Author Topic: FF suggestion: "Childhood's End".  (Read 2223 times)

0 Members and 1 Guest are viewing this topic.

Offline ElectroIrradiator

  • Frequent Contributor
  • **
  • Posts: 614
  • Country: dk
  • More analog than digital.
FF suggestion: "Childhood's End".
« on: September 30, 2013, 12:53:57 pm »
Sorry about the silly subject name, couldn't think of a better one. ;)

If you, Dave, are looking for relatively simple subjects for FF, then how about one focusing on the imperfections encountered when building electronics? One very common mistake among beginners I tend to notice, is that the beginner naturally start their design career using a set of very simple design rules and equations. Initially, if their projects are not too complicated, then these rules-of-thumb works fairly well, and the widgets does as intended.

Yet sooner or later the aspiring EE enthusiast often runs headlong into a brick wall, as he attempt to design more complicated circuits. The simple design rules no longer results in a working widget, with no obvious reason behind the failures.

Examples of what I am thinking of could be:
  • Using a prototyping board for high frequency or high current circuits, where either dead bug construction or a proper ground plane PCB layout would be almost mandatory.
  • Designing a linear PSU from scratch, using only the most crude rules-of-thumb for the design, thus not taking many important factors into account: Transformer losses and transformer cooling, mains voltage variations, diode voltage drops at peak charging current, ripple current limits on filter caps, proper heatsink design, placement and orientation, etc.
  • Assuming all components are noiseless and perfectly linear.
  • Thinking of the wires connecting components as perfect connecting pieces operating at superluminal speed. IE. zero phase delay, inductance, resistance plus zero enclosed area. (Ground loops in audio circuits!)
  • Vbe of any BJT is presumed to be 0.65000000....V at any base current and temperature, with hfe similarly being constant.
  • Insufficient or non-existing local PSU rail decoupling.
  • Not using various 'best practice' parasitic suppressors, often combined with questionable physical layout. This in turn results in parasitic oscillations and thus erratic and unexplained behavior of the circuit.
It is very easy to come up with a lot more examples like these, of course.

The purpose of the FF should not be to educate people on the details of all these subjects, but to notify people that these practical hurdles exist, so people ought to keep an eye open for them. Part of the blog could perhaps show some cases, where a design breaks down in practice, despite looking fine on paper.
« Last Edit: September 30, 2013, 01:03:28 pm by ElectroIrradiator »
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 32005
  • Country: au
    • EEVblog
Re: FF suggestion: "Childhood's End".
« Reply #1 on: September 30, 2013, 01:20:47 pm »
Good suggestions, thanks!
 

Offline Alexei.Polkhanov

  • Frequent Contributor
  • **
  • Posts: 683
  • Country: ca
Re: FF suggestion: "Childhood's End".
« Reply #2 on: September 30, 2013, 05:24:44 pm »
Sorry about the silly subject name, couldn't think of a better one. ;)
« Last Edit: April 03, 2014, 05:54:10 pm by Alexei.Polkhanov »
 

Offline chickenHeadKnob

  • Frequent Contributor
  • **
  • Posts: 892
  • Country: ca
  • doofus programus semi-retiredae
Re: FF suggestion: "Childhood's End".
« Reply #3 on: September 30, 2013, 06:37:45 pm »

Examples of what I am thinking of could be:
  • Designing a linear PSU from scratch, using only the most crude rules-of-thumb for the design, thus not taking many important factors into account: Transformer losses and transformer cooling, mains voltage variations, diode voltage drops at peak charging current, ripple current limits on filter caps, proper heatsink design, placement and orientation, etc.


Isn't this item more of a right of passage than a brick wall? I had to go past this without the help of the internets, no one taught me that thermal design was important I had to learn the hard way. I still eschew calculating and go straight to experimenting with heat sinks that look like about the right size, so maybe I haven't truly advanced past this personal brick wall. What I am trying to say is the first hand tactile experience of failure is more valuable than being coddled past that experience.

I realize this is just a variation of the old man rant, I had to walk to to school uphill in a blizzard both ways , ect ...
But I am getting to be an old man, so I feel entitled.
 

Offline ciccio

  • Frequent Contributor
  • **
  • Posts: 652
  • Country: it
  • Designing analog audio since 1977
    • Oberon Electrophysics
Re: FF suggestion: "Childhood's End".
« Reply #4 on: September 30, 2013, 07:35:28 pm »
I found some "ugly" parts on ebay: http://www.ebay.com/itm/CCCP-Soviet-Military-Grade-Tantalum-Capacitors-K53-28-10uF-32V-K53-/251028301842?pt=LH_DefaultDomain_0&hash=item3a7273e812
Those tantalums were used in military applications,  and I'm sure they were better than those standard "drop type" that arrived in Italy from Soviet Union years ago, when tantalum became rare..
They were the least reliable part I ever used: they shorted after some week of use.
Other funny thing from Soviet era : I bough a stock of NE555 timers (the Russian equivalent) that had a pin pitch of 2.5 mm exact, not 2.54.. The Soviet Union was truly metric.

To stay on topic, I would like to add another example to the OP list:
- a dirty PCB that transforms your first high power audio amplifier in a wonderful HF oscillator,  and your face that becomes bright red when you experienced colleague wash it with some Freon and "voilĂ " : it works...

 
« Last Edit: September 30, 2013, 07:41:42 pm by ciccio »
Ciccio

Strenua Nos Exercet Inertia
 

Offline AndyC_772

  • Super Contributor
  • ***
  • Posts: 3699
  • Country: gb
  • Professional design engineer
    • Cawte Engineering | Reliable Electronics
Re: FF suggestion: "Childhood's End".
« Reply #5 on: September 30, 2013, 07:47:55 pm »
One thing which, in my experience, isn't taught well but is soon learned 'on the job' is the importance of physical layout.

I think we probably all have an idea that parasitic elements can be important, but not necessarily an idea of how significant they are and when.

There are some simple experiments that can show them, and which are well worthwhile (IMHO) to show their magnitude. Things like:

- probe a fast square wave with a conventional scope probe using its ground lead and clip, then again with a short GND spike. Show the effect of lead inductance.

- show the difference in shape of that same square wave at the driving end vs the receiving end. If it's series terminated, explain the 'step' halfway up the wave at the driving end.

- Try replacing the current sense resistor in a SMPS with a piece of copper wire of the same resistance (also shows the effect of lead inductance).

- Show the effect of the DC resistance in a power cable or connector, calculate how much heat is generated in each, and the effect on the supply's efficiency.

- Show why a small ceramic capacitor is still needed on the output of an SMPS even where there's already a much bigger electrolytic

Offline ElectroIrradiator

  • Frequent Contributor
  • **
  • Posts: 614
  • Country: dk
  • More analog than digital.
Re: FF suggestion: "Childhood's End".
« Reply #6 on: September 30, 2013, 08:38:36 pm »
Isn't this item more of a right of passage than a brick wall? I had to go past this without the help of the internets, no one taught me that thermal design was important I had to learn the hard way. I still eschew calculating and go straight to experimenting with heat sinks that look like about the right size, so maybe I haven't truly advanced past this personal brick wall. What I am trying to say is the first hand tactile experience of failure is more valuable than being coddled past that experience.

I may have phrased my suggestion incorrectly.

I am not suggesting a blog showing how to solve particular technical problems. I am suggesting a blog, telling people to be on the lookout for the various problems, which can arise due to non-perfect components and ditto layout techniques.

For instance some people may be stumped, when their latest creation - built on prototyping board - fails to work as expected. Built using 74HC' DIP chips and running at 50 MHz clock frequency.
 

Offline chickenHeadKnob

  • Frequent Contributor
  • **
  • Posts: 892
  • Country: ca
  • doofus programus semi-retiredae
Re: FF suggestion: "Childhood's End".
« Reply #7 on: October 01, 2013, 03:34:32 am »
I may have phrased my suggestion incorrectly.

I am not suggesting a blog showing how to solve particular technical problems. I am suggesting a blog, telling people to be on the lookout for the various problems, which can arise due to non-perfect components and ditto layout techniques.

Fair enough; It is just that I  wouldn't deny some young person the experience that I had. I was a teenager crazy about computers and things digital in the late 1970's and didn't want to mess with power supplies at all but was forced to, so that my creations could run. I could have asked for help. My father was an electrical engineer and my brother a physics student building circuits for a unversity cyclotron, but I was stubbornly self-reliant. My parents didn't even know of my computers until they figured out the herring-bone hash and interferance on the TV was comming from  my room. My open frame computers were terrible RF noise makers  >:D. The act of going to the unversity library and solving the problems myself was the real value of the exercise. I was compeled to exit my interest zone and comfort zone at the same time.
« Last Edit: October 01, 2013, 04:29:03 am by chickenHeadKnob »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf