Traps for young players

[whiskthecat]

I think a good video blog topic would be just listing a bunch of traps for young players. Perhaps we can post some ideas for things that have caught people off guard in the past. Like twisting your high frequency serial wires around and then wondering why your data isn't proper  . Or sending LSB when you should have sent MSB  .

[Psi]

Not strictly a trap for young players but a good tip..

When cutting off the insulation around a multi-core cable using side cutters always let the cutters dig-in then rotate outwards before you cut.
Grab the insulation with the cutters, twist, then snip.
This way you will never cut or mark any of the inner wires.

[DenzilPenberthy]

Part of my job is to help students out with their projects etc. I'd say my top things I see so far are:

If your project involves RS232, you WILL get Rx and Tx muddled up.

I2C needs pull-up resistors to work

If you specify a mixture of 3v3 and 5v parts, you will cause problems for yourself.

When breadboarding/prototyping play the 'long game'. Any time you spend making a super neat professional job will be repaid five to ten times over. I.e wires neatly routed, colour coded, correct length, heatshrink where wires meet connector pins, connectors/switches mounted on a panel not just flapping about etc etc etc. This especially applies to 'I'll just make this one little alteration'. The little alterations add up and your nice neat PCB becomes a rats-nest of flimsy wires and dangling connectors that will never work reliably.

= and == have different meanings in C

[Psi]

never push thick wire into your breadboard, it just ruins the hole. it wont make good contact with normal wire ever again after that.


Never let other people use your "good tools".
Have two of everything and let them use the not so good stuff.
9 times out of 10 if you hand someone your "good" side cutters and turn around for a second when you look back they will be trying to cut 3mm steal coat-hanger wire with them.

Never trust that your gerbers were generated correctly. always review them in a gerber viewer before sending to PCB fab

Always add provision for extra features on your PCBs. It costs nothing to add extra tracks/pads and break out any unused i/o to somewhere. It will save you time in the long run.

[Simon]

the list will go on and on and traps will depend on the particular area.

[abaxas]

Never let anyone involved with the construction industry within 500m of your tools.

[SeanB]

Never let anyone involved with the construction industry within 500m of your tools.

that includes any hammer, spade or other tool that can possibly be used as a hammer or as a chisel.

Builders use every tool as either a hammer or a chisel, including an axe.

[DanielS]

Builders use every tool as either a hammer or a chisel, including an axe.

And engineers may use 20lbs sledge hammers for tweaking precision alignments

[Neilm]

When cutting wires for a project that uses a battery - cut them one at a time. Otherwise you find that you left the battery connected while you measured the wire and there is now a nice hole in your cutters.

[SeanB]

Remember that resistors, diodes and transistors can and do get hot in operation. Thus make sure the power rating for the application is correct, and that you do derate appropriately when they are running at elevated temperature or in confined airflow.

As well remember resistors also have a voltage rating, separate from the power rating. As well power resistors when run at high power are not regarded as being insulated, as if they are running at close to dull red heat the ceramic casing can be slightly conductive leading to unanticipated current paths.

[kingofkya]

Never let anyone involved with the construction industry within 500m of your tools.

Yeah that the truth. OR show an "electrician" them any of your projects i tried to explain watts vs amps to one once, he just could no figure out how you can few amps input could have a 50amp output @5v


 Tools, I watched i guy go though 6 dimond bits after drilling 4 holes the big expensive $20-40 a pop bits. Full power and push as hard as you can when drilling steel right?


The biggest trap i have seen lately is impossible to find issues cased by cheap 5v supplies. Just buy a good supply at least when your testing:)

[sunnyhighway]

Never assume the earth pin in the wall socket is actually connected to earth. Measure the voltage between the earth pin and a metal water pipe or tap.
If you measure something like half the mains AC voltage something is wrong.

I once had my mains fuse-box replaced by a certified electrician because I needed to upgrade from 3 to 12 groups. When all the new the kitchen appliances were installed by some other company it became apparent there was something wrong with the electrical wiring because the metal frame of the new electric stove gave me a nasty shock. I took out my multimeter and tracked the problem down to the mains fuse-box.

After opening the fuse-box the problem was quickly found.
Although all the earth wires to the wall sockets were connected to each other, there was no connection between them and the bare earth wire connected to the actual earth. 

[Cside]

1. Buckland's law states that projects will take twice as long to complete then you predicted - even after taking bucklands law into consideration
2. Most of the features you think you can implement into a project will not make it on time or will be buggy.
3. Instead of designing a machine/device/board with many features and the occasional bug, make a machine/device/board with minimal features but works perfectly first time every time.
4. It may sound logical but it wont work. Your schematic is broken until proven otherwise. Breadboarding is only half the proof.

[w2aew]

1) Circuit simulators are only as good as the models and the care you take in using them.  Always apply common sense!  A circuit simulator will have no problem putting 1000 amps through a 1N914A diode - it doesn't know that you'd let the smoke out of it.

2) The first 90% of any project takes 90% of the time.  The last 10% of the project takes the other 90% of the time.

3) When making a project, you can make it good/fast/cheap -- Pick any two.

[DanielS]

Never assume the earth pin in the wall socket is actually connected to earth. Measure the voltage between the earth pin and a metal water pipe or tap.

Never assume metal pipes are connected to ground: many older buildings that have been recently remodeled contain PEX pipes so you can only trust "metal pipes" as far as you can track them from the feed pipe.

The way I test ground is by connecting a 250W halogen lamp from live to Earth. If I get close to 0mV from neutral to Earth, I know neutral and Earth are shorted together at the outlet. If I read 100+mV extra, then I am relatively confident they come together at the electrical distribution panel: 2A x 8mOhm/m for #14 copper x 10m = 160mV. While you could do this test with the load connected across live-neutral, it would not tell you if the ground is actually capable of passing remotely significant current.... such as something being grounded to a metal pipe which connects to a PEX feed so you end up with a water resistor in the ground circuit.

[VK5RC]

Never switch on your project immediately after you have finished it,  go and have a coffee etc,  come back recheck the circuit then switch it on.

[diegoterc3]

ALWAYS!! check for VCC and GND continuity.

[redtails]

The biggest trap i have seen lately is impossible to find issues cased by cheap 5v supplies. Just buy a good supply at least when your testing:)

  ground loops (earth loops) coupled with cheap power supplies will ALWAYS cause trouble no matter how good of an engineer you are and they WILL cause the weirdest problems ever.

I've seen A/V systems installed by professional specialized companies that were completely swamped by noise (flickering beamers, that typical HF noise on the microphones) simply because someone connected their crappy laptop somewhere in the vicinity. Frankly speaking, I have no solution to this other than carefully selecting which phase you put what equipment on, and not actually connecting the earth to any sensitive audio/video gear

[jay]

I'm still a newbie but I've learned few things.. One is that the example circuits in datasheets cannot be blindly trusted. They rarely specify all the details.

[BillyD]

Try to avoid situations where you have to repeatedly switch your meter between voltage and current mode. Do your current measuring first and get it over with, or dedicate one meter to current measuring and leave it there. If you have to keep swapping over and back, eventually you'll get it wrong and blow something.

Avoid lending tools to anyone. Many of my tools are over 20 years old, and over half of any breakages were done by someone else. One of the problems with unbreakable tools is that they are useful for breaking other tools.

[TMM]

Never assume the earth pin in the wall socket is actually connected to earth. Measure the voltage between the earth pin and a metal water pipe or tap.

Never assume metal pipes are connected to ground: many older buildings that have been recently remodeled contain PEX pipes so you can only trust "metal pipes" as far as you can track them from the feed pipe.

The way I test ground is by connecting a 250W halogen lamp from live to Earth. If I get close to 0mV from neutral to Earth, I know neutral and Earth are shorted together at the outlet. If I read 100+mV extra, then I am relatively confident they come together at the electrical distribution panel: 2A x 8mOhm/m for #14 copper x 10m = 160mV. While you could do this test with the load connected across live-neutral, it would not tell you if the ground is actually capable of passing remotely significant current.... such as something being grounded to a metal pipe which connects to a PEX feed so you end up with a water resistor in the ground circuit.

That test has far too many traps for young players haha. GFCI/RCD protected circuits, faulty devices floating a disconnected ground wire, etc.

What i would do is plug a big extension lead into the outlet and bring it to the distribution board. Isolate the circuit and check continuity from the extension lead to the board, then work out if the distribution board has an adequate ground. Now that's foolproof.

[w2aew]

A few more that I see all the time:

1) Many/most high frequency function generators have a 50 ohm output impedance.  Thus the output amplitude will be different for a 50ohm load and a high impedance load.  Most generators of this type let you *tell* the generator what the load is, so that is can adjust the output so that the displayed voltage amplitude appears at the load.  Many people forget this, or don't know about it.  (see my video on the topic).

2) AC voltage measurements on a DMM are designed for a few kHz or less in frequency, in most cases.  Don't try to measure the amplitude of 10s or 100s of kHz or higher, unless you KNOW the meter is rated for it.

3) Scopes that don't auto-sense 10x probes, or when using probes that don't have the sense pin - be sure to set the probe attenuation factor on your scope channel appropriately.

4) 1x scope probes are ONLY good for low frequency work, since they load the circuit with a fairly heavy capacitive load (>100pF typically).

[abaxas]

Never assume that more expensive is better. Sometimes the cheapest 'badger arse' tools are the most useful/accurate.