![[The Simple N-JFET Board]](NJFETA1.JPG)
Small pieces of double-sided PC board, a sheet metal hole punch, and a soldering iron make easy electronic circuit breadboards.
"Breadboarding" is the process of building electronic circuits in a way which supports easy change.
Here, "change" means anything, such as:
The term "breadboard" comes from the early days of electronic experimentation. In those days, the food staple known as bread was not bought as a sliced loaf, but instead was hand-sliced at home. To prevent damage to countertops and keep the bread clean, almost every american house had a wooden board of appropriate size to slice a loaf of bread.
Because every house used breadboards, and got new ones from time to time, breadboards were available to early electronics experimenters either free or at modest cost. Essentially, breadboards provided an insulating surface on which parts could be mounted. Tube sockets and other components could be held on the board with wood screws, and wires connected between components. Often, the electronic components themselves had connection tabs and few intermediate connections were needed. But when other connecting points were useful, copper tacks or brass screws would provide easy insulated connection nodes.
While the old-style wooden breadboard is long gone from electronic
experimentation, the basic role of the breadboard remains.
Even in this era of SPICE electronic simulation software, it is
often necessary to verify the results in a physical circuit.
The electronics experimenter of any era needs a way to physically
mount parts, then connect those parts, so the operation of a circuit
can be verified, characterized and optimized.
Usually, the whole point of all this is to use the constructed
circuit.
Ideally, a development breadboard would itself become a final,
operating circuit to avoid the need to rebuild it in another form.
My breadboarding approach is related to what ham radio circles commonly call "Manhattan style" construction.
Connection Pads.
Manhattan-style construction typically uses a hand metal punch to
manufacture
The Manhattan-style approach typically uses Super Glue (or similar instant adhesive) to stick the pads in position, but I found the glue to be a problem. Sometimes the glue would not hold, and sometimes it was messy, and often the bottle did not keep long after being opened. So I tried soldering the pads in place, which worked out better for me. Soldering the pads probably does require some effort to develop the technique.
I find it convenient to first "tin" the pad board with solder on both sides. That produces pre-tinned pads when punched. Normally I brush liquid flux on the raw copper. Then I add some solder to the iron and move it across the surface until no more bare copper is left.
Layout. The usual Manhattan-style construction is more than just constructing a circuit on top of PC material with insulating pads. Normally, electrolytic capacitors stand straight up, leaded resistors stand on end, and there is three-dimensional quality to the circuit. Presumably the reason for that is to get more components in a smaller volume, but I am dubious about how much is really gained that way. I prefer a two-dimensional circuit, with components fixed in place and as close as possible to the ground plane, unless there is a reason why not. That allows the breadboard to be a first approximation to a production board layout.
Before mounting pads, it is best to lay out the parts and see where things will go. It is often helpful to arrange the parts similar to the way the schematic diagram is drawn. Indeed, playing with the parts on the surface may lead to the insight that the schematic should be re-drawn in a more reasonable way. An appropriate goal is to construct an arrangement of pads in which the surface-mount resistors connect from pad to pad with no wires needed. Connections from pad to ground are also easy. Some wires will be needed, but they can be relatively few.
The Board. Given a first approximation at a layout, one can develop an idea of the size of board needed. If we have a larger piece we cut it down. We certainly want to use fiberglass PC board and in most cases the double-sided variety. The thinner material, typically used as inside layers of multi-layer boards, is easier to cut. The "1 oz." copper layers are significantly easier to solder than the heavier "2 oz." layers, but either can be used.
I mark the board outline in pencil, using a combination square. Then I cut out the board using sheet metal shears. Since the result is hand cut, I square up the sides with sandpaper on a flat surface. It is important for personal safety to round off all edges and especially the corners of each board. Copper may be a "soft" metal, but it will cut human skin.
It is usually helpful to clean the copper surfaces. One can use a very fine scrubbing powder, or a chemical copper cleaner. I have been using "Tarn-X," but it is not magic. Then wash the board with soap, rinse well, and dry.
Lay Out a Section. This breadboarding technology seems to work best when the circuit is constructed in small sections, instead of placing everything then wiring everything at once. Ideally we would also test each section as completed. Usually I work on an area about 1 inch square or less, and one IC chip at a time. Typically, multiple pads are needed to support each IC chip.
First I arrange the pads and chip on the bare board area to get what I want. Then I mark pad positions with solder bumps or mounds. To do that, I flick a pad away with the iron, then place a small mound of solder at the original position. This marks where the pad should go, tins the board there, and provides extra solder for the upcoming opearation.
For an IC, I may place very small solder mounds around it on four sides. Besides marking future position, these will later conduct heat to glue the chip.
Mounting Pads. When soldering the pads in place, I try to put down several at a time, because the first one heats up the local area and makes soldering easier for the rest. With pad positions already marked in solder (and, thus, pre-tinned), I place a pad on the mound, hold it down with a toothpick, and then bring enough hot solder on the iron tip to tack the pad in place.
After a few pads are tacked in place, I again hold a pad down with the toothpick, and bring enough solder on the iron to solder most of the circumference of the pad. The point is not to solder the edges, but the edges are the only access we have to heat the board under the pad. Since both the board and pad are already tinned, the soldering problem is mostly reduced to providing enough heat. If we get the area hot enough, the pad ought to suck solder into the volume between pad and board. Then it may mush the solder out the edges as the pad is pushed down onto the board. As a result, that pad will not be going anywhere.
It is important to carry more solder on the iron than one might normally use for connections, because the solder mass is bringing heat to a copper surface which will distribute the heat quickly. Typically I have to move the iron around the circumference of the first pad several times until the solder under that pad melts. Subsequent pads in the group generally are easier.
Mounting IC's. To mount IC's I use hot glue, but not in the normal way. Typically, I use diagonal cutters to take a tiny snip from a hot glue stick, and place it on the copper surface. Using the soldering iron at appropriate places around the glue piece, I warm the copper until the glue melts. Then I stick the IC upside down in the hot glue. It is important to use a minimum of glue so the bead that forms around the edges does not get into the IC pins and make them difficult to solder. Modern SOIC packages are tiny, and need only a tiny piece of hot glue. When necessary the glue can be reheated and the part easily repositioned or replaced. Potentiometers and other controls can be stuck in place with larger amounts of hot glue. Terminal strips and other parts can be soldered to pads.
Power Connections. For audio circuits, I typically make the top copper surface "ground." Since connections are easily made to the copper surface, having an appropriate ground can greatly reduce the number of interconnecting wires. I leave the bottom copper surface unused, which allows the completed boards to be placed on a larger conductive surface without accidental shorts (but watch out for the edges).
For fast digital circuits, I again make the top copper surface "ground," but also use the bottom copper surface for Vcc distribution. In this way both Vcc and ground are distributed by full "power plane" surfaces with very high frequency and current capabilities. Tantalum capacitors around the edge easily give almost leadless broadband power bypassing.
For individual digital chips, ground pins are easily connected to the top copper layer with bare wires. A hole drilled by the power pins provides access to the lower layer and Vcc using a tiny bit of insulated wire.
Mounting Resistors. I recommend using tiny surface-mount resistors. I was lucky to find a good assortment of 1 percent 1206 resistors at Jameco. The "12" and "06" refer to length and width in hundredths of an inch, so these are just over a tenth of an inch long. That is fairly large as surface-mount parts go, but about as small as I want to place by hand. Using these parts means developing some amount of expertise in dealing with tiny, hard, slippery objects. I use a magnifying headband, tweezers and a toothpick. The toothpick is helpful because it does not conduct-away needed heat, solder does not stick to it, and it has a little "give."
The surface-mount resistors are solid components without leads. Resistors typically mount from pad to pad, and so the tweezers must set them on a slippery metal surface. Using a toothpick to hold them down, they tend to twist and turn, and in the beginning some would do a tiddly-wink thing and disappear. But using the tweezers during soldering tends to be worse.
Surface-mount parts solder in a given position and must be re-heated to change that position. If the part is soldered at both ends, then both ends must be heated. We do not want to heat one end and push on the part to bend the other end into position, since that may rip the silver contact film off the ceramic base. Nor do we want to take forever at heating the ends, since the silver film may go into tin-lead solution, thus opening the resistor. But I have had no problems re-heating parts quite a few times.
Typically I tack one end of a resistor to a pad by holding the resistor with a toothpick, and bringing solder to the pad on the iron. Then I make a good connection to the opposite side, and re-do the tacked side.
Replacing Resistors. First heat the pad on one side of the part, then the other, repeatedly, until the part can be flicked away by the toothpick. Often, the used resistor is not completely freed but instead sticks to the edge of a pad which then must be re-heated to get the resistor all the way off. Often, the removed resistor irritatingly sticks to my soldering iron tip. Usually I do not attempt to save removed surface-mount resistors, although I may re-use them in the same project.
It is important to remove excess solder from the used pads
with solder-braid, so the replacement resistor will mount well.
Often it is better to also remove the wires to those pads
(bend them up slightly) and then re-connect them after the part
has been replaced.
Here is an early example from my audio preamp development stream.
This particular circuit is a
These transistors connect to terminal strips mounted on pads. The terminal strips allow the effects of different devices to be checked. The large electrolytic capacitor is mounted by its leads, using heavier wire from ground and a pad. External connections use pads at the edges of the board. The diode string at the bottom is unused.
Power comes in at the upper left. Signal comes in at the left, and amplified signal goes out at the lower right.
The board is designed for battery operation, and most batteries can source dramatically large currents into a dead short. One unusual problem with this breadboard technology is some tendency to get very fine solder hairs from pad top to ground. There is a battery-protect resistor (here 47 ohms) to limit the amount of current drawn (here 191mA @9V). When pad shorts happen, they are easy to find (just ohm each pad to ground) and repair (use solder braid to remove excess solder).
Many connections can be made with surface-mount components
soldered between pads or from pad to ground.
When wires are required, I use short bits (typically just an inch
or two long) of #30 AWG solid wire.
I like Teflon insulation, because I tend to damage or melt other
types.
Different colors help track which wire carries what.
Normal blue wire-wrap Tefzel can be used (with some care), and
also is a fine source of bare silver-coated wire for most chip
ground pin connections.
When working with a breadboard, we expect batteries to be
connected and changed and metered for current.
But if a battery is accidentally touched in reverse, even
transiently, that could be bad.
Easy, cheap, and effective reverse-voltage protection is provided
by a single
![[Power N-MOSFET Structure]](NMOSFE4A.GIF)
Normal Operation of Power N-MOSFET. This is the typical enhancement power N-MOSFET structure. The body is made of single-crystal silicon, with impurities diffused to create three distinct regions: N, P and N. The N regions have an excess of electrons and the P region has an excess of holes. Holes are spaces in the crystal lattice which could accommodate an electron, but are unfilled.
The colors generally represent voltage. Red is positive, orange less positive and yellow even less. Blue is negative, green less negative, to yellow again. In an N-MOSFET, the channel area under the gate does not exist until the gate is positive, and even then is very thin.
The source contact is metal on the surface.
Typically that also contacts the P region or substrate.
The gate typically is a polysilicon film on top of a silicon
dioxide glass film.
The drain contact is the bottom of the chip.
In normal operation, the source is typically ground and the gate
is positive with respect to the source.
But what really matters is the gate being positive with respect to
the P region.
That turns the
An N-channel enhancement MOSFET conducts when the gate is positive
with respect to the rest of the P region.
The positive electrostatic field attracts negative electrons in
the P region toward the gate, where they form a thin conductive
channel at the surface of the solid crystal in a region otherwise
lacking electrons.
![[P-MOSFET protection, good power]](PMOSFE4A.GIF)
Normal Operation of Reversed Power P-MOSFET.
Here is a
In these simple breadboards which take at most just a few tens
of milliamps, the voltage drop across the
![[P-MOSFET protection, bad power]](PMOSFE5A.GIF)
Protective Operation of Reversed Power P-MOSFET. Here the battery has been reversed. Now the body diode is reverse-biased and does not conduct, so everything depends upon the gate voltage.
The gate needs to be negative with respect to the N layer or
substrate to form a conductive channel.
Since the gate is not negative, there is no channel.
With the body diode reverse biased, and without a channel, there
is no conduction and the LED is off.
That is how the
Sometimes a breadboard is constructed to develop an intermediate
voltage on the copper surface we would typically call "ground."
In that case, the
Ideally, one might think to protect the MOSFET gate insulation with a zener diode and resistor. In actual practice the simple circuit of just a raw transistor has been very reliable, and I have not had a MOSFET damaged in this application. Without using MOSFET reverse voltage protection I might well have damaged various breadboard circuits by accident.
The simple trick of using a
Another thing about working with a breadboard is that any new construction can have misconnections and short circuits. Having clear, observable wiring can help reduce errors. But a short across the power lines will take as much current as the source can supply. And when the power source is a battery, dramatically large current can flow through a dead short.
One result of unrestrained current is heat.
The other result can be destruction of the battery, which would
otherwise be the future power source for the board.
To protect the batttery, there should be a series resistor, here
typically 47 ohms.
With a 9V battery, that limits the current to about 190mA, which
also would imply about 1.7 watts dissipated in the resistor.
I use a 1/8th watt part.
So, if there is a board short, the resistor probably will die
unless the battery is disconnected quickly.
If the resistor does die, that tells me to fix the short and
replace the resistor.
Always measure current when first powering a new board. That is easy with a 9V battery, by connecting one side to the snaps, and using meter leads to connect the other battery terminal to the other snap. Start out with a brief touch, since the board may have a power to ground short and take an amp or more. If so, inspect the power area for an obvious short. If no problem is visible, disconnect the other sections from power (typically a couple of wires) and then check again. Once the problem is localized, try cleaning up solder on and around power pads using solder braid.
Then make the board do what it is supposed to do. For example, with a preamp, first check the output bias voltage. If there is an on-board bias generator, check that and other bias points on the board. See that the transistors have reasonable bias voltages. Then send in a low-level signal and check for output. An oscilloscope can be very handy at tracing the signal and seeing where and how it goes wrong. Typically: