FEDEVEL
Platform forum

cefreyle , 03-20-2019, 06:19 PM

Hi Robert ,

I decided to write you first of all to thank you for share your knowledge. It is precise and useful for all of us that love electronics. However, more than knowledge I have found in you a model to follow and I think you have experienced similar situations than me. I am currently in a exchange in another country because in my country (Colombia) there is not a real electronics industry (Like Slovakia, I guess). Even I see that usually electronics engineers work in developing apps or web (It's on high demand and well paid). I want to work in electronics and make useful devices (Hardware) but well, I recognize that the process takes a long time and money to become a hardware developer.
I want to ask you about the failures that you have had in this process (mistakes). I want to know from it how to avoid the perfectionism that delays our projects (Like you know, when to stop reviewing the project files, the design guides, other video or datasheets), the fear to fail and all this things in our minds that keeps us apart sometimes of enjoying the projects (Never its good enough). Also I will thank the experienced people in this community if the share their answers to my question.



Thank you

Kevin Amadiva , 03-21-2019, 02:53 AM

Following this string closely, similar case in Kenya.

Paul van Avesaath , 03-21-2019, 06:53 AM

one advice is that in every contact with a client (or even yourself) you have to really set the DOD parameter! DOD = Definition of Done!! if the client is happy then your done basically... but you have to communicate and specify what the cllient expects from you and you from him.. as long as that is clear then you have a strong base to work from. have regular sessions to make sure both parties are still on the same level.. let the client review the schematics in between.. also do a board review with the client at the end. show them why you did stuff a certain way.. if the client approves then you are done.. no need to go any further unless you want it yourself.. (but a client won't pay you for it) it is good practice to try and go the "extra mile" and make a bit nicer than they need or want.. so you will get good reviews of your work (also ask for a recommendation from you client) also good work will get you referrals to other companies.. and also you might end up with more projects at the same client.. keep things clear / open / ask a lot questions and do not make assumptions..listen to what the client wants and needs.. and give feedback on it.. make sure you are on the same page!! do regular updates.. be honest.. and do not take on projects you are not comfortable with.. again back to the DOD.. as long as that is clear then you should be fine! and no one gets dissapointed.. perfectionism is a real downfall in some projects.. but perfection is never achieved in my opninion. just try and make the next board even better than your last.. final piece of advice.. try and take projects that pay by the hour.. not fixed price by project.. because an engineers judgement is most of the time too positive and you will end up doing it for half the pay because you spend twice as many hours on it.. hope this helps a bit

robertferanec , 03-25-2019, 11:16 AM

Designing big and expensive boards is extremely stressful - at least for me. When designing complex boards (such x86), there is not much space for mistakes and therefore it is a lot of pressure - design has to be right at the first time. The time when it is really enjoyable, is when the board boots up for the very first time.

It is not necessary to be perfectionist - important is to be precise and consistent when designing and checking schematic and layout. I have exact steps how I work on designs and I try do not break these steps e.g.
- I add into schematic everything what I have in mind and only remove things when they do not fit on the PCB during initial placement,
- I try do not change schematic once it was fully checked,
- I try to make the initial design similar to the reference board and in the next versions I experiment more and more,
etc...

Some people have different approach - they try to build boards as quickly as possible, test them, adjust the design and manufacture new version of board. This is not usually my approach - even for simple boards I try to build paper models and check everything. Mistakes in boards not only cost money, but they also cost time and cause delays in projects (debugging the problems with hardware and software engineers, fixing the problems, sending the boards into production again) and .... if people make mistakes in one version of board, you never can be sure if they will not make mistakes also in the second version .... I have seen people who needed 3-6 versions to actually finish a board. I usually need 1 or 2 versions.

TIP: Once the PCB is sent to production, do not open the project until the finished board is back

Hope this helps.

Paul van Avesaath , 03-28-2019, 08:55 AM

TIP: Once the PCB is sent to production, do not open the project until the finished board is back

but...but..but.. it's so hard to let go!!

robertferanec , 03-28-2019, 11:06 AM

When I open it, it just always makes me wonder if I have done everything correctly and I start checking what I have already checked multiple times