RF Prototyping on a Shoe Box

[snip]

2 seconds on Cox ;-)

...Jim Thompson

--
|  James E.Thompson, P.E.                           |    mens     |
|  Analog Innovations, Inc.                         |     et      |
|  Analog/Mixed-Signal ASIC\'s and Discrete Systems  |    manus    |
|  Phoenix, Arizona            Voice:(480)460-2350  |             |
|  E-mail Address at Website     Fax:(480)460-2142  |  Brass Rat  |
|       http://www.analog-innovations.com           |    1962     |
             
I love to cook with wine.      Sometimes I even put it in the food.

Hello Kevin,

Then something gets hot, the shoe box lid catches fire, the carpet lights up, sirens start blaring ...

Regards, Joerg

Hi all, I saw this good site that shows how to prototype RF gear on a show box using copper tape. Better than spending a day designing a pcb board only to have to make adjustments when its too late.

Cheers, Kevin.

Hello Kevin,

Dry timber is pretty flammable. I wouldn't use plastics either, you never know how toxic its fumes can be.

Regards, Joerg

Hello SioL,

Yes, and I still remember all the people who died at Duesseldorf airport in Germany. Not from the fire itself but from toxic fumes.

Regards, Joerg

I like Jim William's AN47-1 App Note at Linear Technology on building high speed circuits.

Go to

and use the upper right "Search" box for AN47-1. It's chock full of tips and gotchas for building high speed circuitry. But the note's pretty big. I wouldn't try downloading it on dial-up.

He does it on copper clad pc board with what he calls "ugly breadboard" techniques to minimize parasitics. IIRC, he's got 100MHz Amplifier circuits in the note with very clean results. Though his pictures and data tend to be in the Time Domain. He's also got a nice pulse generator circuit in an Appendix for generating about a 10V 200ps rise and fall time pulse to test these circuits. Another appendix has the "ABC's of Probes" guest written by the Tektronix folks.

One caveat, the note's pretty old and doesn't talk much about SMD components.

Robert

how toxic its fumes can be.

Plexi burns like hell once it really gets going, nasty fumes as well.

--
Siol
------------------------------------------------
Rather than a heartless beep
Or a rude error message,
See these simple words: "File not found."

Wouldn't know. I got it in a paper copy a long time ago.

What do you think of the App Note? Have you seen it?

Robert

You'll forget shoe box construction once you try "Manhattan Construction". Go to:

and scroll down to see what it looks like. Refer to: for details on how to do it. It is easier than foil on shoebox, and far better. You have a ground plane under

100% of the circuit.

Ed

Agreed, timber and plastics are nasty substances in a burn-up, so I'd use an old dinner plate, the local goodwill shop sells them for $2 a bagful. Ceramic, copper, solder, all the makings of a quality low-loss construction, much like Tek used to use in their real quality scopes ?

The only down-side I can think of would be dropping it!

Barry Lennox

It does not provide a ground plane though, so radiation must be pretty high. It is not something I would do.

I prefer what I call 'dead bug sytle' where I cut bits of single sided PCB up, and stick those on top of another bit of singled side PBB. You can make 50 Ohm tracks, by making the widths right.

RF is not quite well as contrained as on a normal board, since there is no dieletric outside the track.

To be more precise, etch simple stright lines on PCB, and make the dielectric 3 x the track width or something. Just stick to lumps of that and pads. I've used this at several hundred MHz.

Joerg wrote in sci.electronics.design:

That was PVC (Polyvinyl chloride) with its nasty HCl fumes. Plexi is acrylic something. The fumes are probably no fun either, but they don't have the poison-gas quality of smoldering PVC.

Anno

that the pads weren't round but usually squares.

only downside is that the peel-and-stick technique

reaches 100F stuff can start to slide before you

own squares.

Hi, Joerg.

At one startup I worked for, we had a little hand tool in the lab to cut small isolated pads in a copper clad sheet. Most of the copper was ground plane, of course. It worked great when common transistor packages were larger. Anymore, I think inexpensive etched prototypes are the way to go. It's very difficult to build stuff by hand as small as it can be when layed out on a regular PCB.

--
--Larry Brasfield
email: donotspam_larry_brasfield@hotmail.com
Above views may belong only to me.

That looks pretty doable. The only think I dislike about homemade cutting tools is that they dull fast. (Maybe a foray into the art of tempering would fix that.)

For some kinds of RF, a circuit you can get your fingers or hands into can still work. For other kinds, the added stray inductance and capacitance that follows from size will make the prototype a poor predictor of performance obtainable once a PCB is carefully layed out and built.

I've always liked the ground plane on the component side where short connections can be made to it without having to drill holes. That's what I like about the method using isolated pads made where needed.

More true than ever as densities have gone up.

--
--Larry Brasfield
email: donotspam_larry_brasfield@hotmail.com
Above views may belong only to me.

Hello Jim,

Wow. It was about 15sec on SBC DSL. It's a 5MB document, so with all overhead that would mean your BW is around 25Mbps?

I really like that classic Tek scope in Fig 31.

Regards, Joerg

Hello Ed,

That's the way to go but it is an old trick. We did that in the 80's, except that the pads weren't round but usually squares. Easier to cut.

Wainwright makes little strips that can be glued on, including SMT pads. The only downside is that the peel-and-stick technique isn't too reliable on really hot summer days when prototyping outside. When it reaches 100F stuff can start to slide before you got some ground wires soldered to it. So I either add some real glue or use my own squares.

Regards, Joerg

I have made several such tools.

Take an old screwdriver, use a file to change the blade into two sharp points. Like a straight fork. Make one of the points a little softer and 1 mm shorter than the other. Cut away material on one side of the rounded point so it becomes a cutting edge.

This means that we need to think through the function. Put the halfmade tool vertically on a piece of pcb board. Turn the tool, see how the longer point digs into the copper and into the pcb material under it. So its position stays fixed. The soft point will just glide around in a circle around the sharp point.

We need to remove material from the soft point to create a flat surface perpendicular to the movement, on one side of the soft point, the side which is leading when the tool is turned clockwise.

When that side presents a flat vertical area it will cut away copper easily. It can be made slightly spoon-shaped with a rotating minidrill tool. Keep the edges of this flat or spoonshaped area sharp.

I use two different sizes, one for 3mm pads, one for 6mm.

Why worry about space?

The board will be inside a box if it is in use, so the pcb design or size are of very little significance.

In fact, it is a great advantage to have a fairly open circuit where there is space enough for changes and repairs.

Use simple pcb laminate, single or doublesided, which is very cheap when you find it in a surplus shop. A lifetime supply can cost 20 dollars or so.

Then just cut suitable pieces, clean it with a kitchen sponge, groove it with a sharp tool, to get the number of pads you think you need.

Start building, and you can add the methods from manhattan and dead bug style, and make new pads with the rotating tool we made above.

If you need a groundplane, or if you want to build on both sides, use doublesided pcb laminate.

Basically you need no holes, which saves a lot of work. Old components can easily be soldered to a flat surface with some ingenuity, cut wires, bend wires, and solder. New smd components are placed over a groove and soldered at both sides.

Integrated smd circuits are a problem, because of the many pins tightly packed, but that is a problem with most simple building methods, and can only be fully solved by designing and etching your own circuit boards.

--
 Roger J.

Back in school, we used single-copper-sided duroid, and used copper tape on the other side. A common tape width was 50 ohms, and we usually drilled a hole to mount the transistors. That way, we could solder the ground connection really short, and just bend the other leads up to solder to the tape.

For bigger, more complicated projects, we had a wide copper tape that we would cover the entire board with, and scribe it with a dull exacto knife in a precision pantograph? table with precision dials so you could get 1 mil resolution if you were careful...

Charlie