JDR Microdevices :: Compendium

The Compendium is a collection of Derick's articles from 1988 through 1993, compiled into one volume

If you don't find it here, maybe it isn't worth knowing about (to copy a phrase).

Over the last couple of decades, I have become more and more dependent on the information contained in the IC Master. The sheer volume of new IC's every year continues to swell the pages, and broaden the diversity of available products.

When I begin a new design project, I want to know that I have considered all of the newest and best available devices for possible inclusion. That quest would be much more difficult without the IC Master.

Within the pages of the IC Master are several indices and cross reference listings. With just a little experience, you will soon find that locating a specific device type or part number is very easy. On more than one occasion, I have begun a search without any foreknowledge of whether such a part may or may not exist.

Happily, I can report that I have never been disappointed in my searches, unless you count the time I abandoned the idea because the idea was too costly.

P.S. -- If you have a CD-ROM drive, you may wish to consider the CD version. Searching for an entry is faster, and the CD takes much less space.

Since the decline of do it yourself electronics in the late 70's, many otherwise curious people have been discouraged from kit building.

Most alarming to me is the number of people who graduate from technical schools and degree programs without the experience of seeing a device of their own construction work. Learning to identify components, solder, and test an inoperable device is a great experience. If you have an interest in electronics, and have never put together a kit, let me encourage you to give it a whirl. The investment is minimal, but the satisfaction (and bragging rights) can be tremendous.

P.S. -- Electronic kits make a great gift for a young person interested in science or searching for a career field.

Should you build your own power supplies or buy off the shelf? I have some advice, but it's free, so consider the source.

I would encourage you to seriously consider building your own low-power low-current supplies for your next project. Integrated devices for power regulation simplify the task and promote success.

For high current requirements, I buy off the shelf. OK, so I'm chicken! Would you be surprised to know that I had been reading a Dr. Seuss book the evening prior to writing this column? Sometimes the titles are more interesting than the column that follows!

Some people take a good thing too far. I have small parts in my junk box that are older than I am! You never know when you'll find a use for an old NIXIE tube (removed from ancient military gear).

In reality, my junk box is actually a group of six small parts cabinets ranging in size from 9 to 50 drawers. In them, I keep an excellent selection of resistors, capacitors, small signal transistors, and a few inductors as well as a declining selection of TTL's and an expanding choice of PAL's and GAL's.

In my experience, I have found that I do a good job of anticipating the components I will require for the major functions of a new design, and frequently overlook one or two of the "incidentals." The time and money I have saved by having a well stocked junk box is far greater than the value of the parts it contains.

And, it has been a significant factor in many small repair jobs over the years too. Last month, a "junk" 40MHz oscillator saved the bacon for a small company's payroll.

JIT, Just In Time, is a wonderful idea if you are an accountant! Give me a choice, and my junk box will be ready ATT, All The Time.

Thermal shock, sustained heat, voltage spikes and voltages out of tolerance are the most dangerous threats to reliability in electronic equipment. Therefore, minimizing those threats should be high on the priority list during design and construction.

What can you do if you are the end user of a piece of equipment you suspect may have been built without those goals in mind?

First, make sure that existing fans are operating and clear of obstructions, including lint and dirt. Next, check for "hot spots" inside the chassis. If you find components that are too hot to touch, they are too hot! Clip-on cooling fans for microprocessors are an excellent idea.

On the power side, move delicate electronics away from fluorescent lamps, heavy motors, and other "noisy" equipment. If possible, add an EMI filter to the device or insert a surge suppressor between the plug and device.

And last, but not least, don't overlook the importance of grounding. Nobody likes to draw an arc when they approach a metallic device, but what only startles you can destroy many electronic devices.

Good wire wrapping technique takes practice and some knowledge of what makes a good wrap. A good wrap is more than good electrical contact. It should be without shorts and last more than fifteen minutes after the job is done. For a wrap to qualify as good, these are the minimum goals.

1.A gas tight connection between the wire and the post where the corners of the post cut into the bare wire. This is to prevent oxidation and later, a faulty connection.

2.Moderate tension on the strung wire so that it neither flops around nor bites into an unintentional connection. Hard bends also increase the chances of breakage.

3.An extra wrap of wire with the insulation at each post. This lessens the chance of the insulation sliding and exposing a short, and takes strain off the last three or four corners of wrap to minimize metal fatigue and breakage.

Using the right equipment will increase the chances that your wire wrapping is acceptable. Choose a wrapping tool that provides a "modified" wrap and then practice on scrap pieces before working on the real thing.

In 1982, I designed my first printed circuit board. It was a full length, double sided board with space for 256Kb of memory, a game port, a serial port and a clock. The design took about two weeks while the layout on clear mylar with tape at 200% took me about two months.

While I can count it as good experience for a complete novice, I will never ever do it again. Hour after hour, until my eyes gave out in fatigue, I would lay down tape only to have to pick it back up again later in order to route another circuit.

In 1983, I purchased my first board design software. It supported both schematic capture and board layout. Finally, I could do the drudge work as quickly as the initial design.

Design software for PC boards improves your ability to see the big picture and develop a concept one piece at a time. Having a professionally printed schematic helps expose errors and encourages logic simplification.

Being able to shuffle components around and trying different arrangements helps to minimize board space and create better looking boards. Later, when you want to make additions or modifications, having an electronic easel makes the job easier and neater.

I am particularly stubborn when confronted with problems having to do with computers. Programs that hang and circuits that don't work properly intrigue me and I have trouble letting go!

Hanging a scope probe on a circuit comes naturally to me; I have been doing it for years. Unfortunately, scopes have also led me down the garden path many times while working on computer problems. You just can't see enough with two or four probes.

That is why I am so enthusiastic about the LA16PC logic analyzer. Not only can I view 16 signals at a time, I can even "look back" to see why various traces are in the state they are in. I have used better analyzers, at 15-20 times the price, but none that were easier to operate or more suitable to the circuit speed and pricing that I can afford.

Since it doesn't require a state of the art PC, even an 8088 system makes a good "Logic Analyzer."

Even the fastest EPROM eraser takes much longer to erase a device than it takes to burn one. If you seldom require an eraser and never erase more than a dozen chips at a time, even the least expensive device will do the job in a couple of hours.

However, if you frequently find yourself working with dozens of chips or erasing chips on a daily basis, then you will want to consider a more capable device.

Higher power units like the PE-240T and PL-265T erase faster, and have a timer that will shut the unit off when erasure is complete. If your volume is smaller, then the PE-140T will take a little longer, but still has a timer with shut off.

All of the erasers JDR sells automatically shut off when the chip drawer is open, preventing UV light from escaping.

The phone woke me with a startle. The "passing acquaintance" on the other end was blubbering something about his good deal HumDinger printer not working when he connected it to a serial port.

To be honest, I didn't want to help Mr. Sponge. I knew that no matter how long it took me to resolve his dilemma, the most I would receive would be a grudging thanks. I also knew he would persist until I gave in, so I packed my tool kit and headed for his home office.

When I arrived, I found he had no manual for the printer and the printer was so old I couldn't even guess at its age--bad news!

Using a breakout box and a manual for the PC interface, I had it working in about 10 minutes. First I attached the breakout box to the printer, leaving the PC unconnected. Then I looked at the LEDs on the box that were illuminated. From them, I knew which lines were being driven from the printer. After that, it was a piece of cake. I was done so fast my acquaintance didn't have time to think up anything else for me to look at.

Fortunately, the person I refer to in this column no longer knows where I live, and he hasn't bugged me in a couple of years. Time heals all wounds, and I anticipate complete healing in about 50 more years. By then I may regret I used his story.

(I say that, but the reality is that I would probably come to his assistance again if he managed to get past my "perimeter defences".)

It was so long ago that I could be mistaken, but I think I bought my first low cost EPROM programmer from JDR in 1986. Before that, I had always built my own for each specific chip type I wanted to program.

I'm amazed by the versatility, reliability, and ease of use that has marked the evolution of low-cost programmers. The complexity of supporting so many different manufacturers with different pin-outs, voltages and timing requirements almost boggles my mind.

The MOD-MEP-1A and -4A represent the best in low cost EPROM programmers. You could spend hundreds (even thousands) more for a programmer that programs a few more chip types, but why would you unless it is really necessary?

At these prices, why have one programmer that is "down the hall in the lab," when every developer can have their own?

I would like to focus your attention on the programmer section called PLCC/QFP Converters. If you are like many readers of this, or any catalog, you probably look first at the page headings, then at the column headings.

If you are already using another brand of programmer, perhaps one that your company specified, you may think you are stuck with that brand's expensive accessories.

Not so if you are merely talking about converting an existing dip socket to PLCC or Quad Flat Pack. Our selection of converters includes the most common EPROM, PAL, GAL and microprocessor sockets which can be used on any brand of programmer.

These converters with Zero Insertion Force sockets can save you hundreds of dollars with each purchase.

The title here was supposed to have a couple of musical notes behind it, but somehow they never made it into the catalog. Do me a favor and sing the title when you read this article.

Experience is a great teacher, and one thing I have learned after building hundreds of prototypes is that I make more "careless" mistakes than errors in basic design. I doubt that I am alone in this observation.

All of my prototypes are built using socketed parts so that the debug is never slowed by questions about a part that cannot be easily removed. And, I prefer either point-to-point with soldertail sockets for small projects or wire wrap sockets for larger projects.

If I have to wire wrap 4 pins in a net, I wire A-B, C-D, then B-C. Until you learn better, most people would wire A-B, B-C, C-D. While the electrical connection is the same, if you ever need to remove the A-B wire, you will have to remove the other two first.

ID Wraps are more than a luxury. They will save you hours of debug time by insuring that you don't miswire an IC in the first place.

And one last helpful comment: If you are working with programmable parts that will be on and off the board many times, invest in a ZIF socket for those parts. When the prototype is complete, you can always remove the parts for later use. (Note: Get a wire wrap receptacle for each ZIF socket. The ZIF socket isn't on 0.1" centers, and you shouldn't solder to it even if you could, but you can't!)

A nearly universal complaint of technical people these days is the control placed on them by management, and particularly by accounting management.

Along with JIT (Just In Time) inventory management for production came the edict that the whole company could benefit from better management of inventory.

Absolutely Right_And Dead Wrong! Some departments are stifled by too much control in the wrong places. One example is the control of small parts for prototyping or test/repair.

When technicians have to fill out requisitions and then wait for them to be filled so they can determine IF a part is bad, minutes or hours are wasted trying to avoid the requisition hassle, and then more time is wasted following through on the procedure.

Every department should have a collection of commonly used parts readily available for use. Where appropriate, a sign-out sheet should be used to restock the parts cabinet. These parts are usually costed out when they go into the cabinet because the value of the inventory is worth less than the cost of maintaining strict control.

I used to work for a VERY BIG computer company. In the beginning of my computer career with them, things were good. Replacement parts were easily available, and production glitches were never the result of poor internal management. The workforce was happy, and we felt good about the product.

Then one day there was a significant change in the makeup of upper management and we learned that the bean counters were in control. Just In Time inventory had arrived. Within a couple of months all locally stocked replacement parts were removed to a "controlled" location, and we began experiencing the results of NIT (Not In Time).

Those NIT's became a real sore spot with the production and sales organizations, but I imagine the bean counters were happy. Soon we were testing products, pulling out the "parts short" portion, installing it into the next machine only to repeat the process.

After a few months of NITwit production, we had hundreds and hundreds of almost finished machines crowding every aisle and empty space in the interconnected campus-like buildings. When the parts finally began to arrive, we worked intense overtime to catch up with our delivery commitments.

Today that company no longer enjoys the dominant position it once had, and has in fact been on the decline since the change described above began to occur. Of course they would point to world competition and the changing market place as the cause, but I think it is significant that people no longer think it is a good place to work.

Got troubles in your life? Solve them quickly with a microcontroller. While I admit this only works for a certain type of problem, it can be just the ticket for many complex tasks.

Once you have used any microcontroller, you will quickly find other applications where your previous experience can pay off again and again.

Intel's "Embedded Applications" handbook is ideal for selecting the right micro for the job at hand. Inside are example circuits and programs for many of life's little problems. Just a quick scan of the topics reveals micros in applications from printer controller, graphic processor, software serial port, DC motor controller to Analog/Digital processing and FFT algorithms.

Many board designs today are based on, and revolve around, the use of PLD's (Programmable Logic Devices). The simplest PLD's are the Programmable Array Logic and General Array Logic groupings.

While many are used for what is called TTL "glue logic" replacement, their versatility and "late in the design" reprogrammability have made them the designer's friend.

PAL's come in a variety of configurations, each of which is tailored for a narrow range of applications. For example, one PAL might have D-type registers for a kind of byte latch, while another has 2 registers and 6 multiple inputs AND/OR gates.

A GAL is very similar to the PAL with several significant improvements. Most important to me is the ability of a GAL to be erased and reprogrammed (just in case I make a mistake). Most GAL's are also low powered, and all GAL's have the ability to mimic the ability of a whole group of PAL's.

This means you only need a couple of types of GAL's in your spare parts location, and this will probably save you money.

That was my mind ricocheting. Acronyms like PAL and GAL conjure up friendly images, but what does ATM mean to you?

The new Logical Devices CUPL starter kit for PAL development software is quite impressive. I had an opportunity to use version 4.2a this weekend, and some of the enhancements are things I will no longer be able to do without.

For me, the most notable change is a customizable "shell" program that acts as an executive for choosing and running programs while developing a new Programmable Logic Device (PLD).

Normally, I am resistant to "shells" because they interfere with the way I think things should be done. MCUPL helps me do things the way I want to get them done, and serves as a gentle reminder to make sure I do all of the things necessary to create, document and use new PLD designs.

After editing MCUPL.CFG to configure MCUPL for my specific machine, editor, and function hot key preferences, I started MCUPL and saw a list of choices on the left of my screen. Logically, the first choice was "Edit a Design File." Quickly, a list of device designs appeared in the box and I cursored to the one I wanted to edit. Editing completed, I quit my editor and snapped back to the menu.

Selecting a second option "Compile a CUPL Design," was a pleasant surprise. MCUPL remembered which file I had just edited and chose it as the default for compiling. After a quick "Enter," I was led through several options that in the past always took me back to the reference card.

While I watched MCUPL build a command line at the bottom of the screen, I was able to select minimization level, listing formats, output formats, and even the target device without feeling I could have done it faster or better on my own.

The new CUPL starter kit is a fine way to test your interest in using PLD's. The nine devices it supports are among the most frequently used PAL devices, and give the new user an excellent opportunity to learn a new way to do things.

The first time I soldered, there was a 15 year veteran looking over my shoulder (we're talking military vet with many stripes!) I was already apprehensive, and didn't really need his attention to make me nervous about the job.

We had just finished a three day "How to Solder" course, and now were getting our first actual experience. After 30 hours of instruction, the last advice he gave was all I have ever really needed:

"Get it clean, get it hot, put the solder on the job, not on the iron, and remove the heat when the solder flows where it should."

If you use a good rosin core solder, and a decent iron, that's really all there is to making an acceptable joint.

(Don't all you experts holler at once! I know all about making a mechanical connection before you start, etc., etc., etc. We're not talking about a space mission!)

I'm often asked if I wire or solder my prototypes. It will surprise many that I prefer point-to-point solder for even multiple chip designs. When the prototype is finished, I spread a layer of hot glue over the back of the card, and make it nearly permanent.

For many hardware developers, PLD's are the answer to a problem they didn't even know they had. It is only after using them in a project or two that the engineers fully recognize just how revolutionary they really are. Ignore for a moment the board space saved, lowered power requirements and faster propagation of signals_I'm talking about making the job easier and more fun.

Frequently programmers are given credit for "artistic" and "creative" skills when designing a program. With PLD's, hardware designers can exercise the same "artistic" abilities.

Instead of wiring the output of one three input NAND gate to the input of a latch, etc., you now have the freedom to think at a higher level and allow the PLD compiler to route most of the wiring.

I still have shoulder cramps from patting myself on the back over the design of the PCODE. I could have created the same functionality using a double handful of TTL parts, but I did it with just two GAL's.

I encourage you to do your initial designs using GAL's. Not only are they more flexible than PAL's, they are also more economical for initial designs because they can be reprogrammed. Your initial high level description can contain errors without incurring the cost of a programmed part that has to be thrown away.

In my own designs, I usually use GAL's until I'm ready for production. Only then are the GAL's replaced by PAL's where it is possible.

The PCODE is such a simple looking card that a couple of close friends have taken me to task for being so proud of it.

It is and was simple, once I decided to use the two GAL's. What made it a neat trick was the circumstances that brought it to life. Here in a nutshell is what happened.

On a Friday afternoon, Chip, one of the buyers at JDR asked me to look at some POST code display cards he had in his office. The lowest cost device was $100 wholesale and one of the cards would have cost us over $300. I was incredulous. The $100 dollar card didn't even use 7-segment displays. All it had was eight individual LED's. I told him they were overpriced and went away steaming.

Over the next two days I designed, prototyped, tested, and laid out the card for photomasking. On Monday morning I gave Jeff Rose a printout of the plans and explained their purpose. (The first iteration worked, but two revisions were required before we really hit our stride. In my haste I did the initial design without finding out the availability of some of the parts I used in the prototype.)

ISHDTTFT: I Should Have Done That The First Time! How often have you said that to yourself? For me, it's a common thought. I built my first floppy disk interface from scrap and surplus parts. The second time I did it, I bought state-of-the-art parts. ISHDTTFT!

For years, I struggled with soldering irons that were either too hot or too cold, and desoldering equipment that wasn't up to the volume of repairs and mistakes I encountered.

Then I bought a real solder/desolder station. Suddenly, I could regulate the temperature of my soldering iron. Copper traces didn't separate from the board. Ground planes no longer soaked up all of the heat from my tip, and I didn't have a clutter of irons, one for each task, lying around. ISHDTTFT!

Desoldering turned out to be the feature I appreciate most. It's nice to know you can remove a 40-pin IC without damaging either it or the board. I could never count on that before. ISHDTTFT!

Once upon a time, batteries "always" leaked when they were drained. And, a brand new battery in a toy car would drain in about 1-2 hours of steady play. I'm old enough to remember throwing away flashlights because the ruptured battery was fused into the cylinder so tight you couldn't remove it.

It's a lot better today, but I don't think we can take that power source for granted yet. Just wait for "Atomic" batteries. NiCad batteries are rechargeable, but require considerable power-on time to fully charge with the trickle current on most PC motherboards. In addition, many have a memory for how much they are willing to discharge based on "normal" usage. If you always use your PC for 3-5 hours per day and leave it off over the weekend, the battery will "remember" that and may refuse to keep your time-of-day clock working when you go on vacation.

For long term storage and fewer problems, I generally recommend a Lithium battery replacement at the first sign of battery difficulties. A Lithium battery will generally last for 6-10 years, and doesn't display the "memory" problems of a NiCad. Don't recharge them.

Of course, it costs more, but for now, it is the most trouble free solution to battery requirements for a PC.

Selecting new products for inclusion in the JDR product line is one of the more interesting tasks we tackle throughout the year. Here are some of the new products that I particularly liked, and think are of interest to the electronic hobbyist, technician, and engineer.

NS16550: The latest in a series of serial port control chips. Contains a 16-byte FIFO buffer. If two control lines are ignored, the chip is pin compatible with existing 8250 and 16450 Universal Asynchronous Receiver/Transmitters.

MAX232CPE: Maxim's +5 volt powered dual transmitter, dual receiver for RS-232 serial applications. Includes its own charge pump to create +12 and -12 volt supplies using only a few capacitors.

ICL8211CPA: This eight-pin chip can be used to create a power on reset signal, warn of a low voltage power interruption, or degate write signals during power up/down. Requires only three or four resistors to set threshold and hysteresis values for detection.

ICL7663SCPA: This is a positive voltage regulator. It operates with an input voltage of between 2.0 and 16.0 volts with current controlled to 40 milliamps. Very low quiescent current and few external parts make these parts ideal for battery powered applications.

ICL7665SCPA: The 7665 is a dual over/under voltage detector. While similar to the 8211, its advantages show up in various warning applications. Use it when you want to know about Low Battery, Power Fail, Battery Backup Switching, Hi/Low Temp, Pressure and Voltage Alarms.

16V8 / 20V8: Use a Generic Array Logic part in nearly any location where you could use a PAL. Because it is reprogrammable, you can experiment with various solutions without throwing away the parts when your experiment turns into a mistake.

8751H: The 8751 microcontroller is an easy to interface 40-pin microprocessor with nearly all of the normally external peripherals built in. This part really shines in embedded applications. It has a programmable serial channel (to communicate with a PC), two 16-bit Timer/Counters, 128 bytes RAM, 32 I/O lines and 4K bytes of EPROM. Port 0 is an open drain and can sink up to eight LS TTL loads.

Several major PLD suppliers now offer logic compilers for under $100. In fact, National Semiconductor offers their "PLAN" logic compiler for free to users of their devices. Thus schools, technicians, hobbyists and engineers can work with programmable logic for under $280.00!!

After you tell the logic compiler to assign signals to pin numbers, the next step is to tell the compiler when you want the decode to become active. This is usually done in a simple text file:

Registered outputs are a little more complicated, but even the example above shows the power of the PLD. After you create the text file, process it into a fuse map. The command to do that might look something like this:

The compiler reads the file DECODE.FIL and processes it to create a standard JEDEC file fuse map for burning into a PLD called a 16L8. Lastly, run the fuse burning software. Simply read in the file DECODE.JED, select the proper device (here a 16L8), and put the chip in the Zero Insertion Force socket.

Once you're familiar with the process, even a moderately complex device can be described and tested in less than an hour.

One last thought. Take a look at JDR's new PCODE Display Card. How many TTL chips do you think it would take to duplicate the function in the 2 GAL20V8's on that board? I estimate about 40 TTL chips, but I never finished the schematic--It would have taken too long.

Have you ever wished your computer could do some small task and then found out that the PC card doesn't exist? Or, if it does, it costs more than you want to spend and does more than you need?

Even if you have never built anything electronic before, don't assume it is beyond your abilities. If in particular, you have written programs for your computer, you do have the basic logical abilities needed.

Books like the "TTL Cookbook" (SAMS-TTL) and "Interfacing the IBM PC" (SAMS-INTRFACE) make the task as easy and safe as possible.

TTL chips and microprocessor support IC's are very inexpensive. As long as you do your experimenting on the "buffered" side of your bus, the most you are likely to damage is your ego and perhaps a few cents worth of IC's when you plug a part in backwards.

JDR sells a variety of microprocessors, and some of them, like the 8088, come in several different speeds. If you are purchasing a motherboard or peripheral that uses a microprocessor, the selection of chips has already been made.

However, if you are about to build your own computer or intelligent peripheral, how can you decide which micro to use? You cannot decide by looking only at processor speed, because an 8 MHz 8088 processes data much slower than a 6 MHz 80286. And you cannot rely on prices to show you which will get the most work done either. So what should you do?

I have a suggestion that will work in most cases. It is based on the notion that if you don't already know what to use, then what you probably need most is something that is easy to program, not necessarily something that is fast. A Z80 might be the best choice for an intelligent hard disk controller, but if it were, chances are that your experience with microprocessor hardware would supersede my advice.

So what is a good chip to use in building an intelligent controller? How about the 8052AH with BASIC? Not long ago I had a need to test some of these chips. In one evening (using the circuit suggested in the data sheets) I built a quick and easy parallel port buffer. The buffer wasn't particularly fast, but then I didn't need it to be. It was a whole lot faster than my printer. I wrote the control software entirely in BASIC (please don't laugh!) and even with a couple of minor bugs, had it working the next day.

Now, what do you need to control? Intrusion detector, automatic lawn sprinkler, or maybe something a little more sophisticated like a solar hot dog cooker? If you're looking for fast, do your homework, and if you're looking for an easy solution, try the 8052AH, you'll like it.

In the last paragraph I mention a "solar hot dog cooker". That reference has special meaning to a friend named Rick and myself. A few years ago Rick and I "cooked" up the idea of using a "Solar Hot Dog Cooker" to roast weenies for a JDR picnic. We went through elaborate preparations to place a huge "Solar Panel" atop the roof of JDR, and ran the wires down to a device that would zap the hot dogs.

As a hot dog cooker, the device was a near flop, but we never really cared about that anyway. Our primary goal was to convince everyone that the "juice" was being generated by the "solar panel". In that we succeeded. Only later did we reveal that the huge "panel" was just a sheet of plywood pointed at the sun, and that the "juice" was 110 VAC.

I've been looking for a solution to low cost diagnostics for quite some time now. As a consultant, VAR, or systems engineer, one often runs into a situation where the PC you just put together won't boot up.

You can try and make sense of the beep code (it will tell you up to 16 failure points if you know how to decipher it), but based on the Power On Self Test, you have more options for diagnosing problems than the beep code can report.

That's why I've devised the PCODE. It's a compact card that plugs into any open slot, reading the data written to I/O port 80 and displaying its results in a hexadecimal code on two 7-segment displays. The data displayed translates to an exact point of system failure dependent on the system's operating BIOS.

There are much more expensive tools that do about the same thing, but for economy, flexibility, and minimum power draw, this is the best solution I've seen.

The 7400 series of TTL (Transistor Logic) devices has greatly benefited from its 28+ year family history. Various branches of the family tree have been optimized to handle all kinds of applications. There are power-stingy slowpokes, power-guzzling speedsters and compromises in between.

While the 74LS series is the largest, least expensive, and supports the majority of applications, some applications may require the special features of other members of the family.

74xx series parts can be intermixed within one design because their input and output requirements are essentially the same. Only the drive capability of certain members of the family will normally prevent mixed designs.

Low power parts driving high speed (high power) parts require the most attention. For example, a 74LS part can only drive four 74S devices without degradation. Conversely, a 74S output may drive as many as fifty 74LS inputs.

Family	Average Delay	Maximum Frequency	Typical Power
54/74	10ns	35MHz	10mW
54LS/74LS	10ns	45MHz	2mW
54ALS/74ALS	4ns	50MHz	1mW
54H/74H	6ns	50MHz	22mW
54S/74S	3ns	125MHz	19mW
54AS/74AS	1.5ns	175MHz	10mW