In my understanding we ended with two separate subjects (at least):
BOM
calculations for design.
For me those are absolutely separate tasks, but I suppose you may be assume that I am all the time speaking about one subject. The second started when you have written:
What I have last written it was about subject 2.
For example the element selection for DCDC converter to get the frequency needed, the current sweep range assumed, the output voltage needed, the max and min current in inductor and many more. I donât use in such spreadsheets exponential notations. As I know L will be in uH I just write equations in spreadsheet to get the result like L[uH]: 232 and not L: 2.32E-4.
Seeing 232 I write into other field that I will use L[uH]: 220 and this value is used for following calculations.
I not only donât wont to count how many zeros I see. I also donât wont to think that 2.34E-7 really means 234nF - let spreadsheet just show me it. In very rare cases when the result can be in big range I have it (one result) in few fields. Like:
Cout[pF]: 3E7
Cout[uF]: 32.5
I donât use spreadsheet scripting (I donât know it).
I:
copy csv (only part type, footprint and reference list) from KiCad to my spreadsheet (the part description and my sorting symbols are automatically filled from other spreadsheet page).
select the range (except few top lines),
sort it by column G and B,
select the resulting region to copy (without column G),
paste it as text (without equations behind them in first spreadsheet) to my output BOM (another spreadsheet).
It is really a little work. Much more work I spend on conservation of my element descriptions as parts become terminated, hard to get in small quantities,⊠So for some elements I write 3âŠ5 replacements.
I have tried it many times. First was Fortran during the 1970âs, then programming an HP45 calculator, then Basic and QuickBasic, Visual Basic in Excel about 10 years ago, then C++ maybe 4 years ago. I have some success, but I am not so good at it and it did not stick. I am much better at circuit design. At least my designs work and donât blow up.
One of my major life learnings (and surprises) is how specialized people are. I worked with one guy who was excellent at circuit modelling but designed circuits which would blow up during power up and power down. Other engineers who were excellent at circuit analysis but could not design a circuit which they had not seen. So my observation is that skills often tend to be very narrow and specialized.
Itâs typically specific to the spreadsheet. The advantage of using it, if available, is that it works on the spreadsheet in situ. The disadvantage is that itâs another language to learn and may not be powerful enough.
Funny, I moved to software because mistakes are easier to edit, in particular smoke generation mistakes. But I wouldnât use my own software without extensive testing to drive crucial systems. Wrong time on a table clock is one thing, overheating oven is more serious.
1.01.2020 one serie of fiscal cashboxes (in Poland) stop working. I donât remember any such problem ever (even at 2000). Many shops canât work as they are not allowed to sell anything without fiscal cashbox. Only what we publicly know is the cashbox writes something about âWrong timeâ and it canât be updated remotely. Probably rules are such that such cashbox after being certified canât allow for software changes. I have read that long queue was formed for repair.
At usenet electronic group people speculate what was the reason someone probably converted 2 digit year from RTC to 1920âŠ2019 range.
At mathematics circle in secondary school the student was invited to tell us something about programming. He was absolutely wrong in teaching anything. Instead of telling something about algorithms he concentrated at what character have to be at what line position in Fortran.
It is possible if there are many people in team. When we (me and my brother) setup in 1988 our small firm I was designing the schematics and writing PC software (DOS, C++), he was designing PCBs and writing microcontroller code (assembler).
Well I guess that sunny nihilists probably have it over moony nihilists. They are brighter and do not have a dark side! But anyway I guess I have that issue. I am somewhat in awe of people who can excel in one field of study, then change fields and excel in a completely different field of study. âPolymathsâ.
Wow! So more went wrong in 2020 than did for the Y2K bug! (Actually so many people were so worried about Y2K that many of the potential problems were fixed ahead of time.)
I think that if itâll work for you, than itâs a good idea.
Personally, I really favor European nomenclature regarding parts values on schematics, ie:
1âŠ2=1.2 ohm
4u7=4.7 MFD
I just like the looks of the values in the left column on a schematicâŠeasier for me to instantly know the value. Also it takes less room on the silk screen. And with an engineering drawing thatâs been copied several times, itâs more legible IMO after a few copies too.
When I fill out the part fields that end up in the BOM, one of them could be engineering notation but usually spread sheet software can sort the values sufficiently for me to get what I need for the engineering staffâs needs, even without using the engineering notation.
I certainly agree that some simple numerical order is better than none. I had that argument years ago regarding power supply part numbers. The other guy claimed that because a systematic part number could not capture all of the un-anticipated variations, there was no point to making any of it systematic. But I also think that numbers which sort better are probably better than numbers which do not sort as well.
I wonder about a couple of things that you said.
I think that such nomenclature seems very good for marking bags of parts. I can certainly read it easily enough, but I do not know what are the rules or âboundaries.â Do the rules say that it is wrong to type ”033 instead of u033 or 33n or 33n0? All of those variations would sort differently in a spreadsheet.
This surprises me too. I remember (ammonia and other) blue line copying machines, but I would think that these days copying would be 99% replaced by viewing or printing digital media such as .pdf files? (Given that we are discussing new designs rather than reproducing achival material)?
In the 033 case you mention, I would try to be consistent, so 33n would be what Iâd use.
I use Ubuntu and their character map is multi-pages all fairly stuffed with crap. After stumbling around for a while trying to find the u symbol, I gave up and just used u for micro. The ⊠I found quickly and just copied over fine to Ki forum.
Back when I worked in engineering, the tech center was a 20 minute drive away, and needing paper copies of schematics was a daily ritual out there. Greasy smudges were the rule. And I wouldnât know which ones Iâd need in all cases. For security reasons, the tech department did not have a readily accessible suite of schematics or engineering drawings on site. There was active industrial espionage (or at least had been at one time) and we handled military and big business contracts. The entire 750,000 square foot building was surround by a 12 foot high dirt berm to discourage pictures of our equipment builds. Kinda fun.
And that was years ago. They now probably have all the equipment to print out a schematic at will.
That is quite the deal you describe. I think that inconsistent nomenclature is a problem across all of electrical engineering and probably engineering in general. In my last consulting gig I did get some pushback on engineering notation, but I think it worked out OK so long as I had conventional notation in another column. Regarding consistency, even now I corrected âScientific notationâ to âengineering notationâ and I think the difference is important because the latter translates most easily to K ohms and nanofarads. I use sort of an old version of Microsoft Office. Excel does do engineering notation but Microsoft does not call it that. Always something to make matters more difficult.
Simple, the value is between 1 and 1000 so the decimal point is always to the right of at least one significant digit. Itâs a form of engineering notation after all.
I am aware of resistor color code but many of the colors are close enough that my eyes canât determine them accurately⊠so I give up and just read it with a meter as it is faster and simpler.
Sounds like programming! Is u versus ÎŒ any different from the other Greek symbols? I was almost ready to give up finding in the character map. AnywayâŠusing engineering notation is EASY. Yeah I knowâŠI am speaking for myself.
It is true that sometimes the color bands look similar but:
Admittedly I do not use axial resistors so often any more. But I do have a bunch of them and they are good for hand wiring which is sometimes necessary. I have rolled my eyes when I was in some otherwise well equipped labs which had no axial resistors. Technicians needed to tack wires onto 0805 chips in order to make a hand-wired circuit mod. That is much more delicate and time consuming then using an axial resistor.
If you often find difficulty in reading the color codes; perhaps you need better light? Seriously⊠There are many life situations where I am happy to work in dim light. But finding things (in general) and reading colors both greatly benefit from having adequate light. My lab bench has an overhead shop light with two LED tubes.
Is u versus ÎŒ any different from the other Greek symbols? I was almost ready to give up finding in the character map. AnywayâŠusing engineering notation is EASY. Yeah I knowâŠI am speaking for myself.