Do NOT Over-Engineer Things!
![]() |
4.4 (5) |
As I pen these words, I am deep in the murky mire. I've done what I usually do, which is to over-engineer something, and now I'm very, very sorry. I should always remember to adhere to the KISS principle ("Keep It Simple, Stupid!")
The problem is that I never learn. I get an idea into my head, I'm carried away on a wave of enthusiasm, and things grow and grow in the telling. Here's a simple example. I'm currently in the process of building something I call a "Man versus Woman Display-O-Meter". I'm not going to talk about this in depth here (that will be the topic of a mega-article sometime in the future).
Suffice it to say that I'm building it in a Steampunk Style, which involves a lot of polished brass and antique black analog switches and meters and suchlike. It will look a little like the example shown below, which was created by a friend of mine in California who is building a Dr. Who TARDIS control console in his front room (don’t ask – I already told you he's from California, it's not his fault).

One example of a Steampunk look-and-feel.
My own creation is going to be mounted in an antique radio cabinet from 1929 but – once again – that's a story for another day. As part of my project, I want to have what appears to be a small coal-fired furnace mounted in the front panel. The sort of look I'm going for is as shown below:

A quick sketch of my proposed furnace.
Be gentle with me... this is just a quick sketch I threw together in Microsoft Visio. In the fullness of time, the main panel will be brass and the "furnace door" will also be formed from brass. Of course I don’t actually intend to have a real coal-fired furnace (I'm not as daft as I look ... but then again, who could be?). I just want something that looks like one.
Behind the furnace door will be a circle of rippled glass (I have this already. I picked it up at a local stained-glass workshop. The effect will be gorgeous!). Through this glass you will see flickering red, orange, and yellow "flames", which I intend to implement using LEDs and a microcontroller. And this is where things started to go wrong...
But before I reveal the true horror of my predicament, I should perhaps point out that I decided to construct a cheap-and-cheerful mock-up with which I can experiment. Thus, I started to play with some card and foam board (I got glue everywhere). First I created a portion of the front panel featuring the furnace door as illustrated below:

A card and foam-board mock-up of the furnace door.
Next I constructed a sort of cradle to hold the body of the furnace; the body itself was created out of plastic pipe I picked up from The Home Depot. Now, the furnace door on the front panel has an outside diameter of 3.5 inches. However, the inner glass area (with the vertical bars) has a diameter of only 2.5 inches. The main body of the furnace is 4 inches in diameter, so I use a 3-to-4 inch expander. The fact that the furnace expands from 3-to-4 inches gives it a nice impression of "depth" when you look through the door. I then sprayed the inside of the furnace black, which – when coupled with the smoked or rippled glass in the door – really intensifies the depth effect.

A rear view of the body of the furnace resting on its cradle.
The idea is to have a bunch of LEDs mounted on the back plate. But which LEDs should I use? Right from the beginning, my chum Alvin told me that I needed only a few very bright LEDs. But would I listen to him? Of course not! I happen to have several hundred 5 mm red, orange, and yellow LEDs at my disposal, so I decided to use those instead.
The result is that I ended up drilling holes for 168 LEDs in the back panel. My friend Joe (who heads up the Hobby Electronics and PIC Microcontrollers communities here on Techbites) built me a special PIC-based controller board that can individually 24 channels of LEDs using Pulse Width Modulation (PWM). So, all I have to do is to wire up my 168 LEDs into three groups of red, orange, and yellow, where each main group comprises eight subgroups of seven LEDs (if you see what I mean).
The following image shows the inside of the furnace back panel with all of the LEDs inserted. They aren’t lit up – all I'm doing is holding it up so that the ambient room light is shining through:

Inside view of the furnace back panel with ambient light illuminating the LEDs.
And here's another image showing the back-side of the back panel with the lets and their terminal wires standing proudly to attention. At this stage things still seem to look nice-and-neat and manageable:

Backside view of the furnace back panel (still looking nice and neat).
Last but certainly not least, the image below shows the current state-of-play, which is that I am embroiled in a wiring nightmare. I now have all of the bunches of LEDs wired together – the next step is to connect everything up to the microcontroller board. I've got so much time and effort invested in this now that I have to see it through, but Oh how I wish I'd listened to Alvin... (don’t tell him, he'll be poking fun at me for ages):

I'm currently up to my armpits wading through a wiring nightmare...
Every day as I drive into work I see banks of super-bright LEDs in the brake and indicator lights of the cars in front of me ... and I think to myself that I could have quickly and easily done the whole job with only 24 of those little scamps (8 each of red, orange, and yellow).
Oh well... I've learned my lesson... at least until the next time (grin)...
User reviews
Average user rating from: 5 user(s)
occupational hazard
Yes, I, too, find myself starting with a reasonable goal only to feature-increment myself to death. If I don't reign myself in, the project will not even get off the ground let alone finished.
At this point, you may want to consider the theory of Sunk Costs: should you continue to throw good money (time) after bad? Humans are notoriously bad at this.
Here's a link: http://en.wikipedia.org/wiki/Sunk_costs
Good luck!
what can I say... o yes - Told you so :-)
Max,
Well some of his college ( who shall remain anonymous) did mention the small fact of OVER ENGINEERING, however the beer managed to wash this from his memory – You still have time to walk away from the red and yellow LED’s and just buy a bunch of flowers…..
PCB's are your friend
Max,
Perhaps having a PCB made would have saved you some time and burnt fingers.
Then again... wheres the fun in that.
I can't want to see this baby working.
Come on baby light my fire
So Max - You call this a "Man versus Woman Display-O-Meter". How exactly is it going to detect this and display it. There is crucial information missing - and I want to know if you intend to patent it. I could feature it in my silly patent series :)
over kill is good
Max,
looks like you've found a great use for those mounds of LED's. The project just wouldn't be the same with a mere 24 LED's! Can't wait for the video of the prototype in action!





