#Need a Professional to take a quick look at design
1 messages · Page 3 of 1
the safety ones
and a lot of places publish free snippets from IPC standards. Sierra Circuit's DFM guide (sign up with your email and you get it free) has a bunch of IPC info in it
safety ones will really depend on what you're building
im 100% gonna need to know the safety stuff
over here building stimulator circuits lol
at a company, they'll have a compliance engineer whose job it is to know that
again, holistically, good to know about it, but not always required
only really small companies don't have compliance engineers.
(others might chime in, but every company I've worked at with > 100 employees had a compliance engineer. never been anywhere smaller than 140)
i think i wil learn just from reading some ipc stuff
im such a rookie ill probably learn something significant
to build your own company, yes you would. To work for a bigger company, you need to be aware of them, and knowing them more intimately is a bonus, but I'd focus on good design and theory fundamentals well before that
Get the Sierra Circuits DFM guide
and read that altium blog post I sent
ok i will thank you daniel
those are great primers
Are you made of money? Safety certs are like years salary kind of cost
Medical startups need a crapton of funding, unfortunately
yeah I didn't even mention this; those standards are big bucks
there are engineering firms that specialize in helping you design for cert (or doing design for you) for those kinds of products, since the cost is so high
again, this plus other reasons is why I wouldn't worry about those certs - a big company will have those certs and it's not your job to know them inside and out
that's why the "compliance engineer" role exists
Im just gonna learn about the design rules they use for certain types of circuits 🤣 🤣 🤣
This is madness
I would have also expected standards to be free, given that they are supposed to be, well, standards.
At least on the front of scientific papers we're very slowly moving into an open access direction, e.g. my university since a couple years enforces that all published material be available open access as well.
I agree would be nice. There’s no fundamental reason they have to charge for them, but that’s the model most of the industry operates on: IEC, IEEE, IPC, JEDEC, MIPI are the ones I’ve dealt with and they all charge $
Yeah standards cost an absurd amount to produce because theyre often designed by committee
A 1 hr meeting with all the stakeholders can be ten thousand dollars when you do the math
This is simply an uninformed opinion and it is wrong.
Standards provide valuable benefits to all. But they not only require significant investment to develop, but also to administer, maintain and service.
It is the later part you subsidize when you purchase a standard.
The PCBs we all develop can be produced inexpensively and with high quality because of these standards. You benefit from them for "free" even if you don't own a copy of them.
I’m on your side here 😉
By “no fundamental reason” I mean that orgs could donate their time. I previously stated that the current model requires investment.
Wish you would have read my earlier comments
@stable leaf
@stable leaf
@stable leaf here
I wish you wouldn’t assume I’m uninformed in this area
@stable leaf
I tempered my responses with “yes I agree it would be nice if everyone made standards free”, and my implication is you could structure the industry ground up around free information for all, but that’s not how it works. I think if you had read the full context you would have seen that.
I’ll grant that the wording “no fundamental reason…” is misleading. What I meant to say is “a model could probably function where information could be freely exchanged, but that goes against the grain of how the industry operates”
Anyways, hope that clears things up 👍
Out of curiosity, are you in school? I had it in my head you were working as an EE in Italy, right?
doing my masters in Switzerland haha
not sure who of the regulars is from Italy
Prob saw that you had an espresso machine and assumed 🤡
Which is funny considering I also have one
lmao tbf it is made in Italy 😂
Maybe that’s what I latched on to
ebastler is in italy
Ahhhh
Are we all german in here?
Or of german ethnicity?
Ive been noticing a lot of people in these chats are
not in over 100 years
What?🤣
@dry grove @stable leaf @finite pebble @tawny wasp how often are you guys doing rf related stuff on your pcbs. Like for example, what are the absolutely necessary things to learn?
Thanks for pointing out my oversight. I do appreciate it. As someone who has spent decades working on standards it is a thankless job and the whole discussion stuck a nerve. Sorry for calling you out so bluntly without having the full context.
All good!! I hadn’t had my coffee either, so it struck a nerve back haha
40+ years of experience is great, and glad to have your perspective
Pretty much all the time. Mostly NFC, WLAN, Bluetooth stuff. Some of it is unusual, but mostly routine.
I think I posted before about this...
Baseline is an understanding of impedance, phase, inductance vs. capacitance, parasitics, ideal models (spice), and Smith charts.
The rest is experience and applied knowledge.
That's my 2 cents.
You know what this feels like, its like I just started a video game and I wondered into an area of the map where everyone is level 99
Hah... You're in Area 51. Just us aliens here 👽
Exactly
Im just trying to make sure im spending my time on the right aspects
I havent gotten into phase yet
I understand parasitic mostly inductance capacitance and model. The smith chart looks terrifying
I used to think so... Then it just clicked when I had to impedance match an antenna.
Rarely
I don't touch RF if its avoidable
I was playing around in flastad trying to wrap my head around impedance matching. Still dont have a deep mental model of it
I kinda understand it but not to the extent that i could create it from scrath
Scratch
It's maximum power transfer theory
Thevenin's theorem but with infinitesimal slices.
Look into w0qe.com and w0qe channel on YouTube. In particular, learn to play with SimSmith.
It can help develop your intuition.
alright now let me ask you this... If I study all these things consistently everyday for years and really put an effort into it, do you think its possible that one day I could physically become a PCB
i think thats a better question for @zenith bone 🤣
Nothing is impossible and besides you are what you eat 😋
ich werde werden PCB
2 times werden behind each other ouch
i dont even speak german lmaooo
Ich werde zu einem PCB
Is what you would say
Ich werde eine Leiterplatte gewesen sein :3
@finite pebble is the drawing sheet in kicad where you did the drawing stuff?
or should i just draw in the pcb editor and have it with my gerbers
Why gewesen sein? You were a PCB once but now you're not anymore? Sadge
I will no longer be a PCB in the future
lmaooo
Yeah but why
I wasnt born, i was manufactured
Isn't being a PCB great ?
I just like the grammatical construct of futur II
werde gewesen sein
Yeah future 2 my beloved funny thing
the bread cutting competition will be held at 6pm
so if my components dont have that dot to indicate pin 1 they wont assemble it correctly?
are you ordering with JLC? They have a nice 3d viewer where you can set your component orientation, and you can also check the component orientation they will use while the PCB is being manufactured, as they do the data preparation for assembly as soon as you order, but assembly needs to wait for the PCB to be manufactured first
i was gonna go with pcbway becuase i heard jlc messes up boards
ive had 0 issues with JLC
actually thats not true
they have put a couple resistors across completely wrong pads because they were close enough together, but i feel like they probably tombstoned and whoever fixed it didnt realize they were supposed to go the other way
complete side note, but good god I do not understand the existence of trinamic
motor go brr
thats the only time ive had anything messed up from them
they also put a diode backwards somewhere on either that same board or another
wanst the whole batch, just that one board with those resistors placed wrong
just seems like their entire offering is underwhelming, lacking in features and really overpriced, but the brand is really hyped for some reason
Dont even get me started
unfortunately many designs using the TMC5160 dont bother to feed them 12V and instead feed them the motor voltage of 48-60V frying the internal LDO causing a tiny pinhole to appear through the plastic package and killing the driver
on JLC at the bottom where it says "remarks"
i thought it was done through like a csv or pdf program
what are all those ipc things and what should I be exporting for manufacturer?
not for notes afaik w JLC
you can upload the component placement information like this in a pdf
zip it with your gerbers and mention it in the text notes
I just use the JLC fabrication toolkit plugin for exporting the gerbers and pnp files
id do that and then add the placement info pdf to the gerber zip is creates
Recommended settings for General PCB design?
depends on who's manufacturing. Also you usually set this before routing the board 😅
you know im nothing but a capacitor
if you're going with PCBway check their capabilities
minimum trac of .15mm
is what I saw was ok
i guess what im really asking is what are the best to avoid higher pcb prices
jlc defaults are .1/.1 (width and clearance), 0.3/0.45 vias iirc, and 0.2 copper to edge clearance
thank you
aisler does .2/.2 and whatever other values by default
im using pcbway calculator to see what makes it more expensive
and .09/.09 on multilayer boards. they do however manage to manufacture even down to .08/.08, but tend to enlarge vias to .3/.56
can do .3/.4 vias
would someone mind taking a quick look at those real quick and tell me if anything is blatently messed up
@finite pebble @stable leaf @tawny wasp @dry grove This is my Final file Im going to send to manufacturer. Is this good enough, is there anything missing or wrong?
is what they claim. Then they just enlarge it without asking
¯_(ツ)_/¯
i use .3/.4 for all my designs
never had any issues
have you looked at the manufacturing files from JLC?
i actually have not
looks like they enlarged the annular to .5
er to .45
hole is still .3
lucky you, I had my vias at .3/.45 and they enlarged them to .3/.56 lol, in most cases reducing the clearance to .08mm lol
all my vias are .00000000000000001mm. They collapse into blackholes the moment they're manufactured
What do you guys think about getting someone to help you with a design and learning from them vs trying to figure it out yourself? Is there a healthy balance or should I lean heavily towards getting people to teach me in my early stages of becoming an engineer?
Both important, you can learn new techniques from people and create and make decisions and trade offs in your projects.
You can learn on your own from data sheets, app notes and youtube
If I have a choice I like to choose things that are fun and interesting
What's different about being an engineering student vs. being an engineer? A student needs to learn how to solve a problem. An engineer has to use both knowledge and experience to define the problem.
It's the later part that you learn by working alongside engineers. That's the mentoring that is most valuable.
I wouldn't be in a rush to skip over what professors will teach you. That is the foundation for the rest of your career.
You guys know of any good swithcing mosfets for boost converters? like ones that are commonly used? My switching will be around 2-3A and the output voltage will be 150V max, 100khz.
Digikey is a good place to start. Infineon are quite good, but there are lots of brands to choose from
I don't think there's really a "commonly used" part
There are so many different PNs to choose from because they all have different specs necessary for different applications
Only one off the top of my head is like a 2N7000
But that's not HV nor gonna be fine with 3A
Do you think that if I look at the MOSFET and concentrate really hard, I might be able to change the gate charge with my mind?
if you charge it up it will slowly leak electrons into the surrounding air, does that count?
I was thinking of affecting the collapse of the electrons wave function when they go into super position with my consciousness but ill settle for the slow leak
I will steal your electrons
How dare you