Is this the lamest "design idea" in EDN...?

[mikeselectricstuff]

Reading EDN today I came across this "design idea" -
How to avoid using a seven-segment decoder
Are they really so short of material that they have to pubslish something as lame and bleeding obvious as this...?

I fully expect to see next month's edition to have something entitled "Save microcontroller pins using an external seven-segment decoder"

[KTP]

Wow...I mean wow.

The only thing now they can do that would be lamer is an issue on how to divide a voltage using resistors. 

[RayJones]

WTF?

They left off the decimal point, now that is lame.

[JohnS_AZ]

Yea, that's pretty lame. That column has been going downhill for, what, the last year and a half? I remember the brilliant "new" idea of creating any color by driving 3 LEDs with PWM. *sigh*

In their defense, however, I imagine all they can do is select the best of what they receive as submissions.
Some of the sharpest engineers I've found on the web are right here on this forum. How many of you would go to the trouble to proto and write up a design idea, edit it a couple times, make the drawings, when all they pay is $150 US for a published idea?

If the $150 sounds like a lot to you, send some stuff in. It sure look like they could use the help!

http://www.edn.com/info/2236-Design_Ideas_Writers_Guide.php

[EEVblog]

Ok, that's pretty dumb-arse.
At least they included the table. Sham it's not valid if you have to use different pin configurations which is not uncommon.

Dave.

[Zero999]

It's pretty shitty.

Here's PIC tutorial which is much better than that.
http://www.gooligum.com.au/tutorials/baseline/PIC_Base_A_8.pdf

A better article would've been how to use a CD4026 to control an LED display with only a few I/O pins, as in the project linked below.
http://www.electronics-lab.com/projects/automotive/006/index.html
Unfortunately, whilst it's a good project, like most on Electronics Lab it contains an error: R1 isn't required because the CMOS IC limits the current and if it didn't, a separate resistor would be required for each LED.

[Bored@Work]

EDN was sold recently. They were on the auction block for some time and had to shed employees. Still they were among the lucky ones, they survived. Their previous owner completely closed other publications.

When the new owner took over they had some serious teething problems migrating the website and putting issues out. It is only recently that the website is almost back to normal. Maybe they get the magazine quality under control in the future.

[Zero999]

I think I'll email them my idea, hopefully they'll pay me for it.

[Kiriakos-GR]

I think I'll email them my idea, hopefully they'll pay me for it.

Tell them that you are nephew of Thomas Edison , they will pay   

[ziq8tsi]

Any such lookup table should include the extra hex digits A to F as well.  Even if you think you only need decimal in a particular project.

It is a useful debugging aid, and the extra storage cost is minimal, especially since it also saves you the need to explicitly check for invalid digits.  (You can just mask off four bits and lookup, rather than having to branch if the value exceeds nine.)

This is especially important on low end PIC devices where lookup tables are implemented by adding to the program counter -- a non-decimal digit lookup could otherwise crash your code.

[JohnS_AZ]

I've rethought this since posting that $150 was too cheap.  If you got three things published that would cover a new Rigol scope.

What's more, work in a link to Dave's site here and I bet he'd get a serious bump in site views.

[jpwack]

maybe it's lame, but useful as hell and a lot of newcomers may not know it. Just a month ago I wrote this:

Code: [Select]

/**
 * Easy integer-to-SevenSegmentDisplay conversor
 *
 * @param input	An integer representation of the number to display in the SevenSegmentDisplay
 * @return	The SevenSegmentDisplay representation of that number
 */
unsigned char sevensd_number( unsigned char input) {
	switch(input) {
		case 0: return 0xE7;
		case 1: return 0x84;
		case 2: return 0xD3;
		case 3: return 0xD6;
		case 4: return 0xB4;
		
		case 5: return 0x76;
		case 6: return 0x77;
		case 7: return 0xC4;
		case 8: return 0xF7;
		case 9: return 0xF6;
		
		case 0xA: return 0xF5;
		case 0xB: return 0x37;
		case 0xC: return 0x63;
		case 0xD: return 0x97;
		case 0xE: return 0x73;
		
		case 0xF: return 0x71;
			
		default: return 0x8;
	}
	return 0x8;
}

it was made for an MSP430 with lots of program space, so don't blame me for it's un-optimized state.

cheers

[Zero999]

I think you shoud've written the numbers in binary not hexadecimal because it would make the code easier to understand.

[mikeselectricstuff]

maybe it's lame, but useful as hell and a lot of newcomers may not know it. Just a month ago I wrote this:

EDN is  a magazine for electronics professionals, not newbies. Publishing something as basic as this in such a publication shows the editorial staff are clueless or desperate.

[mikeselectricstuff]

maybe it's lame, but useful as hell and a lot of newcomers may not know it. Just a month ago I wrote this:

Code: [Select]

/**
 * Easy integer-to-SevenSegmentDisplay conversor
 *
 * @param input	An integer representation of the number to display in the SevenSegmentDisplay
 * @return	The SevenSegmentDisplay representation of that number
 */
unsigned char sevensd_number( unsigned char input) {
	switch(input) {
		case 0: return 0xE7;
		case 1: return 0x84;
		case 2: return 0xD3;
		case 3: return 0xD6;
		case 4: return 0xB4;
		
		case 5: return 0x76;
		case 6: return 0x77;
		case 7: return 0xC4;
		case 8: return 0xF7;
		case 9: return 0xF6;
		
		case 0xA: return 0xF5;
		case 0xB: return 0x37;
		case 0xC: return 0x63;
		case 0xD: return 0x97;
		case 0xE: return 0x73;
		
		case 0xF: return 0x71;
			
		default: return 0x8;
	}
	return 0x8;
}

it was made for an MSP430 with lots of program space, so don't blame me for it's un-optimized state.

cheers

That's way less then un-optimised - Read up on arrays....

Code: [Select]

const char sevenseg[16]={0xE7,0x84,0xD3,0xD6,0xB4,0x76,0x77,0xc4,0xF7,0XF6,0xF5,0x37,0x63,0x97,0x73,0x71};

ledsegs=sevenseg[digit & 0x0f];

[jpwack]

That's way less then un-optimised - Read up on arrays....

Code: [Select]

const char sevenseg[16]={0xE7,0x84,0xD3,0xD6,0xB4,0x76,0x77,0xc4,0xF7,0XF6,0xF5,0x37,0x63,0x97,0x73,0x71};

ledsegs=sevenseg[digit & 0x0f];

woah, talking about optimization and to make it compatible with my shitty function:

Code: [Select]

const unsigned char sevenseg[16]={0xE7,0x84,0xD3,0xD6,0xB4,0x76,0x77,0xC4,0xF7,0XF6,0xF5,0x37,0x63,0x97,0x73,0x71};

#define sevensd_number(__NUMBER__); sevenseg[__NUMBER__ & 0x0F];
#define sevensd_number_dot(__NUMBER__); ((sevenseg[__NUMBER__ & 0x0F]) & 0x8);

where is our money?