quarters of an inch so that a crutch tip can be fitted to the top. If your loop is intended to emerge from the top, I recommend the following procedure:

Thread the washer onto the cord; then feed the cord through the section from the top end. Determine the position of your knot so that the washer traps the elastic with the right amount of loop emerging. Making sure that the washer stays somewhere in the tube, pull the cord out again until the knot is an inch or so outside.

Cut the heat-shrinkable tubing one inch longer than the section. Match the bottom ends up, positioning the extra length at the top. Apply heat so as to shrink from the bottom up (turning the assembly so as to distribute the heat). When you get to the top with the heat source, shrink the extra length; it will shrink down to perhaps 5/16 inches. Then, while the shrink tubing is still warm and malleable, pull the cord so that the knot folds the shrink tubing inside the aluminum tubing. This will produce a finished end which won't hurt your palm when the cane encounters something.

Some canes may not require all of these steps. For example, the Superfold cane has plastic fittings at the joints; the one in the top section is perfectly adequate for retaining a knot in the cord, and all that is necessary is to position the knot further down. Depending on how far the modified handle piece has moved the position where the elastic is retained, you may find it necessary to shorten the cord to get the tension back up where it belongs.

If holes in the handle have been drilled (for installation of a loop, for example), a tapered reamer is the best tool for opening the shrink tubing, although careful work with a pointed knife would do as well.

Earlier models of canes had few sections--four being typical. One of my tricks was to order five shorter canes, use one for parts, and make four long ones having five sections. You can shorten the bundled-up cane by two or three inches that way.

Notes on the AFB Superfold

When it works, the Superfold is my favorite folding cane (see the first paragraph of the previous section). It has Delrin fittings at each joint that carry a pair of rubber O-rings. When pressed together, the cane does not wobble at the joints, and it does not rattle when in contact with rough terrain.

These O-rings make it possible to make a 6-section cane that outperforms canes with metal-to-metal joints. Unfortunately, these joints are somewhat vulnerable to breakage. Also, fittings can come loose of the pieces they are supposed to stay with.

Out of four Superfolds, I got three joints (15% of them) in which the plastic fittings had to be cemented into the aluminum with SuperGlue.

Right out of the package, the Superfold is very hard to press together and disassemble. The O-rings squeak and refuse to seat properly, making a wobbly assembly and causing internal hemorrhages when pulling them apart. There is an easy solution for this, and I'm surprised it is not done at the factory.

The same snugness of fit can be had if the O-rings are "glazed" with Vaseline. Generously smear each O-ring (two per fitting) with Vaseline; then, press the joint together and twist it back and forth a few times. Next, pull each joint apart and wipe it off with a paper towel.

Both the O-rings and the Vaseline are petroleum products, and they "take to each other." The glaze will remain, even though the fitting will not be greasy enough to soil your pocket. (They could glaze the batches of O-rings before they put them on; they would go on easier and it would save us the trouble.)

On the Appropriateness of Folding Canes

If you are a long-cane traveler, your cane is most important for your safety. You are more valuable than an inanimate stick, so if a stick must be sacrificed now and then, so what. On the other hand, it must be there for you when the going gets rough.

I know people who swear by the most dainty of canes; they say, "I only have to replace it every once in a while," or, "You just have to take care of it." Balderdash! Codswallop!

When I was a kid, there were product advertisements that said, "This makes an excellent 'theater cane.'" Pardon my opinion here, but a theater is full of oddly spaced tiers, seats that can trap your cane, people who, with a small misstep, can render you caneless--unforgivable.

If you need a cane, you need a cane. Travel styles and situations differ, however. I travel in benign environments--smooth, wide sidewalks--most of the time, and I am fairly gentle with my canes; I usually break them when I am abrupt with them. Some of you live in towns with no sidewalks--perhaps with car after car parked in your way (the bumpers of which eat folding canes). Still others are heavy handed with canes; they don't solder for a living, so why shouldn't they be.

When you test drive a cane, solid or hollow, folding or not, put thought in your confidence in it. If your cane is too delicate, you will travel as if you are protecting it.

Protecting your cane is dangerous. In bailing it out of trouble, you can side-step and fall off an edge, you can unexpectedly back up and be hit by a quiet bicyclist, or worst of all, you might be avoiding putting it where it best belongs--taking the hits for you.

I need subscribers. I cannot afford to lose you, so do me this favor. At the slightest inkling that you might be favoring your cane, cast it aside and choose a mightier staff.

Products and Numbers

The AFB Superfold is available from the American Foundation for the Blind Product Center as the C601xx (where "xx is replaced by the length; my choice being the C60156). Extra tips are No. C61100. AFB Product Center, PO Box 7034, Dover, DE 19903-7044; Phone: (800) 829-0500.

The telescoping fiber cane is an NFB product. The length in inches follows the prefix ACG, with the suffix T meaning it telescopes. Thus, mine would be an ACG57T. Order it from the National Federation of the Blind Materials Center, 1800 Johnson St., Baltimore, MD 21230; Phone: (410) 659-9317.

My favorite tips (and canes, for that matter) come from the Rainshine Company, PO Box 5615, Madison, WI 53705; Phone: (608) 437-8018. Their canes are noncollapsible solid fiberglass, and I still have the one I got in high school. It has survived being tripped over and run over by trucks, as well as being bent in a 20dg arc so as to stuff it into my gym locker hundreds of times.