A) Using DVM's and Scopes

Do you think you already know how to use a DVM or multimeter correctly?  Are you sure?



1) Using DVM's or multimeters.

The following are fundamentals to know about DVM's, if you are an intern.  There is a level of familiarity that is sort of ok if you are doing a project around the house.  If you are working in engineering, the stakes are higher.


- Connect the probe leads correctly for the measurement at hand.  

Typically, voltage and current measurements use different probe sockets.  A very common mistake is to measure 'amps' by moving the dial on the DVM, but not move the probes to the correct configuration.  Making that mistake will often damage the meter, and sometimes smoke will come out.


- Beginners very commonly confuse "OL" on a DVM's screen, for "zero ohms."

OL refers to 'overload' or such a high ohms value, that the meter cannot resolve the value.  Typically in the megohms range.  For beginners however, perhaps the "O" looks like a zero.  If you are a beginner, avoid this mistake.  If you are working with a beginner, make sure they know the difference.


- Do not use the 'beep' sound to confirm a connection.  

Yes, it is awfully convenient to hear 'beep' to confirm a connection.  Do not do this.  Measure the resistance in ohms.  Always.  This is engineering, not a home project.  "5.2 ohms" is worth a lot more than "I heard it beep."  For best results, subtract the probe resistance before reporting low ohms values.


- Do not take a meter's reported result at face value.

DVM's are not magic, and require a small induced current to operate.  If a wire vanishes into a wall and comes back out, and you read 35 ohms across it, is that accurate?  Not necessarily.  
If in doubt, measure the resistance both ways by swapping the probe tips and measuring again.  If you see a significant difference, there is something else interfering with your measurement.  The conductor may be influenced by something else you can't see.


- DC voltage measurements presume that you already know the signal is a DC voltage.  

This is a very common mistake.  Often, good meters have some way of indicating to you that the signal is varying.  But not always.  When in any doubt, check the signal with a scope.


- All measurements have value, type (dc volts, ac amps, &c), and 'sig figs.' 

NASA lost a $327 million Mars orbiter due to a units mismatch error.  Nobody 'feels anything' when measurements are glossed over, if the number sounds sort of right.  The mistake is painless at the moment it's made.  Two choices: either a habit of clearly stating measurements, or a habit of fixing consequences.


- There are plenty of other mistakes to make.  

This is not a comprehensive list of mistakes.  Rather, just enough to get the point across.  For instance, it is best to know when measuring AC voltages, what "True RMS" and "Total Harmonic Distortion" means.  Currents can be read very wrongly if on the wrong ampere scale, and so forth.  Check your results against expectations.  When in doubt, break out a scope.

In sum: 

- Your meter measurements are only as good as you are.  




2) Using oscilloscopes.

Using a scope can often go spectacularly wrong for people new to the field.

Let's get the lesser issue out of the way first.  


- Don't rely on the 'auto' button.

Need to use the auto button, to set a scope up?  Instead, get familiar enough that you can trigger on any signal by setting it up yourself.   If you can't set up a scope for yourself every time, and have to rely on 'auto' to find your signal, you aren't good enough for real lab work yet.  Engineers can set up scopes for themselves.   

If you aren't at that point yet, that is ok.  Get familiar, get help if you need it.



- Now the big one, and this is safety related.  This is long, but important.

Some scopes are 'isolated' and some scopes aren't.  Knowing the difference will allow you to take correct measurements.  And more importantly, not accidentally start a fire.  


Non Isolated Scopes:

Every standard scope probe has a tip, and a reference clip that the measurement is referenced against.  Non isolated scopes are designed to have their probe reference clips earthed through the power cable.

If your scope has a three prong power plug, and the plug's Earth pin is electrically connected (a few ohms tops) directly to the probe reference clips, it is a non isolated scope.  Typically, those clips are all electrically connected to each other, as well.  

The classic mistake made with non isolated scopes, is to measure across the positive and negative power supply rails of an amplifier. 

Say the amplifier is running at +15 VDC and -15 VDC with respect to earth.  Clipping the probe tip onto +15 VDC harms nothing.  But connecting an earthed probe reference clip to -15 VDC, will short the power supply's output to earth.  Through the scope.  This can damage your circuit, damage equipment, hurt you, and/or start a fire.

Clipping a non isolated scope probe's reference clips to anything other than earth potential, is asking for trouble.

Some people get around this by 'floating' the scope.  Meaning that they defeat the power plug's ability to connect to earth, by using a 3 pin to 2 pin power adapter.  Without a connection to earth, you can't short to earth.  However, typically, all the probe reference clips are still shorted to each other.  Thus this really isn't a substitute for a fully isolated scope.


Isolated Scopes:

Each probe and reference clip is completely isolated on an isolated scope.  Allowing you to clip onto pretty much anything, within the voltage limits of the probe.


Not Quite Isolated Scopes:

'Floated' scopes were mentioned above.  How about USB scopes?  Where you plug a little box into a laptop's USB connection, and run a program to see the traces.

Typically, the probe clips are connected to the USB cable's backshell.  And then on to the laptop's chassis.  If the laptop's chassis is earthed through the laptop's power supply, you've got a non isolated scope.  If the laptop's power supply has only two pins and not three, you've got a 'floated' scope unless it is earthed by some other pathway. 

If you have a USB scope that is specifically designed to be an isolated in the first place, it will likely be marked as such, and be a lot more expensive than most USB scopes.


If you are an intern reading this page, and have truly understood everything above this line, you are a lot less likely to harm yourself, destroy equipment or inadvertently start a fire in the lab.  Your measurements might also be reliable too.  Hopefully this has been a positive thing, that will help you for a lifetime.

1) The Inverted Tree.

The reason this is not being called earth, which it typically represents, is because it can possibly mean other things.  If you doubt this, look up the IEEE spec for symbol usage.

When dealing with certification labs and compliance people, they will pretty much enforce the 'earth' interpretation of this symbol.  Unfortunately, it's very common for engineers to use this same symbol for:

- a 'zero volts' reference
- connected to the negative side of a power supply
- connected to the chassis, which may or may not be earthed, somewhere
- connected to earth, but not with a conductor capable of dumping heavy currents
- the counterpoise of an antenna, which may or may not be actually earthed
- ... and so on.

The possibilities for confusion are vast and endless.  

For you, the new engineer or intern, use the symbol to represent 'earthed' and only that.  Unless specifically instructed to do otherwise by a senior developer, your manager or for historical consistency with the document at hand.

Your schematic should typically have just one of these symbols.  Or none, if it's not earthed.  It should represent the point where a nearby grounding rod is attached, via a heavy current conducting wire.  Typically a green conductor with a yellow stripe (North America).  Or where a large copper lug is attached, to serve the same purpose of safety grounding.

In larger system schematics, multiple 'earth' symbols may not be electrically connected well at all.  Particularly if there are multiple ground rods involved, tens or hundreds of feet away from each other.

But, just how confusing could it be, to use this symbol as flexibly as the IEEE spec allows?

Here is a basic example.  In North America the white or 'neutral' power conductor is supposed to be earthed back at the source of power.  The green conductor (with yellow stripe) at your equipment is meant to be earthed more locally for immediate safety reasons.

One conductor is for safety, the other for power.  Even though the white neutral line is earthed elsewhere, and is called 'neutral,' it is still supplying half of the power to the load.  Look up and understand the 'Poynting Vector' concept, if you doubt this.  Power is delivered to a load through both supply wires, not just the 'hot' wire.  Unless something has gone very wrong, the green earthed conductor should not be supplying power.

Can you see the confusion that would result if both the green and white wires were labeled simply: 'earth?'  And yet, this is exactly the case for all other circuits where this symbol is used, not specifically referring to 'earthed with a heavy current carrying conductor, to very nearby ground rod.'  You have to guess at the intended meaning of the schematic's author.  Muddling the meaning of this symbol has created countless unwanted ground loops, incorrectly referenced sensors, injection of power noise into sensitive circuits, and on and on.  

When an outside consultant is brought in to look for an electronic issue with 'mysterious, intermittent occurrences' - signal referencing and 'ground loops' are often the first thing they will quietly check.  It's an issue that evades some of the very smartest people.  Because it's just so easy to assume that the innocuous earth symbol means what you think it means, at the time.

Best solution for all of the above: use the symbol only for its best main purpose.  'Earthed.'



2) The Rake.

This symbol refers to Chassis, typically the conductive chassis or frame of your enclosure, vehicle, and so forth.

Also, this is where pragmatism starts to take over, and muddle things up.  Chassis can be used as follows:

- a current path to Earth for safety reasons (chassis typically earthed by a copper lug, somewhere)
- a means to return current to an electrical power source, such as a 12VDC battery in a car
- a signal reference voltage, such that 'single wire' sensors bolted to the chassis can work.
- as a counterpoise for an antenna, or multiple antennas
- all of those and more

If the chassis is earthed, there should be some indication of where the chassis (symbol) is connected to the earth (symbol).  However it's common for the chassis to be poorly documented, electrically speaking.  Many companies have detailed schematics of their circuit boards, but only an assembly drawing indicating where the grounding lug is attached to the chassis.

As such, 'shortcuts' abound.  Schematics may show literally everything earthed with no chassis symbol at all, earth used in place of it.  Or worse yet, the chassis connection is shown but there is no evidence whatsoever if the chassis is earthed or not.  Not even a note on the schematic.  Sometimes, a chassis is inadvertently earthed by assemblers using metal screws, and then suddenly that connection goes away one day, when plastic screws are used instead.

For you, the intern: if at all possible, use Chassis only as a means to connect to Earth for safety reasons.  

If you do have a clear case where it's harmless enough to use Chassis as a reference potential, or any other purpose, know that you are opening your design up to any noise source that someone might attach to the chassis in the future.  Document what you are doing, and why.  Put a note on the schematic.

Also:

It is somewhat of an art to route power return currents through a chassis without causing other problems.  This is hard to capture on a schematic alone.  

If you are trying to get a stable zero volt reference from a chassis that also returns 5 DC amps to your power supply, there can be big problems depending how it is done.  Look up a technique known as 'star grounding' or better yet, use dedicated wire harness conductors instead of the chassis.

Chassis connections are also notorious for spreading circuit noise around.  If you connect a noisy circuit to the chassis, just about every other circuit can pick it up.  If two circuits source or sink current along the same section of chassis, there is going to be some level of crosstalk between the two.


3) The Triangle.

This is an abstract thing, utterly unlike the first two.  Except of course, when the earth symbol is used for this same purpose, generating a lot of confusion.

The triangle may be known as common, circuit common, zero volts, signal ground, or a number of other more obscure things.  For the sake of simplicity, let's just call it Common here.  

This is an abstraction.  It is where the author of the schematic declares: 'This is zero volts.  Measure voltages with respect to this potential.'

Very often, there is a subscript associated with this symbol.  A small 'A' or 'D' or other letter.  Typically, 'A' denotes Analog and 'D' denotes digital, but anything could be used.  

You may see several Commons in a schematic, all referencing different voltage potentials.  One might be directly referenced to Earth.  Others may be completely isolated, some may be offset by 1000 volts, and so forth.

What this symbol does, is allow you to more easily understand what is going on.  Analog circuits may use a Common symbol subscripted with an 'A' to help you see how an amplifier is referenced.  Which may be offset from your Digital and Power circuits.

On rare occasions, you might even see a Common symbol attached to a waveform that is oscillating with respect to Earth.  That is 100% valid, if it helps to clarify what a circuit riding on that waveform is doing.  Although in such case it would be really decent to put a note on the schematic describing what is going on.

If you have a circuit that is 'air gapped' or inductively coupled, you can add the subscript 'ISOL' for isolated to a Common symbol.  It is also good practise to put isolated circuits in their own section, and avoid 'mixing' Common symbols all over the page.  

With schematic capture software, there is also a very terrible thing that can happen.  Symbols may be visually indistinguishable, while the connectivity (such as Analog Common or Digital Common) is 'hidden in the electronic properties' and buried in each symbol.  In which case, you can't visually tell what is going on.  Even worse, the schematic appears to show that everything is connected, when it isn't.  Avoid this situation at all costs, but also be aware that it can occur.  If a circuit board doesn't seem to match its schematic, you may have run into this.

For an intern or new engineer, when referencing what *you* mean by 'zero volts' in your circuit, place a Common symbol and give it a relevant subscript.  

Just remember that this is purely an abstract symbol, and there is no such thing as a universal 'zero volts' in this universe.  Except as a reference that you are making for purposes of measurement.  This is the zero volts reference symbol that is the most appropriate to use in most cases, for analog or digital circuits.

In sum:

- Use the Earth symbol only for Earth.
- Use the Chassis symbol only for Chassis.
- Use the Common symbol as a reference to 'zero volts' as you yourself have defined it.
- If any of these symbols do connect, explicitly show that.  Leave no room for assumption.

The above will keep you out of a lot of trouble.  


* * * * *

RF Ground.

A brief note about another horribly confusing term, often without any symbol at all: "RF ground."

"RF Ground" is meant to refer to the reference potential (and geometric reference plane) of an antenna's RF function.  

- Antennas on the ground may be heavily Earthed near the antenna base, with Earthed radial conductors and even 'mesh mats' of conductive material.

- Antennas on the roof of a building have no such access to Earth.  There may be a wire mesh, or plate, that is referred to as "RF Ground" in such case.

- Antennas on vehicles may use the vehicle Chassis as "RF Ground."  (Chassis gets a lot of use)

But what is really going on here?  What does this have to do with the actual ground?

The real concept in play here is something called counterpoise.  We won't get into that here, other than to say this is related to antenna efficiency.  A horrible analogy might be: an antenna broadcasts better when it can 'push against' something, electromagnetically.  Go read up on it for the real story.

But whatever you do, don't muddle 'RF Ground' in with the concepts discussed above.  It is a completely separate sort of thing, for purposes of antenna efficiency and driving radiated fields.

* * * * *

A final note about the misuse of symbols, for interns.

Please do not go around correcting senior engineers with your newfound knowledge.  They get by just fine, and have learned to navigate all the confusing stuff over the years.  The points made here, are to allow *you* to be crystal clear in your communication with *them*.  And possibly help you get your products agency certified and approved without redrawing the schematic several times.

C) Embrace Electrodynamics.

For a high school student or member of the general public, electronics is often explained by analogy.

"Voltage and Current are like pressure, and the size of the pipe!"

"Electrons move around in wires, like marbles in a tube!"

That's fine if you are a regular person, not an engineer.


But if you are anywhere near electronic design, the above is worse than useless.  In fact, it is worse than not knowing anything at all.  The best place to start is basic electrodynamics.

- understand how current propagates through a conductor
- understand how current creates a field external to the conductor
- understand how power flows through a circuit (look up: Poynting Vector - at least this)
- understand how radiating electromagnetic fields propagate through space
- understand how currents and fields behave very differently at different frequencies

There is a big cut off between junior and senior people.  Of course, senior people were junior people... but... what made the difference?

Usually, at junior level, it's possible to set up little microcontroller projects, make high school competition robots, maybe set up a server somewhere or have an app control a switch, and so forth.  Maybe a ham radio project too.  All of this is right, good, and necessary!

Unfortunately, the skies turn dark typically in college, in the physics electrodynamics course.  At this point, electrodynamics is new to you, purely conceptual, and not much of a problem solving friend (yet).  But never sell your old book.

The following are some of the things that separate good engineers from great ones.

- High Speed Design.  

In the worst case, at high frequencies an input trace radiates like an antenna, skips over a fancy circuit entirely, and that circuit's output trace picks up the raw input signal just like an antenna.  Countless engineers have underestimated high speed design, and worse, didn't understand the nature of the problem they encountered.  If you are doing anything from 100 MHz to GHz and above, no matter how trivial a job it may seem, read up on high speed design first.


- Information Theory.  

This is another area where consultants make a lot of money.  What usually happens, is that a software development team will run smack into physical layer limitations.  It is common in the software world to run at GHz speeds with Gbits worth of resources.  Unfortunately, the 'last mile' or the 'edge' where the electrical engineers are working on things, has only 1990's dialup levels of throughput.  This always comes back to fundamentals: how much data can you really propagate at the wavelength used?  Nyquist-Shannon stuff.  You will run into it. Information is physical, and in your case, it's electrodynamic too.


- Electromagnetic Effects.  

There are a lot of things that are obvious in retrospect, but surprising to young engineers.  Let's pick an easy one: AC current through a wire.  Would it be surprising to know that the type of insulation around a wire has an effect on the current?  Consider that a varying current produces an electromagnetic field.  If something around the wire acts against the electromagnetic field's formation, then that directly impacts the current in the wire as well.  The frequency, material and other factors contribute to the size of the effect, as do other nearby objects and fields.  Learn to envision fields.  Are you putting two large inductors next to each other?  Will their fields overlap, and interfere?  It's obvious if you think about it, and have at least a vague sense of what the field 'looks like' and 'how big it is' compared to nearby components.


- Relativistic Effects.  

It's surprisingly easy to accelerate electrons to relativistic speeds (outside of a material, of course).  This isn't a big deal.  But what is a big deal, is that electromagnetic fields appear differently, depending upon your reference frame.  

Remember that a moving charge creates a field?  There are reference frames in which the charge isn't moving at all.  Such as the reference frame of the moving electron itself!  Maxwell's equations have relativistic solutions, everything works out, but it's a very easy mistake to presume that a field 'looks the same' to everything in and around it.  

Electrons, particulary radiated ones, 'live' in a world full of relativistic effects, time dilation, and so forth.  You will eventually run into this.  To give an example, old TV's (using cathode ray tubes) would be impossibly blurry, were it not for the fact that electrons get accelerated to a substantial fraction of the speed of light in the picture tube.


- Quantum Effects.  

You may have heard of the double slit experiment, where photons collectively act in a wavelike manner, going through two slits.  Put enough photons through the double slit, and you will clearly see interference patterns where the photons landed on the far side.  Did you know that this works for electrons too?

The point here, is that electrons aren't really just tiny charged balls.  Yes, their field can be considered 'round' in a sense, but they also follow a very strange (to most people) set of rules.  They are Fermions, which means that that two electrons cannot occupy the same state as each other... hence complex atoms can exist.  They also have Spin, which is sort of like a top spinning... except they aren't physically spinning at all.  And that spin is quantised to one of two fixed values.  Sensors, light emitting diodes, transistors, superconductors, and many other effects rely upon the quantum nature of electrons.  

Even if you don't fully understand the quantum nature of electrons, get comfortable with using quantum effects.  Many semiconductors are built out of one or more quantum effects.


- Physics in General.  

Tube electronics uses something called thermionic emission.  Which is a fancy way of saying that electrons can fly off heated metal if there is a more positive charge nearby to attract them.  Perhaps you haven't learned of this effect, but it still occurs, regardless of your level of knowledge.  Certain types of radioactive decay can also produce small currents as well.  Anything that is emitting charged particles, is creating a current.  

Other common scenarios are just rife with charge accumulation.  In high school physics, one of the best ways to accumulate charge is to rub asphalt and rubber together.  Which is what every car going down the road does, and why in dry weather you can often draw a spark touching a car door.  That's hundreds of kilovolts showing up, regardless if planned for, or not.  Modern semiconductor gates don't do too well when hit with a few hundred kilovolts, even if there isn't a lot of current associated with it.  If that is new information to you, you might understand it only as 'mysterious failures during dry weather.'

All of these effects and more are active all the time.  Sooner or later one of them will interfere with what you are trying to do.

Unless you are aware of such things, you won't be able to do anything about them.  Embrace physics, electrodynamics especially, and suddenly a lot of 'mysterious' problems become easy to deal with.