The Celtic Engineer is a weekly newsletter produced by Celtic Engineering Solutions. We hope you enjoy it. If you have any suggestions for topics, would like to give feedback or want your email added to the distribution list please send an email to [email protected].
Russia has been in the news lately, a bit, well some. So, I thought it was fair that we start off this newsletter with a Russian proverb – “Doveryai, no proveryai.” This means, “Trust, but verify.” If you think about it, that is why we put test points on our printed circuit boards. You and I know, that the circuit will work, the first time, every time. Perhaps we should add a few more test points.
Where in the world is my ground?
One of the most infuriating things as a designer of prototypes is after building a board up, after spending days or weeks designing it, sitting down and running a smoke test and then… Where did I put the ground? I am sure you have never done that, you always remembered to put a test point connected to ground, several actually across the board so it is easy to do the next step which is always checking the major voltages. Then again when you want to check signal integrity, look at a wave form, check timing, determine the S/N. Yeah, we’ve probably all forgotten it on one board or another.
On first prototypes, and on circuits where there are lots of unknows, I like to sprinkle them all over. It makes debugging so much easier. As a design matures, you often don’t need as many, although you should always leave some connected to ground.
Sprinkle is a relative term
When I say sprinkle, it might indicate that I randomly put test point throughout the project. I have often shown my work to none professionals, who give me the eyebrow raise, which means they think its really complicated but don’t want to say so out loud. There are plenty of complicated circuits out there, most circuits are not as complicated as they look, or at least their complexity has some bounds.
There is a certain flow to a circuit, an in, and an out. Those are good places to draw boundaries. I don’t know
if this is a good place to put a math analogy, but here it goes. Remember in undergrad physics when you had to
do an integration problem? There were two types of students. There were the ones who selected an appropriate
coordinate system, did the integration in two steps and handed in the test. The other group thought everything
looked cartesian. They handed in a 20-page test and hoped for partial credit. My point here is that if you draw
a box around an MCU it looks chaotic. How could anyone understand that? Draw the boxes around the
temperature unit, the pressure unit, the comm port and things start to look a lot more reasonable. Even signal
conditioning doesn’t look that bad if you think of it as gain, compensation, anti-aliasing filtering and an analog
to digital converter. It is at the boundaries that we often want to check a signal.
Why can’t I just poke at it?
I have worked with people who would rather sharpen the probe then include a test point. And you can do that,
but it can damage the board or a pin on a chip. That is if you are lucky enough to have the signal present on the
top or bottom layer. These people will stick to their option until that board where they have to look at one
signal on the top layer and another on the bottom at the same time.
Isn’t every test point just an antenna?
No, but they could be. How you place a test point can affect your circuit. The large test points shown in the
picture are usually placed on a test station and may be located many feet from the point they are measuring. To
be sure this would be an antenna and a noise magnet if that is how you were using it.
One way to deal with this is to place a test-point next to a trace and a short connection between the test-point
and the trace. This can still cause a problem for some circuits, but is acceptable most of the time.
Another technique is to buffer the signal you want to measure and send that copy, or amplified copy, to the test
point to be measured. This has the added advantage of not loading your original circuit with capacitance when
you place the probe.
Do you have a favorite?
I have tired many types of test points and I have a few favorites. The small trapezoidal one on the front right of
the photo is a great TP to use when space is at a premium. They can be a bit hard to solder, but after you have
placed a few thousand, it becomes old hat. Oh, and they cost $0.25 each, ouch!
When you have the ability to be a bit more expansive, the TP on the front left is a great choice. This is a loop of
wire that has a colored plastic bead to support it. Both of them work well with a queue-ball test lead. The later
comes in a rainbow of colors and a variety of sizes. The one shown is one of the smallest.
One technique that should not be overlooked is a bare pad. Sometimes it is useful to have a pad with no soldermask
that you can drop down a probe for a quick look. They don’t cost anything to have and can make your
life, or that of your technician, much easier.
This newsletter is sponsored by Celtic Engineering Solutions LLC, a design engineering firm based out of West
Jordan, Utah, which can be found on the web at: www.celticengineeringsolutions.com. You can find the
newsletter on the company blog, LinkedIn or in your inbox by subscribing. Send your emails to The Celtic
Engineer at: [email protected], with the subject line SUBSCRIBE.