@rawbrawb I don't think it's just "I don't know the answer so it's off topic." The question is also suffering from OP's limited English skills. I had to read it 3 times to figure out what is really being asked. And only the clue of Bobbi Bennet's answer really made the question make sense to me.

The fact he mentions "8 bits" in the title and in the text, when it seems like he really only needs a 1-bit control signal may be throwing people off. Certainly if he really needs to do 8 bits (for some function he never explained) without using a microcontroller and no other justification makes the question "not a real question".

@rawbrawb when i saw that question last night, i just knew the reaction you mention would ensue.

And morning all!

Anyway, upvoted the question, also wrote a comment I may be lambasted for.

@AnindoGhosh I don't think your comment is legit, if you think about OP's request for 8-bit data transfer. Solving his real problem without using a micro seems to be reasonable. Solving an arbitrary 8-bit data transfer problem (with no info about bit rate, but some people might have assumed more than 1 Hz) and ruling out a micro with no justification is silly/"not a real question".

@ThePhoton I read you rewrite, I thinks it appropos. I will grant you the title - with 8 bits mentions - was a non sequitur and misleading. However, there was nothing in the rest that would lead a curious, open minded individual in the wrong direction. I think my analysis is applicable. They have a very limited view of what EE is and are willing to force that view on others.

@ThePhoton I feel like the root of the issue is that he does not understand why you would go to serial for that, he has in his head some misconception that serial is some sort of issue. He also might have a valid reason, but he has not explained why, I think someone should have just explained a way to do it well with serial. Being not a real question is a bit over the top, just because he has misconceptions does not mean we should skip answering his question.

Or a reason we do not realize.

@rawbrawb The original text ended with "Is there any simple circuit that allows me to send such small amount of data, without the use of a micro controller and serial data transfer, all I need is 8 bit parallel data transfer, to send this data." which recapitulates the request for 8-bit data.

Getting some more housework done, off I go.

@ThePhoton fair enough, I just assumed it was a writing problem. And even if it wasn't what's wrong with providing solutions ?

@Kortuk Well, I did not vote to close. But if someone answered with "here's a good way to do it with a micro", and OP commented "I simply cannot use a micro and need 8 bits" with no other explanation, I'd vote to close.

@ThePhoton Yes, the user gives that last bit of rep for an accept but who cares, just upvote the answer that explains why a micro makes it easy if you think that is the best Idea. If the poster does not like it, they will flag what fixed it for them, but you can still have another answer with more upvotes.

I did not mean to single you out, but you had made the last comment and I was responding.

@rawbrawb I agree. If OP returns to say why he needs or 8 bits or realizes he doesn't, I'll suggest to BobbiBennet to turn his/her comment into an answer.

@Kortuk Problem is, OP hasn't stated his question clearly enough to know if a micro is needed or not. If he really just needs an on/off signal, with infrequent switching, he probably doesn't need or want a micro. If he needs 8-bit data at > 1 Hz, he probably does. We could just hi-jack his question to choose one of these assumtions or the other and let people answer it accordingly. But it would be better if OP comes back and clarifies their requirements.

If someone wants to answer about how to do something else from what this OP wants, they should create a new question and ask that, then self-answer. Not just hijack this one to answer whatever they want.

@ThePhoton actually, I think the best way is for someone to take an educated guess and try to add detail and if he wants something different completely when he returns in a weeks then have him post that also. If he returns right away he will see what detail he needs.

@Kortuk I dunno, that's a good way to get a "stranded answer". I've done stuff like that before...try to figure out what OP "really" needs, and answer that. Then OP comes back and edits the question to ask something totally different, and my answer is sitting there making no sense at all in the context of the new question. Waste of time.

@ThePhoton Or we just roll it back, tell them they were not clear and suggest they post a new question. That is the way to have the highest quality output. Out last little review the site had did well!

be back in a bit.

And a question where an answer has a good chance of becoming a waste of time is "not a real question". OP should clarify their question, to the point that people answering know they're not wasting time answering it.

@ThePhoton yes, in a general sense, but here, specifically it can be (if we assume the 8 bit is a red Herring) the only thing. The key point/issue I have is the assumption that some people have that they are the be all and end in terms of knowledge. If one doesn't know PLC (the original PLC ) they should just not answer.

The current edit by The Photon i think works well.

If this OP never clarifies and 3 weeks from now there's still no complete answers and the question is popped up to the front page by Community, then I'll vote to close.

@AnindoGhosh I agree.

@rawbrawb I definitely see people voting to close just because something is outside of their knowledge. But I don't think that's what's going on here.

@ThePhoton well i disagree.

Look at the question linked by Kenny in the comments. Olin answered that one, so presumably he doesn't think power-line communication is outside the realm of the site or outside his own knowledge.

@rawbrawb Well if I vote to close it will be because I think OP's requirements are contradictory, not because I think the question is off-topic for the site.

Also, the votes to close that were already made are for "too localized", not "off topic".

@rawbrawb My solution is to rewire the building so the big loads are on circuits that only get powered by the "government" line. But then it's a DIY question, not EE.

@ThePhoton i will add my voice to rawbrawb's that it is being close voted on the "not my type of EE" sentiment, namely using MCUs being the new holy grail.

Analog still works and solves problems... I am often faced with a "why am i bothering to digitize this" quandary in some types of problems... I still remember the analog ways as being cheaper and sometimes simpler.

Bobbi's solution is a good example.

@AnindoGhosh I think analog will solve the "real" problem. But if the "real" problem turns out to actually require 8-bit paralllel data transfer as originally specified, then rejecting a microcontroller as a potential solution is arbitrary to the point of silliness.

@ThePhoton absolutely, that isn't my issue though.

@rawbrawb It's not your issue, but it could be the reason others voted to close...

I'm saying, with my experience and ignoring contradicory stuff in my favor, I will provide a positive solution w/o assuming the poster is an idiot. I may end up wrong. Who care.

The other answers are focusing, like me, on a aspect on the qquestion that is contradictory. saying that it's useless and throwing the hands in the air.

@ThePhoton can you guess who I fire when I run across them?

@rawbrawb Nothing against you trying to answer. But without more clarity from OP, your answer is only helpful and informative, it isn't "the answer". It's a positive contribution, but it doesn't answer the question (because the question as stated is partially nonsense).

@rawbrawb Guys with too much facial hair?

from a presentation that just crossed my desk.

cool huh?

EE IS more than beards say!

When I was in high school my 2nd cousin forwarded me an IEEE Spectrum issue on MEMS...I think that is what convinced me to pick EE over other stuff I was considering doing in college.

Gotta go...off to dinner.

@rawbrawb that looks like a stepper on my phone screen, what is it?

@AnindoGhosh It's an early step in processing for what most people call MEMs gyro, but in fact is a rate gyro. It's just so pretty I thought I'd put it up there.

- It's not my work.

@rawbrawb could you explain that a little more?

@NickHalden what aspect?

well i only have a basic understanding of MEMS devices

what is a rate gyro?

a gyro measures angle, a rate gyro measures rate of angular change, thus the rate part.

so all the "gyros" people mention as parts of smart phones are actually measuring rate of angular change and integrating to get angle. So it drifts.

ah, i see.

and of course the orginal spinning disk gyro.

after processing to free the rotating structure.

god

NOt sure of scale here, 35 um across?

this is from a Euro company that starts with B....

so whats the idea here

the spinning portion somehow pushes on those little pieces around it

staright bar at bottom of picture is 100 u, so this structure is about 250 um across.

and probably creates some potential energy there

like a voltage

@NickHalden no spinning, => thus the rate aspect.

the spinning thing doesn't move?

it must move...

the last one vibrates, and any angular changes , change the depth that the vibration goes to so the signal is unbalanced during rotational acceration.

in MEMs you can have things that move as long as they are connected. Some people make spinning things, but not for long.... i.e. it breaks. in production anyways.

ok but what about those first two?

those look like fan blades?

Aaah that also vibrates rationally and the fan blades are torsional springs.

just saw, diameter 1 mm

in first one, so the 35 um must ahve been the gap underneath.

the fan blade ones are actually more subtle than I state, need to look at it more closely.

gotta run.

hello all

what is up except the ceiling?

@Rick_2047 heya, i've been designing my portfolio most of the day

Hi

@MLM cool

@AnindoGhosh hi

@AnindoGhosh taking your advice and trying to implement a second order filter, using a lm358

Try the lmV324 instead... might work better for you from what i remember of your req. Cannot check data sheets as i am traveling.

Also there is a maxim filter ic, that might be good... Max275, maybe? It does higher order filters.

strange things are happening, it just halves the PWM signal doesn't filter it

Yes, just checked. MAX274 and 275. Very nice

@AnindoGhosh checking datasheets right nos

if I am willing to add a lot of resistors, it will give me an 8th order low pass filter, interesting

Yes

Since these are continuous time, no digitization aliasing will happen.

but the prices are more than the DAC for this

it would not work

Well, it is always good to know what is available. Now go back to your lmv324

:-)

@AnindoGhosh :)

why insist on lmv324?

any special parameter?

Gets closer to the rails, lower quiescent

Retail cost about the same in India

okey, got it

what about an lm324?

i have an lm324 and lm358 on hand

Check common mode input and output voltages for your vcc of choice. Also check gbw.

I cannot access datasheets on my phone so can't help.

I cannot access datasheets on my phone so can't help.

gbw is same for all the chips, 1Mhz

input common mode is same too

they are pretty much alike (lm358 v/s lmv324)

Ahh check ca3240 please?

wait

ca3240 has higher input swing

No, something is wrong, the lmv cannot be same spec as the lm, specifically for low voltage single sided supply.

wait

lm series specs are given for V+ = 30

you are right, I read it wrong

its supposed to be V+-1.5V, if I give 5V it would give me swing of 3.5v

lmv and ca3240 actually have rail to rail

There you go. For a moment you had me thinking my memory is failing.

:)

but I will just give then a 9v rails, so that problem will not arise

i am attaching a 9v battery to this anyways

Lower rail still a concern

Since you wanted a 0 to x analog output.

yeah, that will be a concern

but it doesn't have to be absolute zero

i mean it can tolerate between 0 to 250 mili or something

and it is 0 to V-1.5v common mode input for lm358

that should work

Ok.

And gbw

its 1MHz for both of them

Hmm and your signal frequency is?

the PWM frequency?

its 16MHz and 8 bit resolution right now, so its 2MHz

Might work since you are doing a low pass anyway. If it dies not suffice, go to the component shops and get some ca3240, they are some Rs 5 to 10 each.

okey, will check that

Oh, 2 mhz. Are you sure + checked on scope?

with a second order RCRC passive filter I am getting a 160mV error

@AnindoGhosh if I am getting a 160mV error with the RCRC passive filter, will I get a better result with an active second order filter?

one more question, we can use an electrolytic capacitor for an active filter right?

Since i have no idea what type of error you are getting, i cannot guess. Electro caps have horrid linearity across voltages and frequencies. Best avoided.

okey then I would just use these polyester caps

Error = standard deviation? rms error?

@AnindoGhosh peak to peak error

Sorry insufficient data to speculate while in a moving car!

let me make an active filter, check its output and then get back to you

should I post a question about rep on ee meta or so meta?

I've posted on ee, but now I'm having second thoughts

@StaceyAnne If EE isn't right, mods can probably migrate it to SO Meta.

good morning @all

hey all

@AnindoGhosh tried the filter circuits, the response was too slow for my applications, rejected

anyone who can help me with labview?

hi, Is anyone here familiar with SPI?

@Gomu what do you need help with?

Serial Peripheral Interface

yes, what in SPI?

what are you trying to interface?

There are three registers SPSR, SPCR & SPDR. Do they reside in both slave & master?

what controller are you using?

are you interfacing a controller an eeprom or something?

No. I'm not doing it practically yet. I'll have to understand its basics and only then I'll be allowed to work on it.

ARM

craneboard or anything in general

only master has those

but those three registers are not for SPI

SPSR is special purpose status register

As per this pdf atmel.in/Images/doc2585.pdf

SPCR is special purpose control register

Serial Peripheral Control Register, Status Register & Data Register.

okey, if that is the full form then only the master will have it

only the master has the control over things

a status register and data register will be on the slave

to tell you that data has been read or written and what is the data that has been read and written

but, its mentioned that the SPIE flag (in SPCR) in slave & master mode

in Page 13

then SPCR must be present in slave also. Am I right?

@JohanLarsson Am having doubts in Serial Peripheral Interface. Can you help me solve it?

/me is out to the office

@StaceyAnne Am having doubts in Serial Peripheral Interface. Can you help me solve it?

@Gomu I'm just a lurker here, pretty clueless when it comes to this type of stuff :D

:)

and also, don't ping everyone, ask your question and wait for an hour or so, I'm sure if someone sees it and can help they will.

Pinging everyone in the room is considered rude by some, this is a chat and not a helpdesk after all

oh... thanks! I'll have that in mind!

@Gomu why not ask your question as a question on the site?

@Rick_2047 for some reason that made no sense to me... An analog filter was too slow?

@AnindoGhosh, how fast is the internet in india?

and how much does it cost?

As fast as one can afford... Mine at home is 2mbps, 2mbps and 1 mbps backup line. Some of my friends, gamers, have 4 or 8 mbps non-shared lines.

is it reliable?

I'm looking at freelancing from india for a while

I pay around us $25 per month for main link, around 20 for backup, and another 20 for the 3G fail over.

@AnindoGhosh Thats quite cheap for your main connection

compared with the UK.

@Dean Yes, if you use exchange rates. Cost of living wise, it might not be so great. A good meal for two at an award winning restaurant would cost me less than that $20.

logging on to 0 flags, dont know if it was fate or clabaccio, but you dont look a gift horse in the mouth.

coffee reward time.

@AnindoGhosh Morning mate!

@Kortuk Just flagged some stuff for you :P

@Dean ><

:-D

@AnindoGhosh You spend a lot for internet!

You would have limited data plans though ?

Please check this candidate for reopening. As it is currently written, the English is still a little dodgy but I think the question is perfectly clear and on topic.

@ThePhoton good on you!

@rawbrawb W5VO did most of the work re-wording the question.

@ThePhoton well then @W5VO "good on you two" ;)

@ThePhoton I'm a little confused in your solution. How are you modeling the Id = Kd * E^(Q^2) relationship?

@ThePhoton Why not just use VerilogA ducks

@rawbrawb Not sure what you mean? The relation I have in mind is I = dQ/dt.

@rawbrawb Wait, now I see what you mean. I think that's a red herring, not part of what OP is really asking about.

@rawbrawb I'm surprised on the minimum W, since at 130nm you can have minimum sized square transistors

Since V = kQ^2, then exp(Q^2/kd) is just exp(V/kd'). So it's just a normal diode behavior (with some slop about the I_s term.

I don't even know what the other symbol is supposed to be, the circle with the horizontal line through it.

@W5VO OK, different processes then, the STI side wall etch and depth of etch mainly dictates that aspect ratio as well as Well depth. Also are you talking dog-bone transistors? because can you even get a contact down on that ?

@rawbrawb it probably is dogbone at that point. It's been a while since I've messed around with the 130nm kit we have

@ThePhoton OK, fair enough. I missed that.

@ThePhoton I think that a fair assumption. Even the way that the subcct is drawn is confusing because it should be two different currents - that's what confused me.

@rawbrawb I think he does seperate them out. One is I_d, the other is just I.

@ThePhoton now to tease you - so it seems that making educated guesses and providing an answer (even if it ends up wrong) is now in your bag of tricks? ;) But only just teasing you ...

@rawbrawb Only if I think the question is clear enough that OP won't come back and totally change it.

@ThePhoton LOL

@rawbrawb In this case, I suspect OP is long-gone anyway.

@ThePhoton still worth while to answer. as long as people search for these things. I +1 for that reason.

This question is somewhat dangerous

As I can't imagine a situation where in the person has access to the tool and NOT have access to the documents unless the tool is a cracked version.

@rawbrawb A lot of places, not all of the users have access to the company's support account.

@rawbrawb Sure, and I see that risk, but often the tutorials supplied with the tools assume you know what you're doing.

@ThePhoton Really? Under Cadence and Synopsys they have different classes of access.

I worked in MentorGraphics PCB tools for years without ever being told how to access any of Mentor's documentation

@rawbrawb I have to access my documentation through an account liason - there's no easy way to see what's available

fundamentally, I think he needs a tutorial on P&R, and then you can start looking at specific tool documentation

@W5VO OK, I've never had that issue, and I've actually been in on the purchase and had that as a requirement , which was fully supported. So It's not clear to me WHY other organizations should allow that.

@W5VO and as @ThePhoton mentioned too, the learning curve and base assumptions can be a total bearcat. So perhaps there are looking for a gentle intor. I will grant you that.

@rawbrawb We're an educational institution paying fractions of the list price. We give up things like dedicated support

@W5VO I grok that, but in reality they place tools into Colleges for indoctrination reasons. ;) I think it's more the "grad students as slave labor" mentality more than the vendor...

@rawbrawb probably

@rawbrawb I might have said that, if I knew anything at all about VLSI place & route.

@rawbrawb On top of that, it is not our business to decide if we answer a question because they might have a hacked tool.

OK so my experience is different, have never used those Synopsys tools so I can't help.

@rawbrawb Not all companies work that way, I have had both results.

@rawbrawb I think Cadence is a bit better with their support for us, but then they still insist on that %#*@( scripting language they came up with

@W5VO ROTFL and so ironically named.

@rawbrawb I mean Synopsys uses Tcl, but at least it *looks like a programming language

@Kortuk Certainly companies make their own decisions, but my statement is not based upon guessing. I have friends who used to be/ are VP level so I have some idea.

@rawbrawb Oh, I dont think you need to have contacts to know that indoctrination is a very effective long term marketing mehtod, but some companies consider that supporting such a product is a bad idea, quite oblivious to how poor incorrect usage can make any product look.

@W5VO Tcl is a little goofy, but it can do some cool stuff too (higher-level/functional programming).

@ThePhoton It just takes a while to wrap your head around the concept of "everything is a string"

but it's not nearly as bad as Skill

@W5VO I avoided it for years, but when I was finally forced to do a project in TCL I really liked it. Of course it's slower than January molasses, but

I hated the idea that () and [] had such different functions, and figured I'd never keep them straight in my head...

@W5VO do you think that Skil being lisp -like is the issue? or is it the Kludged wrappers and libraries that Cadence has surrounded it with? I knew a company (since bought by cadanece ) that could do some amazing stuff in skil. but lisp does hurt my little brain a lot.

Kind of like how I now avoid Python because of significant whitespace, even though Python programmers seem to cope with it okay.

@ThePhoton The same people who don't like Python whitespace think that the Obfuscated C Contest demonstrates best practices. ;)

@rawbrawb The lisp link is definitely a barrier for me, but the implementation doesn't make it better

@Kortuk Thats not what I was saying, I was saying that those vendors wanted to provide documentation support in colleges.

@rawbrawb Hmm, i may be missunderstanding you and we are both saying the same thing, or I am just lost. I am going to call it even and give up. :)

@W5VO And a freind of mine used magic and scheme to make a tool that we used to assemble chips with. so in the right hands ... but not mine.

@Kortuk fair enough. who really knows anyways? :D

@rawbrawb Never me :)

good morning @everyone

@jippie jipster!

@rawbrawb hi there

@jippie you were in Budapest, right?

I deny everything, why?

@jippie I don't know why you deny everything... just wondering if you were still there.

no, i was there only for 3 days

4

kitchen for a moment

@jippie morning

did i miss any major advances in electronics or fundamental physics today?

@AshRj No, my primary and fail-over link are unlimited plans, the Tata Docomo 3G plan is limited. Yes, I spend a fair bit because I need reliable connectivity.

g'night all. Being back home feels good.

@AnindoGhosh Night mate!

bed time for me. Enjoy @everybody

Please go upvote my answer: electronics.stackexchange.com/questions/58386/… and my related question electronics.stackexchange.com/questions/31048/… while I'm asleep so I will be surprised with 5k+ rep when I wake up. :o)

@jippie Nearly there, 'grats

two someones are playing games

I have a screenshot saying 5016 and a screenshot 4996, in that order!

I'll figure you guys out!

@jippie Someone probably got the vote script, but it could be me editing your answers and removing upvotes, you will never know.

@jippie Mods have super downvotes - worth 10 points each.

@jippie If the difference is 15 or 30, that's someone unaccepting your answers. Happened to me a few times.