Electronic Prototyping: Tips and Pitfalls

This page will discuss electronic prototyping and construction methods. It was originally written in November of 1994; obviously much has changed in 14 years. Please note that many documents accessed through this page are taken from postings on the Usenet newsgroup "sci.electronics" and the information is the personal opinion of the person who posted it. We do not endorse or accept responsibility for any of these postings and provide them simply as potentially useful advice. Many thanks to Fil and his sci.electronics FAQ archive, referenced through several links here.

The on-line Electronics Design Checklist compiled by Hank Wallace compiles a wealth of suggestions from electronic designers. Check it out!
7-23-1998 - Superb articles on debugging embedded systems are found at The Ganssle Group website.

Index:[Time] [Basic Tips] [Breadboards] [Wire Wrapping] [Perf Boards] [Printed Circuits] [Data Books]

Time

Based on the projects we've done as research technicians, and on the years of student projects we've observed, the greatest miscalculation students make when working on electronics projects that actually have to work is how much time it yakes. While people may allow enough time to design and build a project, they often don't realize that even in a really good project, the design and construction will probably be only half of the time required. It will probably take as much time to debug, fix, and/or redesign as it did to create it in the first place. It's very rare for new designs to work right off the bat, due to

Mistakes in construction - shorts, faulty connections, wiring mistakes
Errors in design - even if it is wired perfectly, you may have flaws in your basic concept.
Need for additions - you may not be able to tell until you test it that there are elements that have to be added to make it work as you want.

We have seen the blithe assumption that a project will work as per design without significant troubleshooting threaten the graduation of a student with a job waiting; don't let it happen to you. Budget yourself plenty of time to debug and re-engineer your project.

Basic Tips

These are a few other things that apply no matter what prototyping technique you choose:

Assume you will have to make changes. Whether it is correcting design flaws, fixing mistakes in wiring, adding extra circuitry, tweaking component values, or re-engineering your entire concept, the odds that you will NOT have to make any changes to your circuit are very close to zero. So choose methods that allow for easy changes, leave yourself enough space on your boards, and don't set anything in stone until you are certain it's working how you want it.

Account for ALL your pins. Whether or not they appear on your schematic or in an example circuit you are borrowing from, make sure you take a look at the data sheet for each IC and be sure that every pin is appropriately connected if it needs to be. The classic mistake here is the student who can't figure out why his op-amp circuit doesn't work, when they didn't bother hooking up the + and - voltage supply pins that didn't appear on the schematic. Besides power and ground connections, there may be chip enables, clocks,resets, and other similar inputs that have to be made happy before things will work. Unused inputs on extra gates should be connected to ground or the logic supply: left unconnected, some kinds of gates can oscillate and cause wierd problems. Unused outputs can be left open, as a rule.

Despike your chips. This means putting a .01 to .1 uFd ceramic capacitor from the +5 volt supply to ground right at the chip. power line spikes occur when there are sudden current changes far from the power supply, and it's amazing how much trouble this can cause.

Make good power supply and ground connections. A skinny, daisy-chained wire-wrap connection from chip to chip is probably not going to be enough. The more robust you can make power supply and ground leads, the better.

Keep your digital and analog circuitry physically separate whenever you can. Digital switching, especially at microprocessor buss or video card speeds, can throw all sorts of noise and trash into analog or audio circuitry.

"It's must be a bad chip" . . .NOT!!! When a circuit design doesn't work, the first impulse is to blame a failed component because we just know the design and wiring are right. In practice, it is really quite amazing how seldom the chips are at fault, and how much abuse (wrong wiring, wrong power supplies) many common chips will withstand without damage. (DO watch that static electricity, though.) The problem is almost invariably somewhere else. (One exception: if you put a standard EPROM in a socket backwards, which is easily done, and apply power, it's gone. Throw it away.)

Don't use silicon sealant to mount wire-wrap sockets or to seal/insulate circuitry. This stuff is handy and common but it is not an insulator. It will leak small currents, which may not matter in logic circuits but can wreak havoc in high-impedance analog circuits.

Build one whole device first, if you are intending to make several identical units. It's tempting to save time by, for example, drilling all the chasses for all the units while you have your drill set up: soldering all the boards at once; etc. etc. But if you then find you've got to undo or redo something that didn't work the way you thought it would, you've multiplied your mistake across all the units. Finish one unit completely to find all the mistakes and optimize the design, then go into "mass production."

Make sure you can get the parts before basing a design on it. You may find the ideal integrated circuit for your application in a data book, but it may not be in production, may be unavailable from distributors, or may be too expensive. Especially if you are creating a design you hope to produce for a while, it's wise to choose devices that are widely available and that (you hope) won't be discontinued.

NEVER plug untested circuitry into a computer backplane slot!!! Never never never, if you love your PC. All it takes is a simple little wiring error and your motherboard and disk controllers will be toast. While IBM-PC/AT buss interfacing isn't rocket science, it's all too easy to do expensive damage to your machine. Consider other interfacing methods (parallel/serial ports, commercial control/data acquisition boards) first. There are buffer cards that transparently permit prototypes cards to be used while protecting the computer buss, but they are not cheap.

Use IC sockets on prototypes. After you have things debugged, you might consider soldering chips directly to boards to save cost and eliminate connections that can oxidize or come loose, but during the design/testing phase you need to be able to swap chips without repeated desoldering.

More to come . . .

White Protoboards

Nobody does electronics work for long without running into the classic white plastic prototyping breadboard, made famous by Global. These are the de facto standard for whipping up a circuit just to see if it works. They do a good job of this and I don't know what I'd do without them, but remember that they have limitations.

Breadboard connections are not as good as wire-wrapped or soldered connections. If you are doing designs with high frequencies or high currents (anything above 100 mA) these connections may not be reliable.

Breadboards tend to wear out with use, especially if you've tried to jam some fat resistor wires into the holes. This makes the connections even less reliable or may make some holes unusable. If you are having problems with your circuit, use an ohmmeter to make sure there really is continuity where you think there is.

Breadboards don't always let you supply power reliably with the power buss along their edges. Make sure you use despiking capacitors on all your chips on the breadboard.

Wire Wrapping

Wire wrapping is a common prototyping method for good reasons: it is easy to do, and can be very reliable when done correctly. Most importantly, it permits very easy modifications and corrections. If done right, a good wire-wrapped prototype can be used as your final product instead of just a test of your design, making it faster than wiring something on a white breadboard and then transferring it to a printed circuit board.

The disadvantage comes in needing to use wire-wrapping sockets for all parts, and in having a circuit board that is thicker due to the long pins. I will be writing more on this here before long but for the moment, take a look at what an engineer with Ampex has written on the subject by clicking here. In addition, another person has a note on how NOT to wire wrap correctly.

2008 Addendum These days neither the EE Shop nor students use wire-wrapping much, probably because it only works well with through-hole .1" devices, not so much fine-pitched surface mount devices.

Perf Boards

"Perf board" is phenolic or fiberglass circuit board with perforations every 1/10th inch, allowing common electronic devices to be mounted. The cheapest have no copper pads. Who hasn't made a quick and dirty circuit by sticking the parts on one of these and bending the leads underneath to tack together with solder? It might work, but it's messy and unreliable and not very solid. Fortunately, they make perf boards that have copper soldering pads around the holes so parts can be soldered in. Some boards just have a pad per hole; some have a variety of strip patterns suitable for DIP integrated circuits. Especially handy are the ones that have the same layout as the Global breadboards, as these permit you to easily transfer your breadboarded circuits. It's often possible to mix wirewrapping and soldered connections on such boards. These boards are a straightforward way to make projects and are what our Engineering Shop often uses for our one-of-a-kind small designs.

A disadvantage to these is that, like any board with solder pads, if you solder and unsolder a pad more than a couple of times, it's likely to come off of the board.

Any catalogue that carries electronic components is likely to carry these, but we've been pleasantly surprised to see that Radio Shack generally has quite a broad assortment of solder-pad perf board. More here soon.

Printed Circuit Boards

A key point for a student designer to remember is that printed circuit boards are often NOT the best choice when creating a new design. You really should not be thinking about drawing a circuit board until you have a circuit that you know works because you've breadboarded and tested it already. It's hard to modify an etched board; and I can almost guarantee you that an untried design will require changes. The time and effort needed to create the board may also be a drain on your limited time resources if you are doing a class project.

2008 Addendum: The advance of technology has changed this advice. Many integrated circuits can not be obtained in through-hole, .1" center packages any longer but are only made in fine-pitch or BGA surface mount packages. You CAN'T stick these into a white breadboard even if you tried. It is possible to buy or have made an adapter that is basically a tiny circuit board just big enough for such a chip, that brings the wires out to dual-inline .1" pins for traditional breadboarding. Obviously this is more hassle and expense, but what are you gonna do?

By all means take a look at The Green CirKit PCB prototyping on-line manual, an excellent on-line manual with downloadable versions available. (Of course you can always dump the job of laying out your board on a third party like pclayout.com ).

Processes

2008 Addendum The UN-L Engineering Electronics Shop no longer uses any photochemical exposure and etching processes to make circuit boards. All PCB manufacturing is done with an LPKF routing machine that drills and mills copper around traces on plain copper-clad boards, using files generated by electronic CAD programs. Talk to Tom Grady in the EE Shop for more details. However, I don't like to see "legacy" design information vanish from the Web, so I've left the following information about older processes in place - just understand that we no longer use the chemical stuff. It was nasty and scary stuff anyway.

Most printed circuit board manfacturing is done by putting the pattern of the circuit traces and pads onto a sheet of copper-clad phenolic or fiberglass board using substances that will resist the action of copper-etching solutions. This can be done by directly drawing on a board with a resist pen or paint (here's one method of resist-pen plotting) ; by silkscreening the design on to the copper; or by a photographic process where the portions of a sensitized board that are struck by ultraviolet light passing through a negative image of the traces become, following development, resistant to etchant. A new and relatively easy method is the use of special sheets of material on which a circuit design can be printed with a laser printer or photocopier, after which the pattern is transferred to a copper clad board with an ordinary steam iron. This process is reviewed here. Also, a new process using toner transfer is advertised here.

In the UN-L E.E. Shop we have most frequently used the negative-photoresist etching process to make PCBs. In this system, the parts of the board that are exposed to UV light become resistant to etching, so any artwork of the circuit board must be turned into a high-contrast photographic contact negative that is transparent where one wishes there to be copper and is black where it is to be etched away. There ARE boards and chemicals that permit a positive-photoresist process; wherever the transparency is black, is where there will ultimately be copper. This can be more convenient from the standpoint of creating the photographic mask from artwork, in that one can try things like laser printing or photocopying on overhead transparencies sheets designed for this purpose. (I should comment that in our experiments it can be difficult to make the traces sufficiently black using this method. HOWEVER, , there's a way to fix that problem: read this method .) In addition, one can purchase rub-on tapes of circuit traces and pads that are laid down on translucent material to make a positive image.

However, we have found the positive photoresist process to be much less forgiving of incorrect exposure times, development times, and solution temperatures, often resulting in several tries being required to create a satisfactory board. As a result we stay with negative photoresist materials, which work reliably almost every time and are far less critical in use.

!!PLEASE NOTE: Developing and etching PCBs is a process that involves chemicals that can burn your skin, have hazardous neurotoxic fumes, can be absorbed into your bloodstream, and CANNOT be disposed of by dumping into ordinary sanitary drains. It should NEVER be done without powered ventilation or in an area accessible by children or pets. Of lesser but significant note is that it is common for these materials to wreak havoc on flooring, furniture, and especially on clothing. (Splash some etchant on your clothes and you might not notice until after you wash them that you suddenly have what looks like a shotgun blast through your jeans.) In short: don't fool with it unless you have proper facilities to do it right, and follow ALL safety cautions mentioned on the materials. You may well be better off to farm out the process to a commercial board maker; some are listed in this list of PCB Fabricators. New . . . you might also want to check with PCB-Quote Free multiple quotes for PCB Layout, Fabrication, Assembly. The E.E. Shop can also help you fabricate a printed circuit board: click here for details on what we need.

Creating PCB Artwork

One can also draw boards with any computer drawing program that permits you to space pads and traces exactly on .1" centers. Often there is a "Grid" variable that can provide guidance, and many programs allow you to "Snap" to that grid: i.e., all lines and items will end up being exactly on the grid dots simply by getting anywhere close to one. (One widely used program that can do this and more is AutoDesk's AutoCAD , and there is an add-in package to do PCB work called AutoCAD-PCB .) For these drawings to be any good, you have to be able to print out these drawings exactly to scale, on a device that can give you very sharp, black images. Laser printers can, and plotters will also if they have a good pen. (We have refillable plotter pens with Rapidograph-type tips that we use with India ink for smooth, black results.) You may need better-than-average paper to optimize the sharpness and contrast.

2008 Addendum : OK, these days you are more likely to not even need a hard copy - you probably are just sending a computer file to a place that fabricates the boards directly. Also, this page used to warn folks not to put too much faith in using electronic CAD, on the grounds that good packages are very expensive; that having to spend time learning the software sucks up precious project time; and that autorouting is not magic and can make a lot of whack decisions. These objections are much less true now: there are more packages available at lower cost and better quality. See our Software page for many useful links to this stuff. Also, given the prevalence of surface mount components, you really may not have the option NOT to use CAD to design boards.

There are links to a bunch of good PCB guides, resources, and software on Lazar's web page here.

The documentation for AutoCAD-PCB contains many good tips about using any software program to produce PCB artwork: an edited version of this file that contains just the "generic" information can be read by clicking here.

Links to a zillion PC board software and service bureau vendors can be found through Printed Circuit Design Magazine's buyer's guide and directories.

Other methods . . . there are also newer methods of creating PCBs: for example see how it's done at PCB Milling .

Surface Mounting

The shrinking of electronic parts into surface-mount devices has not only permitted drastic miniaturization of equipment but has improved the ability to automate the production of circuit boards. But what is easy for an assembly machine can be hell for a hobbyist or student trying to put together a project with nearly-invisible parts and leads that are incredibly close together. Some components are only available in surface mount packages. You can find some adapters that let you mount large surface mount ICs onto a small board with pins that can go onto an "old" through-hole PCB. Or you can learn to solder surface mount devices. To do that, you should read Hints and Tips for prototyping with Surface Mount Devices, an excellent tutorial that includes further links and resources.

Data Books

2008 Addendum Note that paper/printed databooks are a dying breed. Our current Electrical Engineering students assume that any device they want to use will have a manufacturer's datasheet available on the Internet. Fortunately this is largely correct. Some of the comments below apply to the data sheets whether you have a book of them in your hand or you are looking at a PDF you downloaded.

Data sheets from semiconductor manufacturers are your most valuable resource in creating designs, but you have to bear some things in mind.

First of all, they usually have all the crucial information needed to apply a device but will have it in very brief, no-frills format. Many books are not going to discuss the application at length and may give few, if any, examples. An absolutely critical detail about wiring or programming a chip may be mentioned only once in the text or on a chart: you need to read everything at least once and if you are having troubles you need to scour the information again.

Fortunately, some books are better than that. The data books from National Semiconductor have a justified reputation as a priceless reference set because they include lengthy discussions of their devices and are generous with providing examples of typical applications. These books can be purchased from many electronics suppliers and are well worth buying.

Another item to remember about data books: DON'T TRUST THE INFORMATION IF THE DATA SHEET IS LABELED "Preliminary Information". This means that the data sheet was written before the chip was actually in production, and the device may change significantly by the time it is actually released. The device performance may be different; still more important, the pinout may be completely rearranged in the final design. Preliminary data can help you choose a device but don't set your design in stone based on it.
Just as important, you may find datasheets for the perfect device only to learn that it has been DISCONTINUED. This can be a really severe problem, particularly with some special function devices. If you are just making a one-off project for your own amusement and you can get your hands on a discontinued chip to make it happen, fine ... but for a new design you hope to continue to use in the future, check with online catalogs to make sure your selected devices are widely available.

UN-L Engineering Electronics Shop