« first day (2314 days earlier)      last day (2612 days later) » 
04:00 - 21:0021:00 - 00:00

4:34 AM
1
Q: Controlling an MCU with remote fast real time bidirectional communication

codinginsaneI'm trying to make a robot that can send real-time data to a computer with a decent processing power, so the computer can process the data and make decisions to autonomously drive the robot. Some background information: I have 6 DC motors (Tank like chassis 6 Wheels). Motor specs are 365rpm 29...

 
4:49 AM
Can someone please help me with this problem:electronics.stackexchange.com/questions/289096/…
0
Q: Is it okay to take \$\beta=I_c/I_b=98/2\$ in this problem ?

2017 In an npn transistor the collector current is 20mA. If 98 percent of the electrons injected into base region reach collector the base current in mA is nearly ? Here, in this problem is it okay to take \$\beta=I_c/I_b=98/2\$ ?

 
 
3 hours later…
7:52 AM
needlessly unapologetic
 
 
3 hours later…
10:56 AM
@PlasmaHH Why you can never work from home.
https://t.co/LJflckUBCE
 
 
4 hours later…
3:17 PM
 
3:33 PM
@PlasmaHH Der Empfänger war beim 1. Zustellversuch nicht anwesend. Ein 2. Versuch erfolgt.
 
Curious how I can narrow this question?
3
A: Compact DC/AC converter

Jack CreaseyEmbedding a Walwart inside your project box seems like a retrograde step. More sensible (and already suggested) is to use power supply products designed to go into project enclosures. If you want to go in this direction you take on a whole lot of responsibility if you intend to sell product. If ...

The short version is "Would this component work as a compact power supply?" Not sure how much narrower I can make it.
As usual, voting to close without providing a comment with a reason is stupid and useless.
 
3:53 PM
@Asmyldof I blame the wife, her responsibility to stay home ^^
 
@PlasmaHH o.O
I'm sure leaving a Daughter in charge will be fine for a 29kg package
@AndyD273 You ask whether something would work for you, by providing none of the details needed to judge whether it would. As some answers indicate most people suppose the device you indicate won't work, because you are probably not looking to build your own SMPS with a controller module, which that is. But how the hell can we judge and suggest if we know none of the requirements other than "Smaller than an unnamed and unreferenced wall adapter"
 
@Asmyldof That's fair (and also why a comment to that effect would have been useful so I could fix it)
And also why I'm here asking how to fix it
 
JRE
@Asmyldof Paid.
 
4:11 PM
@JRE Yay. Your package needs to travel further, though both went through the same hub at the same time in Germany, first delivery tomorrow
@AndyD273 I'm not saying everyone else in that Q-chain acted in the spirit of the site, but generally EE is difficult enough when you know the specs, such that a "management type" request quickly falls on deaf or dumb ears here
 
@Asmyldof I'm still learning my way around, and also what different components do. I added some details. I don't know if that makes it any better
like for instance, this: digikey.com/product-detail/en/VSK-S15-5U-T/102-2608-ND looks like one of these: digikey.com/product-detail/en/cui-inc/VSK-S1-5U/102-2595-ND/… with some screw terminals attached. Does it actually need a lot more stuff to make it into a functional AC/DC converter?
 
4:46 PM
@AndyD273 Do you want a thing that you solder on to your pcb, or do you want a thing that hooks up with wires to your pcb?
@AndyD273 Do you know how much current your load will need?
 
on board would be fine with me.
 
@AndyD273 The first unit you linked says "15 W" and the second one says "1 W", so I don't think they're the same.
@AndyD273 That said I'm using something like the first one (but 50 W) on my current project and it seems okay.
 
Around 5V. One of the applications is a relay, which has a 5V coil. The other part will be something like an ESP8266
 
@AndyD273 How much current?
 
The wallwart I'm currently using is 220mA
@ThePhoton Yeah, I saw that. It was mostly an example of what I'm hoping for. Something compact
 
4:50 PM
@AndyD273 The only think about soldering on to your pcb is you will need to understand creepage and clearance distances (you can google it) to make the pcb design safe. Also choose a connector to bring the mains voltage on to the pcb.
 
We should advance from using plain old electricity to something more advanced that doesn't throw away the majority in conversion losses...
 
Maybe I don't know enough yet to handle this project yet. :/
 
@AndyD273 5 V @ 200 mA is 1 W, so even the smaller one "should" be okay.
@PlasmaHH donkeys
 
photons...
 
@ThePhoton The wasted energy is pretty high on those
 
4:53 PM
@AndyD273 If you could just train another donkey to muck out the stable, though.
@AndyD273 Or, if the intended application is "garbage disposal", get a goat instead of a donkey.
 
@ThePhoton Right. I'm looking at the data sheet to see what I need. Test configuration looks like a couple caps for current smoothing, and the "typical application" has a lot more...
All this makes me wish I could take some DC transformers plans back so that Edison could have won the power wars instead of Tesla. To be fair, I would give Tesla the patents.
 
@AndyD273 Take them back from where?
 
@ThePhoton Time travel
IIRC, during that time period we had AC transformers, but not DC transformers, so Edisons idea of running DC power everywhere wouldn't work because either the current would be way to high, or you'd need a power station on every block. AC worked because they could run it at high current, and reduce it at substations for household use, which is what Tesla advocated.
As is, we run AC everywhere, and then mostly convert it to DC power at a loss
You know, I have an LED nightlight that somehow manages to turn AC power into DC to run an LED and fits it all into a very small package. I really aught to take that apart and see what kind of current it has...
 
@AndyD273 Problem is, we still don't have DC transformers.
So you're going to need to go to the future first to get the plans, and then back to the past to get them to Edison.
@AndyD273 Look around for "off-line" power supply designs. The trade-off is no safety isolation.
 
5:11 PM
Huh. Interesting. For some reason I though we had ways to step down current from very high to low. my bad.
k
 
@AndyD273 We have, but they're not (AFAIK) called transformers. And Edison would have a hard time doing anything with the plans, since there was no purified silicon (along with a lot of other things) available to him to make the ICs.
A high-voltage, direct current (HVDC) electric power transmission system (also called a power super highway or an electrical super highway) uses direct current for the bulk transmission of electrical power, in contrast with the more common alternating current (AC) systems. For long-distance transmission, HVDC systems may be less expensive and suffer lower electrical losses. For underwater power cables, HVDC avoids the heavy currents required to charge and discharge the cable capacitance each cycle. For shorter distances, the higher cost of DC conversion equipment compared to an AC system may still...
 
Not to mention the fact that we then would all be stuck with a single phase DC distribution system, which would suck for losses
 
JRE
5:41 PM
@AndyD273 That's "high voltage.". You generate AC at high voltage, so that you have a lower current for the needed power. Lower current means lower losses over long distances. You transport the AC at high voltage until it gets close to where it is needed, then use a transformet to bring it down to a more easily handled voltage. Now you need more current for the same power, but the losses are acceptable because the distance is shorter
 
@JRE Right, I was mostly wishing for a way to do high voltage DC, so I could avoid all this AC/DC conversion stuff... I get that I don't know enough to really carry on any intelligent conversation in here yet.
from The Photon's link it looks like its at least possible, and may have some advantages
 
Given a pad drawing like this one, what are the general rules to produce a footprint. In other words, how much slack should I add in each dimension?
 
@DanielSank You should be able to find good recommendations for pad layout on the datasheet for a particular chip.
 
@ThePhoton Oddly enough, I don't see it anywhere on the data sheet.
I figured AD might have separate documentation for their recommended footprints, since they generally try to not repeat their documentation (which is nice), but I haven't found that either.
The closest thing I found was their library program, but that will be a rather large PITA for me to use because of how security works at my job.
 
5:56 PM
@DanielSank You could look to other vendors' chips with similar packages.
Or ask your assembly shop for a recommendation.
 
@ThePhoton ..and extract rules of thumb.
@ThePhoton Oh that's a good idea.
 
The assembly shop will (if it's a full-service shop) have access to the appropriate IPC specs, so be able to give a good recommendation.
Or you could get the IPC specs and try to work it out yourself.
But it's pretty messy.
@DanielSank If you have Altium, there's a footprint wizard that should be based on IPC also.
 
@ThePhoton Oh?
Oh, yes, looky there. A footprint wizard!
 
@DanielSank You'll need to double-check that the results meet your fab's capabilities. For example, if it generates too thin a solder mask sliver between pads, you might have to change the design to use gang relief (with increased risk of shorts between pads).
Or narrow the pads slightly to increase the sliver.
Or whatever
 
Just my luck that my part's package type isn't in the wizard.
 
6:05 PM
:35717337
@DanielSank Try the QFN wizard.
 
@ThePhoton I just came to that same conclusion ;-)
Thanks!
 
@DanielSank Just ran through it and looks like it came up with an okay footprint. You might have to reduce the heel distance or the solder mask clearance to avoid the solder mask openings for pads 1 & 40 (for example) overlapping.
 
@ThePhoton ...working...
 
If you're working with a good assembly shop, you might want to send the results to them for comment. (mark up the main dimensions for them, and also send the package drawing from Analog)
And if you work someplace where "how security works here" is a hassle, hopefully the trade-off is they are willing to spend on using manufacturing partners that give more support.
 
@DanielSank All of the required dimensions are in that drawing
 
6:16 PM
@ThePhoton For sure. I can spend a bit of money for good assembly. My colleague recently paid a bit for blind vias and nobody batted an eye about the cost.
@Asmyldof Yes, indeed, but that doesn't mean that I know the relation between the pad drawing and the recommended footprint.
 
@DanielSank Become an IPC member and check out all the docs?
 
@Asmyldof Yes, that would be a solution. My colleague reports having done that and recommended I not do it as he found the documentation difficult to navigate. Another solution is to be lazy and hope someone here could summarize the important bits for me.
ThePhoton has already helped by pointing out the IPC compliant wizard in Altium. At least now I can look at what it produces and compare to the pad drawing, thus learning something at least about what Altium thinks it should do.
 
@ThePhoton unfortunately the component rule check doesn't support (AFAIK) checking copper clearance yet. Which is a PITA. Really loved those checks
@DanielSank The drawing you linked yourself: analog.com/media/en/package-pcb-resources/package/pkg_pdf/… is all you need to enter the right numbers to get the IPC wizzard to make something
As mentioned, just need to check the corner pads for your clearance rules
 
@Asmyldof I know that.
I did not know that wizard existed until ThePhoton told me.
 
@Asmyldof But you also need to know how much heel and toe allowance to make your assembly shop happy.
Although mostly the IPC reccomended values should work
If you go to the actual IPC specs, you'd probably have to do some calculation dependent on the stencil thickness.
 
6:26 PM
@ThePhoton Most fabs are fine with medium density as a standard
Low density has never resulted in any problems, high density has here and there
Finding that out was a PITA in itself with one of the US based fabs
 
@Asmyldof fab or assembly?
 
Both
The people we deal with assembly, but they have close friends do the fab
Or something
"It didn't reflow right. Some pins aren't connected. Can you look at the solder mask openings and comment please?"
-> "Solder mask openings looks fine. Probably wrong pin spacing" <-- !?!?!?!
"Are you sure? Because the pins are spaced fine, but it seems there's too little solder, or otherwise the reflow profile was wrong. It does seem like maybe the chip lifted a little"
-> "No reflow was right. Can't have been a problem"
... etc etc etc (2 days)
-> "We think maybe the solder mask opening is wrong"
Soldermask opening --> Paste mask
I'm tired
 
@ThePhoton I see what you mean about the heel now.
 
6:42 PM
Is it still a "green wire" if one lone trace is cut?
 
@MickLH Then it's a cut, not a jump
 
I'm not sure what the appropriate informal derogatory term is for these boards. They require a cut to become functional.
 
@MickLH Some dissembly required
 
@MickLH "bad" ?
:-P
 
"DIY"
 
6:48 PM
I was hoping for something that fits into "... piece of shit", like "green wired piece of shit" really has a ring to it
 
green wired?
Bodged?
Bodgey
 
GWPOS sounds pretty reasonable.
How about "Half assed"?
 
too accurate
 
r
Using ass to say "half arsed" is in and of itself half arsed
 
6:50 PM
But that expletive value... ok how about just "fucked up"
 
^ That works but lacks originality.
 
I have now ventured into the world of git-branching an Altium design set and it's hilarious the way Altium tries to keep up with my branch switches
But, it does work properly
 
oh dear...
I've seen a colleague branch-switch while a python interactive shell is open, reload various modules, and carry on with his life.
 
Result of people wanting 60% of the design to be the same, just the modules plugged in are a different size
 
It's kindof amazing that Git works at all, that text difference voodoo feels sketchy to me
 
6:52 PM
@MickLH diffs have been around for a long time.
 
They feel sketchy to me.
 
It's one of the most space and time efficient systems to keep track of mixed folders
 
I've had bad experiences with projects where two people inadvertently re-write some support code in their patches at the same time, causing spontaneous Git fuckage
 
Sometimes I think it's just Linus going back to his first idea of what a filesystem should do and implementing it on top of all of the filesystems without us noticing
@MickLH And have you ever tried to merge ANYTHING with SVN in the same branch, at all?
At least with Git I can have it all fixed in less than a minute
 
No I refuse to actually use SVN for source control
or CVS
 
6:54 PM
git is awesome but it does have some interesting problems.
 
But I now have a CFP and a CFP2 branch of the same test board and it works magnificently
 
Ever try submodules?
 
I haven't formed an opinion on Perforce because I haven't had any clients force it on me yet
 
@DanielSank YEs
 
@Asmyldof Yeah so, they work, but there are some subtle ways you can shoot yourself in the foot.
 
6:55 PM
Name one
 
Also, several of the default configuration settings for submodules are just ridiculous.
@Asmyldof You are really argumentative and confrontational, eh?
Sure, I'll name one.
Gimme a sec to reconstruct it...
 
For most people, likely yes, but you can set those up
 
@Asmyldof Yes you can. I didn't say you can't. I said the defaults are bad.
 
Yes, and you can change those
 
@DanielSank Pffff that's not even really argumentative and confrontational yet!
MY BOY ASMYLDOF BRINGS DOWN THAT HAMMER SON GET BACK
 
6:57 PM
@DanielSank I am asking, because I haven't had any problems that weren't my own fault
 
@Asmyldof Ah well, with a computer you can always blame everything on the user.
 
@DanielSank true (program exited successfully)
 
The interesting questions is whether or not the software could have been improved such that the mistake in question wouldn't have been made.
It's not interesting to take every mistake that happens when using software and say it's the user's fault.
 
@DanielSank I'll conjecture "false" by the "you can't fix stupid" theorem
 
@MickLH He said "the mistake", not "all mistakes"
 
6:59 PM
(ok ok secretly, my intention is to produce a counter-example because I'd really like stupid to be fixable)
 
@DanielSank When I try to do something with Git with the wrong tool in the toolbox and everything blows up I don't go around the office shouting Git broke my stuff, is what I'm saying. Many people, hopefully not you, but one doesn't know yet, on the internet do
 
@Asmyldof Of course I don't go around shouting. That's never helpful.
I don't understand how you got from my statement that submodules contain subtle traps to talking about shouting around the office. It's almost like you're not actually discussing this issue in good faith.
 
So stuff sometimes goes wrong here as well, but never has it taken more then a minute to realise that "of course this shouldn't have been a cherry-pick", or some such
 
In any case, if you're actually interested, go here and search for "first-order trap" and read the surrounding discussion.
It explains how you can accidentally nuke commits from a submodule if you aren't really paying attention.
@Asmyldof Good for you.
 
I straight up scream at the top of my lungs and rabidly cause violent chaos until I can be sure there's at least 1m property damage
 
7:02 PM
^ nice
 
@DanielSank You turn genuine interest into confrontationalism and then a genuine claim of experience to date into "being able to turn everything into a problem with the user". Why accuse me here?
 
I have one project with a very large submodule that I often need to change with the project, but that I need to not automatically update to new versions of the submodule unless they are verified by hand first
It's... a tedious workflow on that
 
@Asmyldof Cease fire? Check out the link if you're interested.
 
Guys piss or get off the pot, this isn't really aggressive enough of confrontation to be entertaining, give me a more intellectual flavor of entertainment
3
 
@MickLH Hmmmm, ok... uh... all experiments demonstrating a violation of Bell's inequality are fundamentally flawed because mutter mutter.
 
7:08 PM
@DanielSank That could be called a trap. It's never hurt me because I have an allergy for asterisks, angle brackets and all other stuff that indicates my flow isn't fixed yet
But I can see how it can trap people
On the other hand, the choice not to automatically check out after fetch had.... some reason....
Forgot it, probably not the best one at the date of 2017
 
@Asmyldof Yes this is what I said before: a lot of the defaults for submodules are sub-optimal.
 
You are free to add something like:
git config --global alias.pullall '!f(){ git pull "$@" && git submodule update --init --recursive; }; f'
 
Yes, of course.
An expert user can always figure out how to do the right thing.
 
As I said way back, I agree some defaults are fuckky, but everything can be modified
 
And as I said way back, the interesting question is how the software can be improved so that users don't make mistakes.
 
7:12 PM
The fact it can be modified on all levels is one of the reasons I like Git
 
Indeed.
 
Way back when with SVN we needed to hack the server software to get behavioural change
 
Does SVN still require a network connection to do work?
 
Once I remember what the reason was for not auto-checkout I'll let you know
No
But then you have to have your server software on the same computer
:-P
 
"...Way back when with SVN..." apparently, asmyldof is a rapper now
Guys what ASIC tech is good for cheap these days?
 
7:15 PM
I've been a Poet Laureate without the official appointment for years
@MickLH Hung Chow Famous Fujian Technology Co Ltd
 
@MickLH For runs of less than 100k units?
 
Yeah :/ I doubt there's a huge market
 
@MickLH I've heard that Jazz Semi is "cheap" if you need 10 Gb/s.
Meaning cheaper than other processes that can do 10 Gb/s
 
Thanks for the lead, will have to check them out
 
7:33 PM
o.O Jazz has some history for a company started after I started university
 
To un-route all or not to un-route all, that is the question.
 
?
You know about s-p right?
 
@Asmyldof s-p what?
Select connected copper?
 
8:14 PM
@DanielSank I would hope that at work there is a network connection...
 
@ThePhoton Yes. s-p, s-u and s-i help me modify existing design quickly and efficiently (select connected copper, select touching area of rectangle, select inside (normal select, but forced))
designs**
 
8:28 PM
@PlasmaHH Yes but then one gets on a plane/train/etc and wishes to do work...
Man, what is it with the Captain Obvious snarkified comments around here?
^ Of which I suppose that is one.
 
@Asmyldof Handy command, but not much use on ground connections
 
@DanielSank not if I don't get paid for it. and even then I can wait until I check in, should that need ever arise
 
@PlasmaHH Ok but you're not everybody.
 
Your mom's not everybody!
 
^ True!
Not sure. The physics room isn't like this.
 
8:31 PM
@DanielSank everybody that uses svn that I care about in that regard :P
@DanielSank thats because everyone there is afraid of thermonuclear retaliation
 
@PlasmaHH So, if you're on a six hour flight, you don't mind not being able to commit? When using git I like to make small, dedicated commits so that the reviewer doesn't have to comprehend a bazillion changes all at once.
 
(whereas here we embrace thermonuclear retaliation)
 
@PlasmaHH No no no you've got it all wrong. We're into black holes.
 
@DanielSank if you are, how can we communicate, or aare we too?
 
...and I might want to work on two different things, i.e. on two different branches.
@PlasmaHH I do not understand.
 
8:34 PM
Have you studied the art of recklessly code bombing the maintainers? Shit's great.
 
@MickLH If by "studied" you mean "looked in horror as a maintainer", then yes.
 
@DanielSank you went into a black hole... afaik its impossible to communicate with the outside then
 
@PlasmaHH No no no, we build intergalactic black hole launchers. We don't personally go into the holes ourselves.
 
the best two features of svn that git does not have is externals and no pull requests
 
@PlasmaHH What is wrong with pull requests?
 
8:35 PM
@DanielSank they are very annoying
 
Also, pull requests are a github thing, not a git thing AFAIK.
@PlasmaHH Oh? Do tell.
 
Code bombing man... I'm telling you, all the ego appeal of doing the project completely yourself, with the added bonus of getting to blame the maintainer for stalling progress and question his competence
 
@DanielSank do you have family? or do you have time to review and modify multiple pull requests a day?
 
@PlasmaHH Why is that an either/or? In my group's code workflow, review is a required step.
Why would you assume outright that I either have a family or review code? That makes no sense.
 
@DanielSank well, get some kids and then we talk again about obvious things...
 
8:38 PM
@PlasmaHH Uh, so... do you think that people with children and jobs at software companies do not review code? I'm really confused.
Is it normal for people in this room to go after each other like this?
 
@DanielSank I am talking about pull requests, you know that github stuff for your spare time free software projects that you have to squeeze into the time bteween work and your family
@DanielSank yes it is because we are grown up people and enjoy an argument to exchange opinions about things nobody else dares to talk about
 
@PlasmaHH Uhhh, pull requests are used for things other than my free time projects...
 
@DanielSank so you do open source for a living?
 
holy!! woah woah you need to calm the fuck down daaaaamn
 
@PlasmaHH Right, because "get some kids and then we can talk again about obvious things" is mature conversation. Keep telling yourself that.
 
8:41 PM
RIOT!
 
@PlasmaHH Nah, work in superconducting qubit research (experimental). We write a lot of code to control the hardware.
 
@MickLH I tried it once and didnt get it to work
 
This has been a surprisingly unpleasant conversation. Goodbye.
 
@DanielSank and why do you need pull requests there?
 
jokes aside, @DanielSank, I think you're having a fundamental seriousness mismatch
 
8:43 PM
@MickLH I have a low tolerance for wasting words for no reason other than to stroke one's ego.
It's just boring.
Boring is unpleasant.
 
I think the main things available on SE chat are entertainment, and hope for humanity
 
@MickLH I don't think there is much left when people want to ban words like "victim" and replace it by some terms like "experiencing person"
 
@PlasmaHH Oh that seems worse than it really is, people have always wanted retarded and/or horrible things. People just seem to have regressed in their ability to quiet down, suck it up, and take it like a man.
 
@MickLH which unfortunately leads to these things being implemented because people lose the energy to fight it, because if you do, you are a regressive bad person
 
@ThePhoton Shelve polygons first?
And also at everyone: Pull requests are an every type of versioning in a larger group thing and should be managed by dedicated time
 
8:50 PM
@PlasmaHH Eh, I mostly just don't believe in people's crap and don't care about their opinions, even when they start acting crazy with the insults and the "close your mind onto our specific version of being a liberal" shit.
I'll be honest, I'm pretty rude about it too because if someone wants to take up a negative view of me I'll give them an actual fucking good reason to.
 
Hi
Trump, right?
 
where? * ducks *
 
@AccidentalFourierTransform nah, trannies
 
accidental? there is a story behind that screaming to be let go
 
8:54 PM
that's nice too
I chose that username because that is what my momma used to call me
 
must resist obvious follow up ...
 
@PlasmaHH I literally find this hilarious, but partly because I'm confident that such issues will resolve themself as people realize it's just that the transgender drugs cause mental instability
 
with or with the fourier transform?
 
lol
 
8:56 PM
* bada tish *
 
@MickLH the condition itself is surely bad enough for the people to get pretty unstable, I would not want to be in their skin (and they either)
@AccidentalFourierTransform or as these days you would probably hear: tää tääää, tää tääää, tää täää
 
I mean I might be wrong, but I've observed an extremely specific flavor of psychosis that is specific only to them, so I've got a hunch
 
@MickLH hm, how should I put it, I think this is a flavour of pyschosis that you find at many places, but it has "trigger parameters" that are unique to everyone
 
04:00 - 21:0021:00 - 00:00

« first day (2314 days earlier)      last day (2612 days later) »