« first day (890 days earlier)      last day (4117 days later) » 
00:00 - 16:0016:00 - 00:00

4:01 PM
‘Secondary cells’ having an ‘energy density’ exceeding 300 Wh/kg...
@rawbrawb You just have to rename things. So instead of "Chemical Vapor Deposition Oven", you say, "Pizza Oven".
so no LiFePo batteries for my RC cars?
Oh sorry. Note: 3A001.e. does not control batteries, including single-cell batteries.
@DavidKessner another trick you have to watch for it endoscopes, the FDA flags those, so if you're sending those across borders they have to have the right paper work, so you call them "borescopes" and all is fine.
As silly as it all is, the results are very draconian. There is a researcher who had some eqt in his lab, his Chinese grad student was in the same lab space, likely never even got near the eqt. the state dept. threw him in jail. Another Prof idiotically travelled to China with documents in hs laptop. They were never accessed when he was in PRC and when he landed back state side he got into deep doo doo.
gotta run, good "seeing" you guys again.
@rawbrawb A coworker of mine was flying to Wash DC to do some work with the Dept. of Homeland Security. At the time, the U.S. Senate was changing the way they do ID badges. His badge had expired, but the new system was not in place to give him a new badge. The senate was honoring the expired badges during this time....
Anyway, the TSA did a random search of his checked-in luggage and found the blueprints to the Homeland Security office and an expired Senate ID Badge. Many panicked phone calls ensued!
@rawbrawb Take care, and take a holiday sometime :-)
@DavidKessner Something like that deserves panic.
4:14 PM
@AnindoGhosh While he was in the air, his boss got a call from the U.S. Senate. That's not ever a good call to get. But in the end it was all fine. His mistake was that he should have never had the blueprints and the badge in his checked in luggage-- it should not have been outside of his possession.
@DavidKessner Scary stuff.
The US Govt is so weird. They have massive security on some things, and zero security on others. And no real reason for either.
@DavidKessner some of the restrictions are, I surmise, motivated by Big Business.
@DavidKessner The RMS of my career is quite good ;)
4:23 PM
@W5VO Liked that, did you?
@DavidKessner It totally needed an edit... I swear, so many words for "I want to FM a clock using DDS"
@W5VO Yeah, that could be taken so many ways!
@AnindoGhosh Restricted for export doesn't mean banned from all exports. It might just mean banned from export to Syria, Iran, and North Korea.
and Cuba.
@ThePhoton I appreciate that... I was considering worst case scenarios, but I see your point.
@ThePhoton well, there's different levels of export control
4:37 PM
@W5VO Exactly my point.
like ITAR, EAR99
I bought some resistors that had export restrictions on them!
@DavidKessner Regarding previous discussion around FPGAs, please see this - Is it a decent starting point for teaching myself to use FPGAs?
Also, the item I see about 40 MHz MCUs only applies if they're "manufactured from a compound semiconductor", which applies to approximately none of the MCUs I've ever seen.
@W5VO I used to design stuff that I was banned from talking about with one of my coworkers, because that would be considered exporting the technology.
@AnindoGhosh The 2C64A is a 64-Macrocell CPLD, not an FPGA. For what it does, it is a good part. But don't expect to do lots of cool things with it, since there are only 64 flip flops in there.
@ThePhoton Uhmmm wouldn't anything but a bare die be a compound semiconductor?
4:40 PM
@AnindoGhosh Compound semiconductors are like GaAs, InP, ...
@DavidKessner Ok, I'll continue with my eBay scan. Thanks.
Silicon is not a compound semiconductor.
@ThePhoton Ahh ok. Then the general microprocessor section later in the document probably applies.
Neither is Germanium ;)
@AnindoGhosh But for $10, it is useful for a lot of things. I buy just the chip right now for about $1.50 in 1K qty.
4:41 PM
Of course the Bureau of Stupid Rules and Idiodic Restrictions might have their own definitions.
Also, I've had a revelation - when companies brag about power density, that basically means it's going to weigh a ton
@ThePhoton Also, the relevance is low since we source stuff from outside the USA anyway, due to US manufacturer pricing.
a. “Microprocessor microcircuits”,
“microcomputer microcircuits”, and
microcontroller microcircuits having any of the
a.1. A performance speed of 5 GFLOPS or
more and an arithmetic logic unit with an access
width of 32 bit or more;
a.2. A clock frequency rate exceeding 25 MHz;
a.3. More than one data or instruction bus or
serial communication port that provides a direct
external interconnection between parallel
“microprocessor microcircuits” with a transfer
rate of 2.5 Mbyte/s.
@AnindoGhosh If the technology originates in the USA (like an i7 processor), everyone along the way should have signed a contract not to re-export it against US regulations. And it applies to the technology itself, not just the physical goods.
So if you got an i7 legally, it means India isn't on the list of countries where it's not allowed to export that technology, or Intel got a special license to send those parts to Indai, both of which are possible.
@ThePhoton By that logic, would the i7 not violate a bunch of the sections in what I just pasted?
Not if Intel got a license to export i7's to India (as physical goods only, without sharing the knowledge of how to build them, for example)
4:45 PM
@ThePhoton That seems most likely. India isn't on the "send them back to the paleolithic" list, I would hazard.
@ThePhoton Intel makes a lot of noise at seminars about how their design and dev for the latest CPUs is done in Bangalore, India. Interesting quandary.
Look at the top of the section where those requirements are stated. There's a subsection called "reasons for control" with some 2-letter codes.
@AnindoGhosh This one looks reasonable: ebay.com/itm/…
You go look up those 2-letter codes at the very beginning of the document (another chapter not included in your PDF). It will tell you which countries are on the black-list for each of the 2-letter codes.
@AnindoGhosh India comes and goes from the "we don't want them to <strike>have</strike> improve their nuclear weapons" list.
@DavidKessner Outside my fooling-around budget for the moment. How about ebay.com/itm/… ?
@ThePhoton Funny, that.
Actually, @DavidKessner I think I'll stay away from the FPGAs till my current project is delivered. Then I'll have spare time and spare funds. Thanks for looking though.
@AnindoGhosh Might be OK, but I'm not as familiar with that chip. Keep in mind that with any of these things you need a way to program the part (and/or Flash). Some boards might have a USB interface, but others like this one might require something else. Factor that into your budget.
4:52 PM
Is it because I'm on the west coast that my "deliver by 4 pm" FedEx shipments always arrive before 10 am?
good mornign @all
Box just came in from Thief River Falls, so now I gotta go work.
@ThePhoton Today, my "work" involves spending 5-10 minutes setting up a test. And then waiting for an hour as the test runs...
@AnindoGhosh I'm not entirely sure. I thought you needed an external osc. once you set the fuses for an external oscillator, but it is also pretty forgiving when changing frequency between 1MHz..2..8..16MHz RC-osc.
I'm not really prepared right now to test and pull out the HV programmer sketch and related stuff.
@jippie Yes, that was my understanding as well.
4:57 PM
I am really curious though if the answer given is correct, as people may brick their controller if it isn't
I stopped experimenting clock fuses when I bricked a controller by setting it to 128kHz clock. :o)
@jippie generally the clock drivers just need to be set for the right range. like the HS setting in a PIC will handle anything above 1MHz, it just has to have the correct higher current driver on.
HS is short for ... ?
Hola everyone
I'll wait until someone puts an exorbitant bounty on the question ;o)
@jippie I started to answer it but then deleted my answer. I only really know ISP: never used JTAG or HVSP/HVPP
5:03 PM
HV programming is pretty straight forward, if you know the right tricks :o)
have to go feed my cats, they're driving me crazy
@jippie Guess you weren't the one with the cat feeder question :)
i had a cat feeder question once\
@jippie "High Speed"
5:33 PM
@AnindoGhosh @Kortuk if I prove it for an ATtiny45, would that be sufficient for ATmega32?
@jippie if their setup is the same for osc then it is.
they are alike in many things
@jippie I would bet it is the same.
to put it differently: I don't have a mega lying around
@jippie You guys talking about the "Do I need the crystal for programming" question?
5:36 PM
@angelatlarge to my knowledge HV programming is just like LVP but you have access to changing more things, it is not really more complicated. Actually, I consider it less complicated, you can always change everything.
@angelatlarge yes
@jippie I couldn't figure out from glancing at the site/datasheet what method that programmer for actually programming the chips. Do you have one of these?
@angelatlarge I am just rambling on the internet.
@Kortuk HVP is useful for unbricking the chips, I am told. Never done it though.
I have an ATtiny45 running at 2MHz RC and it outputs a 994Hz square wave, a simple program. Next step is to program it to use an external 1.8432MHz oscillator and it should still output little under 1kHz.
then I detach the oscillator and I try to reprogram it
5:38 PM
@angelatlarge I am just saying, it is exactly like LVP, except it can unbrick fuse settings.
so now I'm trying to identify which pin to use for the osc.
@angelatlarge just making sure you realize that HVP does not mean there are dangerous high voltages or anything.
@Kortuk Oh, yes, that, I knew. :) It's 22V or something like that, no?
@angelatlarge yeah, roughly a voltage like that. I have seen a little lower I think.
@angelatlarge I would say, I have a 210A 2V PS, so danger is all relative. I would not lick that one, but I dont really worry about it.
@Kortuk I usually have enough crystals on hand that I've never needed to unbrick. Yet. And I've never set the SPIEN fuse to 0. Yet.
@Kortuk I don't have a good understanding of the relationship between licking discomfort and V/I ratings. Maybe I should post this as a question. Intuitively, I think of V-rating as correlating with licking pain.
5:45 PM
@angelatlarge no, I would not suggest asking.
Not enough people here have licked enough different power supplies to know.
OK I have 994Hz with the external clock
@Kortuk I bet it would get a ton of upvotes :)
@angelatlarge your tongue is a dead short, how much it hurts is a function of how much current it can supply.
so now I remove the clock
@angelatlarge that might be true!
5:46 PM
and I try to program the 2MHz RC
avrdude: Device signature = 0xffffff
avrdude: Yikes! Invalid device signature.
connect the oscillator
avrdude done. Thank you.
@Kortuk One time I tasted a beautiful vintage 400V 150mA adjustable supply. It had a nice pop with a general mouth feel of lightly cooked meat and copper. The nose on the supply was an exotic warm dusty phenolic with hints of overheated carbon composition resistors. I would rate it 97/100
A: ATmega32 programming without external clock?

jippieYes! When you configure the fuses to make the controller run from an external clock, you also need that external clock for reprogramming. I didn't test this with an ATmega, but rather with an ATtiny45, but both AVR's are very similar in these aspects. This is what I did: I wrote a small progr...

6:03 PM
@W5VO No hint of Ozone? I gotta have me some Ozone...
@jippie Is that for that particular programmer, or for any programming method? I knew that you needed the external crystal for ISP programming if that's how your fuses are set, but I wasn't sure about other programming methods.
@angelatlarge good point
@DavidKessner I've tried some recent supplies with 4kV outputs, and I'm just not a fan of strong ozone power supplies. I much prefer the smokey supplies, but that taste isn't for everyone.
@jippie I started writing up an answer, but then realized that my answer was ISP-specific, and I really couldn't figure out how his programmer worked: ISP? JTAG? HVSP?
@jippie So I deleted my answer. For ISP you are entirely right though.
@angelatlarge His programmer is technically able to generate high voltages, it says so in the manual. But what it does for this MCU is another question entirely.
6:07 PM
I added the fact that I tested this with an Arduino Serial Programmer
Let me throw in the Serial part
@jippie Following your lead, I undeleted my answer :) Yours is more complete though.
@jippie I suspect that his programmer can do more than ISP simply based on the fact that there is a ZIF socket. I think JTAG programming would need your chip to be clocked, right?
@angelatlarge At least I figured out something that I was wondering about for a while myself.
@jippie You've never temporarily bricked a chip by accidentally programming the fuses for external oscillator?
@jippie (when you were expecting to use the internal one)
@angelatlarge I bricked one by setting 128kHz internal clock
then I made an Arduino High Voltage programmer to unbrick it :)
@jippie What kind of HV programmer did you make? I think I need one too :)
6:16 PM
Arduino based
Make a question 'how can I HV program an AVR using Arduino' :-p
@jippie Unlike some posters, I do possess Googling powers :)
haven't used the sketch for ages, but it still compiles
@jippie I think I've seen this before: mightyohm.com/blog/2008/09/…
// AVR High-voltage Serial Programmer
// Originally created by Paul Willoughby 03/20/2010
// http://www.rickety.us/2010/03/arduino-avr-high-voltage-serial-programmer/
// Inspired by Jeff Keyzer http://mightyohm.com
// Serial Programming routines from ATtiny25/45/85 datasheet
@jippie But if you need the rep, I'll be happy to post a question :)
6:20 PM
there is no fun if you google'd the answer yourself already
@jippie BTW, if you want to help a person out, maybe invite cpf to chat?
@W5VO PS with hints of peat. Smokey and matches my scotch perfectly.
Q: ArduinoISP fails with Atmega368, 168 - Device signature = 0x000000

cpfSo, apparantly this is a common error, which many people have fixed many different ways, none of which have worked for me. I have two different Arduinos - an NG and an Uno, with an Atmega168 and 328 respectively. Both of these have the bootloader burnt onto them. I'm looking to burn the Arduino...

I spent some time on chat with cpf yesterday, but didn't get anywhere: we tried the cap/resistor across the reset line thing, though it looks like cpf is not doing that consistently, from the pictures cpf posted.
@angelatlarge 'then chip will need the external crystal/resonator to run at all, there is not fallback option such as "Oh, no external crystal? I'll just use the internal resonator")'
Funny to see that the XTAL-IN overrides the internal RC-oscillator even when the RC osc is selected :)
@angelatlarge I don't see a resistor nor capacitor on the reset line either.
@jippie Yeah, exactly.
6:36 PM
it should be pretty easy to add an extra clock output on arduino ISP using a spare timer.
@W5VO Was it a Belgian power supply?
@ThePhoton Nice!
@ThePhoton Pretty sure it was Scottish
@W5VO It was the hint of burnt toast that gave it away
@jippie You mean if cpf's chips are bricked? Right, you'd have to modify the ISP sketch.
6:39 PM
no in general in arduinoISP
version 3m5 :-p
@angelatlarge With a name like Angel, I knew you would start speaking Spanish eventually.
@ThePhoton No te gusta cuando la gente hablan Castellano? Lo siento.
@ThePhoton I used that line on my future wife-to-be. She was mad. Her name isn't Angel.
@angelatlarge I was gonna reply to your "hello" the other day with an Hola, then I thought, if his name is really Angel he might challenge me to follow up with more Spanish...
@angelatlarge Ha, I was replying to the wrong msg. I ment that for @ThePhoton.
6:44 PM
@DavidKessner You should have said, "with a heart like an angel, I expected..."
@ThePhoton Well, I'm married now so it's all good! :)
I really can't figure out what this Wellon VP-290 actually does.
Nothing on AVRfreaks
'Wellon VP-290'?
@jippie You know, the programmer in the "external clock" question.
nope, to be honest I totally read over that link
@angelatlarge and I assumed a simpe ISP 6 pin programmer
6:59 PM
@jippie LOL. Yeah, that happens to me a lot :)
Does anyone know of specific AVR chips that do not support ISP programming?
The BFG 10k^H^H^H^H^H err... Wellon VP-290 has a supported device list (but not much else in the way of docs). If no non-ISP chips are there we can surmise (maybe) that it is an ISP-only programmer.
@angelatlarge oh wait, is that where I saw that name before?
@jippie Not sure what you mean.
That 'Wellon' name sounded familiar, but I didn't know why :o)
It was from the oscillator question
@jippie Ah.
Ok, so ATtiny45 isn't an ISP chip?
No, that's false.
@angelatlarge ATtiny45 is a really small ATmega32
7:12 PM
@jippie Well, funny thing is that BFG (I mean Wellon VP-290) doesn't support it.
ATtiny10/11/28 are not ISP capable, and the BFG doesn't support them either.
I don't know what that means.
this scp guy is right, if avrdude reports wrong ID due to the resistor or capacitor issue, then it'd report the controller ID for the master Arduino.
@jippie Do you know any AVR chips that can't be programmed via ISP, but only via JTAG?
all zero means there is no data at all. it is probably open drain, maybe a line stuck or a pull up missing.
@angelatlarge no, but I have never searched for any.
@jippie I thought the resistor/cap was to prevent the master Arduino from resetting.
@jippie And if it resets you'd expect 0000 back
@jippie No?
@angelatlarge yes and then you read the wrong controller's id.
7:15 PM
@Kortuk / @W5VO The new user Suirnder needs to be gently nudged against posting his email address in answers and asking people to contact him thereby.
@angelatlarge Let me show you how I know that.
@jippie Maybe? Wouldn't it depend on where in the protocol the reset line is pulled low?
Q: Why does my ATtiny13 report wrong device ID?

jippieMy programmer reports: avrdude: Device signature = 0x1e9406 avrdude: Expected signature for ATtiny13 is 1E 90 07 The -F flag to force programming does not override the ID in this case. I know I can reset the ID with a high voltage programmer, but why did is the device wrongly reporting its I...

it's almost like a broken wire between master and slave
@angelatlarge No, the AVR family are ICSP-programmable by design, I believe.
@jippie Good work on the external crystal question. I'm out of votes for today, but I'll be sure to vote on it tomorrow.
I know there are AVR's that do not support both serial and parallel, but can't remember which it supports always.
You should always save some votes for me @AnindoGhosh. How am I ever reaching 10k rep otherwise?
7:18 PM
@jippie From my reading, smaller ones do not support parallel. But I have been mistaken often before :-)
@AnindoGhosh since when do you vote so much anyway?
@jippie I was on a vote binge today, sorry
@AnindoGhosh that does make sense, the smaller ones probably have too few pins.
@jippie I always vote on questions that I think are likely to be searched for by learners. I also sometimes vote on answers that I respect for reaching the beginner level and yet not being condescending. Plus I vote for you, don't I? You, Phil Frost, ThePhoton frequently get my votes.
@jippie Isn't ICSP the official AVR term for it, rather than ISP?
@AnindoGhosh ATtiny10 cannot be ISP-programmed
7:21 PM
@AnindoGhosh you should have known I couldn't resist to figure out the question even if I didn't know the answer from the top of my mind
@angelatlarge Noted. Interesting.
@angelatlarge is it so small that it doesn't have any flash at all ;o)
@jippie Yes, that's what I admire about you - the willingness and tenaciousness to drill through to an answer.
@AnindoGhosh I was trying to figure out if the programmer in the "external clock" question was ISP-only, or what. Very little documentation for it.
@AnindoGhosh I am not sure that's true. Atmel uses the term "ISP" in their docs. www.atmel.com/images/doc0943.pdf
@angelatlarge Tough call - I wouldn't buy it without documentation, since there are enough programmers for AVR which are well-documented, spanning ICSP, JTAG and everything else.
7:25 PM
@jippie That's right. No flash, no programs, no problem.
@angelatlarge OK, so ICSP is someone else's official term for it.
@AnindoGhosh Agreed 100%. But without knowing what it supports we can't 100% answer the question.
@angelatlarge Was there some MCU I read about, that only comes pre-programmed from the factory with code provided as part of the purchase order? It has RAM but no flash. I wish I could confirm that this isn't my imagination.
@jippie I'd have to think it through all the way: I don't full understand how PC->ArduinoISP reset line is used relative to what is happening on ArduinoISP->TargetChip end. And I am feeling a bit intellectually lazy right now.
@AnindoGhosh You mean the "program" is permanently burned on there? Like an ASIC?
@angelatlarge It is a bit like a RISC architecture (Reduced Instruction Set Computer): By making the architecture more simple by reducing the number of instructions, you can make the CPU as a whole faster. This means that if you continue doing that, you create processor that can infinitely fast do absolutely nothing.
7:28 PM
@angelatlarge Important point to note here: TDoes the OP say they have that precise programmer, or that they have something like that one? The latter could simply mean they don't know the difference between various programmers, so long as they have a socket on them.
@AnindoGhosh Ah true! I am still a n00b in the skill of translating from what OP said to what actually might be the case in the real world. Good point. Let's see...
"I'm using a universal programmer (like...."
@AnindoGhosh So yes, open question.
util/delay is a quite a bit more optimized than a for () nop loop :) @angelatlarge
it is accurate to the cpu clock cycle :)
@jippie I really don't know what "optimization" means in that context.
@jippie You mean "optimized" in terms of "well-suited for the task" I see
@jippie Good point.
if you ask for 1ms, you actually get 1.000000000 ms, rounded to the nearest CPU clock cycle
@AnindoGhosh Mmmm... that might be a bit strong, no? The OP did say "universal"
7:34 PM
@angelatlarge Universal = supports multiple different AVRs.
@AnindoGhosh Perhaps? It could also be "I am using this one" and they said it wrong?
@angelatlarge You need to work on your paranoia skills ;-)
@AnindoGhosh IRL: definitely not. Here: very true. I am amazed at your guys ability to mind read sometimes.
@angelatlarge My skill comes not from here, but from years of teaching to newbies.
@jippie You still here?
7:38 PM
@angelatlarge searching for easy rep on the stack
@jippie I think if your "Why does my ATtiny13 report wrong device ID?" had something in the question about what programmer setup you are using, that might be helpful. But I don't want to try to edit it because I don't know.
@jippie My offer of asking "How do you build a HV programmer out of an Arduino" still stands :) It'll be like shooting fish in a bucket :)
@angelatlarge it was an Arduino with ATmega168
@jippie Done. In the review stack now. Thanks, I think that would have helped me.
@AnindoGhosh thanks for the heads up.
@Kortuk How does on nudge a user like that?
7:44 PM
@angelatlarge so if I accept the edit, you will upvote the question?
@angelatlarge Approved the edit. I think it needs another approval, maybe not.
@angelatlarge a comment to them on a post and editing it out is the best way. If it continues, let us know.
@jippie Sure! I didn't know you needed the rep :)
@AnindoGhosh It does
@angelatlarge @jippie is a huge rep whore
whats going on guys?
@NickHalden Hallo.
7:47 PM
@NickHalden no I'm not!
This question, I think, the OP needs some fundamental hand-holding, he seems to be well-meaning enough, rather than one of those gimme da codez types. I need to get offline now, else I would've.
Q: What kind of Relay Reacts to Ohms?

jp2codeI need to connect a relay to my old truck to turn on the "soon to add" electric fans. The ECM has an engine coolant temperature sensor that uses the output resistance to determine coolant temperature. From the service manual, I have this data: If you can not see the pic above: __Condition__...

@NickHalden that is a gigantic understatement.
@NickHalden He's not huge. Just plump.
7:59 PM
@jippie Ah, sorry, I guess my vocabulary isn't developed enough to describe how big of a rep whore you are =P
@NickHalden not sure if words exist to describe it properly
@jippie lol, fair nuff
@NickHalden so did you upvote 4 questions of mine today?
@jippie nope, haven't looked at anything today... pretty busy at work
apparently 4 is a max (whoever decided that)
@NickHalden how come you have time for chat and not for important things?
8:03 PM
Stoopid question: How does one make the formulas larger when typing in MathJax? I am getting very small equations?
@angelatlarge ask it on meta
mathjax doesn't work in chat, so that'd be pretty difficult to discus here
@jippie Are you rep whoring on meta too?
@angelatlarge Yeah, was annoying me earlier.
@angelatlarge meta has no rep, if you did not know.
@Kortuk Oh, true
@angelatlarge hit CTRL + to enlarge
8:07 PM
@jippie :)
@angelatlarge If your fractions keep getting smaller, try using dfrac instead of frac
Do you have a specific example to see?
A: Test the new \$LaTeX\$ markdown in this Sandbox question!

jippieYes, I've been shopping around in various (upvoted) answers, but how about a ... CHEAT SHEET To use a formula like \$\frac{Math}{Jax}\$ inline with the text of an answer, surround it with \$, like this: \$\frac{Math}{Jax}\$ To create a formula centered $$\frac{Math}{Jax}$$ within the article...

A: What kind of Relay Reacts to Ohms?

angelatlargeIf you don't need a lot of precision in your circuit, you can build a simple voltage divider as follows simulate this circuit – Schematic created using CircuitLab If you calculate the values in such as way that the voltage at output reaches the relay threshhold voltage, then when R1 get l...

@angelatlarge Use dfrac instead of frac. I'll edit it for you
@W5VO Yeah, it is much moar better!
8:17 PM
@AnindoGhosh I am just gonna migrate that one and see what happens. Ignore my comment.
@W5VO Thanks, but don't you have better things to do (like protect questions, ban users, and generally moderate stuff) than editing my posts?
@angelatlarge you never said anything to me when I was being helpful, this @W5VO guy gets all the credit.
@Kortuk Apparently, Mr T doesn't love you quite as much.
@angelatlarge My job is to help and improve the site. This is much more enjoyable than playing whack-a-troll
@Kortuk It's a dup anyway, right? Didn't we see this like 2 days ago.
@W5VO W-a-T. I like that.
8:21 PM
@angelatlarge another Latex tip is that you can use \large and \huge and other "size words" to tweak the size of text for a relative effect. there are a few examples on Meta where an absolute font size is given
@jippie Nice
@jippie cuz time stops when you're in chat?
right guys?
@NickHalden approved.
8:25 PM
@Kortuk fantastic
8:38 PM
Man I really need to get some actual work done, something that isn't SO/ee.SE
@angelatlarge a constant issue.
@angelatlarge way past beer o'clock here.
@jippie NL?
what is the question?
8:53 PM
@jippie You are in the nederlands, no?
I am in the netherlands.
9:04 PM
He is a "Need-err-land-err"
if you wondered how I pronounce that.
9:18 PM
@Kortuk last night i was drinking a beer and wondering how you pronounce that
@NickHalden your welcome :)O
I'm not sure this is mathematically valid: For example, a 1Hz crystal with 1% jitter would give you 1 sec +/- 1% ticks. A 1kHz clock with 1% jitter going through three divide by 10 chips will give you 1 sec +/- 0.001 % ticks.
@AnindoGhosh Agreed. Seems fishy. If there is a 1% error over time t, then there is also 1% error over time t*1000000.
@AnindoGhosh Not fishy, just wrong I think. Unless I misunderstood something.
@AnindoGhosh But I am NOT going to comment.
yes 1% is 1%
and @Kortuk's pronunciation is ... poor, to say the least
relative error is constant when dividing by an exact number (like 2)
@jippie That's what I thought. But I need to chill out a bit in the comments: I've been a stick-in-the-mud a lot lately
9:29 PM
@jippie I saw it with a Deutsch accent that way also
Generally @angelatlarge @AnindoGhosh @jippie crystals specify in PPM, which in that case your error goes down for the same PPM with a different crystal, I just think they used the wrong units.
@Kortuk as long as you enjoy thinking you're good at it ;o)
because 1% is 1%
I dont think I am good at it, I just enjoy it.
@Kortuk What's PPM in this case?
bagel is pronounced Bag - ell when I say it.
Because why not.
@angelatlarge parts per million
it is an error in number of extra clocks you can have per million normal clocks.
with dividing, you add relative errors, but with an exact number like 2 is, the relative error stays the same.
if you have a really cool crystal it would be ppb
9:31 PM
So a 1MHz clock with +-80ppm would count between 80 too few and 80 too many ticks after 1 second(1 second being 1 million oscillations at 1MHz.)
really really cool then
@Kortuk And a 1Hz clock with +-80ppm would have the exact same accuracy, no?
yes every 1 million seconds, you can be up to +/- 80 seconds off
@jippie +1
@angelatlarge wait, am I doing this wrong.
9:40 PM
@Kortuk Oh good.
@Kortuk ...the story of my life...
That answer got two upvotes already. Wow.
Given the speed of sound in typical solids, a 1Hz crystal fabricated the same way as watch crystals would likely be about a thousand feet long..
that would make a 32kHz crystal still 1000/32000 ft?
oh that is about 1mm?
that should be OK thn
time for bed
Is the asic tag a mistake on the RTL question? I don't understand why it is there
EE hangout: "in your timezone, that's 7:00 on Saturday" are you kidding me @Kortuk!?
9:51 PM
@jippie does that mean 7AM?
and it would say 19:00 if it was PM?
I think @Kortuk must have hurt his head badly
wait 1:00
Starts in 7 hours
9:51 PM
means its 1AM
/me is out
@jippie gnight
@jippie Agreed
@jippie 6am in the UK :'(
we can do our own hangout later ;o)
10:28 PM
@AnindoGhosh @AnindoGhosh -- I tried to give the explanation in a comment. Take a peek @ fig 6 in silabs.com/Support%20Documents/TechnicalDocs/… and look how the phase noise goes down. Same phenomenon. The table that follows shows that the jitter expresses as time, not %, stays largely the same as division takes place, meaning that the % jitter goes down (same numerator, larger denominator).
@ScottSeidman Are you saying that at any particular point the error will be a single "tick", and it is better to be off by a small tick than a large tick?
@angelatlarge @angelatlarge, no. To me it looks like folks are interpreting what I call "jitter" to be the frequency tolerance. It's not. Frequency tolerance speaks to what the measured frequency actually is with respect to the labelled frequency. Jitter refers to how much the period can vary, given the actual frequency, on any given cycle.
@ScottSeidman I was confused by this as well, that's why I am asking. I think what you are saying is that even though each cycle can be +/- (say) 1%, it is extremely unlikely that on average this +1/-1% doesn't accumulate, because you are equally likely to have +1 cycles as -1 cycles. Did I understand you correctly?
10:49 PM
@angelatlarge Jitter is "random variation of a timing instant from the ideal value", and it normally has very high frequency. If there's a slowly varying timing error, it's often called "wander" instead of "jitter". If there's an error in frequency that varies slowly enought to accumulate in a many-times-divided-down clock, I'd call it "drift" instead of jitter.
Of course you can always find somebody to call your problem "jitter" no matter what the problem actually is.
@ThePhoton Thanks, that makes sense to me, but I had hard time figuring that out from the original answer.
I will admit the boundary between jitter and wander is not well-defined.
@ThePhoton In the audio world, we call that wow and flutter! :)
@DavidKessner In my world, we call audio "DC".
@ThePhoton Bigots! :)
11:02 PM
It's actually one of my big complaints. A lot of engineers here seem to think that anything that happens below 1 GHz is not worth their attention. We get some pretty crappy designs as a result.
I gotta run. Have a good weekend, all!
@ThePhoton And audio people ignore anything above maybe 100 MHz. :)
With equally crappy results.
@DavidKessner Later.
@angelatlarge @angelatlarge -- the error has zero mean and doesn't accumulate. If the error accumulated, the frequency would be wandering.
@ScottSeidman Right, thanks for the clarification, I didn't understand that. It might be helpful to edit that into your answer.
11:45 PM
@angelatlarge @angelatlarge -- Done, I added the silabs pdf explaining the phenomenon -- but jitter has a pretty commonly accepted definition
@ScottSeidman Thanks! Point about standard definition taken.
00:00 - 16:0016:00 - 00:00

« first day (890 days earlier)      last day (4117 days later) »