Hardware Design - Eliminate the silly things

 

There are some steps in the hardware design and testing which are not given preference and may hound you badly later. Below are some of those which we are categorizing as silly:

1. The first most among this criteria is the jumper wires which are used to connect two points on a board. Test Engineers while using jumpers tend to use self made jumpers with nice tinning which could turn out to be the culprits most of the times. It is better to buy those cheap commercially available jumpers to have a neat connection and eliminate breaking your head in debugging and spending sleepless nights.
2. Criss cross problem - This is the universal problem faced by the designers while designing circuits. For example, a reversal of TX and RX in circuit. Most % of the designers world wide would have gone through this criss cross issue. Study the pin description of the interfaced ICs thoroughly.
3. Missing instructions - This is one of the silliest hardware problems. Most of the designers are experts and doesn't pass on the instructions required to next level. For example, passing the required instructions to PCB Layout engineer, test engineer is very important. A simple case could be providing an input to PCB layout engineer on the placement of Acclerometer IC.
4. The oz problem - This is another silly problem where you and vendor get the terms, costing for a specific copper thickness but in the Gerber you mention different copper thickness size. While PCB Layout engineer do review the Manufacturing instructions, Designer has to look into it.
5. Missing updates - Most of the times we do have changes to versions of the schematic. Designers do create documentation once and tend to miss the updates. This results in inaccurate information flowing down the cross functional teams.
6. Legacy library - This is one of the major miss seen in some of the designs. There shall not be a formal library management system and designers tend to use old erroneous libraries leading to catastrophic issues.
7. Avoiding Self review - While this cannot be categorized as a silly thing, it has become a silly thing because of the exhausting design process involved and designers showing laziness to do a self review there by errors propagating to next level. Designer knows in and out of the design,so, self review is the best review
8. Powering up a board without proper checks - Once a new board is received, many test engineers believe the quality process of the Assembly house and cut short their checks. They tend to restrict the checks to just short without any component checks there by burning the board on first power up attempt due to other issues like component orientation, wrong component placed, unintentional shorts on the board, etc.
9. Zero documentation - Lesser the documentation the more the organization shall have problems and the focus must be on 100% documentation

The major requirement to eliminate all these silly issues is to have a proper work flow defined with a strict third party review mechanism in place. Reviewer should be given the time which results in a quality output from the project team.