Electrical Engineering

vicatcu

04:07

I'm desperate for some attention on this question: electronics.stackexchange.com/questions/65754/…

The Photon

@vicatcu I don't think there's a "one best answer." Using bigger separations gets you more total range with less steps. Using smaller separations gets you better resolution within each range.

vicatcu

@ThePhoton I need to come up with an algorithm for picking a value

but i'm barely conscious right now, so i'm off to bed

Anindo Ghosh

04:24

grEEtz

The Photon

@AnindoGhosh Howdee

W5VO

05:29

@angelatlarge He wasn't omitting vowels, it's just hexadecimal ;)

-2

Q: S RECORD QUESTION

Plz tell me how we can data from S-record file Details of controller and hardware is mc56f8014 and file is like this S30D0000000054E15C0054E15C00D0 4 byte address 0D00 DECIMAL EQUAL 3328 4 BYTE FOR ADDRESS AND 1FOR CHECK SUM PLZ TELL ME SOMETHING HOW WE CAN GET DATA FRO TI HOW CAN BREAK IT .

microcontroller

jippie

05:56

good morning @all

Power-Inside

06:18

Hi everyone, I'm a bit new to electronics, I have made this setup and connected it to a 9.6V DC adapter.. but now my multimeter reads the volts as 22 (which is probably exceeding for the LM386). if not connected to the circuit, the 9.6V adapter gives around 17 Volts.. weird.

I'm confused now and I don't know what to do.. when I ran on 22 V using this 9.6V adapter, my LM386 got really hot and probably broke.. It was also too loud even without the 200 Gain boost cap..

jippie

@Power-Inside Measuring DC while your multimeter was in AC? That gives wrong readings.

An unregulated adapter can give a higher voltage.

So you have to check the reading on the adapter.

Power-Inside

multimeter was in DC mode. It is an unregulated adapter and the rating on the adapter says 9.6V output..

jippie

06:45

@Power-Inside if it is unregulated, there is probably no way of telling what the voltage will be when it is (nearly) unloaded and it can be much higher than at rated current.

Power-Inside

@jippie Okay.. So is there any way of limiting the voltage for the circuit? Or should I just buy a regulated adapter?

jippie

@Power-Inside a regulated adapter is the easiest way to go, but you can do the regulation yourself too. On the other hand, if you are looking at 22V unloaded, then you are facing pretty much heat dissipation in the regulator.

Anindo Ghosh

@W5VO "PLZ" is not hexadecimal, it's barely even human.

jippie

@Power-Inside You may have an old mobile phone adapter lying around that can be used.

Power-Inside

@jippie Actually, this is one old mobile phone "quick charger" from china..

jippie

06:53

@AnindoGhosh yeah it mustn't be written in all captitals. 'plz' will do fine.

@Power-Inside hmm ..

Anindo Ghosh

@Power-Inside If you must use a charger as adapter, at least find a USB charger (they theoretically provide a 5 Volt regulated output) preferably of a well-regarded mobile phone brand.

jippie

@Power-Inside I'd personally ditch this particular one for this project.

@AnindoGhosh Maybe 5V is a bit low for an LM3-something audio amp, but otherwise you are right and it may just work fine.

was it LM386 or so?

/me is out

Anindo Ghosh

@Power-Inside Better yet, just get yourself an 7805, 7809 or 7812, depending on what actual voltage you want, use that at least for the experimentation. Those parts are dirt cheap almost anywhere in the world. Once you've established the set-up works, buy yourself a DC-DC buck regulator off ebay for $3-5, so you don't have an overheating linear regulator.

@jippie Fair point. The unregulated mobile phone chargers just piss me off, the output is totally random.

@Power-Inside Tantalum capacitors are not non-polarized.

jippie

not non-polarized ...

double

I had to read twice

anyways

on my way out.

did you hear about the boston bombing this morning @AnindoGhosh?

Anindo Ghosh

@Power-Inside [From Wikipedia](http://en.wikipedia.org/wiki/Tantalum_capacitor):

**Polarity**

SMD tantalum and aluminum capacitor polarity markings. Some types are marked on the positive terminals, but others are marked on the negative terminals
Most tantalum capacitors are polarized devices, with distinctly marked positive and negative terminals. When subjected to reversed polarity (even briefly), the capacitor depolarizes and the dielectric oxide layer breaks down, which can cause it to fail even when later operated with correct polarity. If the failure is a short circuit (the most com…

jippie

07:01

2 fragmentation bombs at the marathon of Boston exploded

Anindo Ghosh

@jippie Yes, I was awake and reading Reddit when the news appeared everywhere.

jippie

listening to the news for a moment now, before I go

Anindo Ghosh

@jippie I wanted to hear from @rawbrawb as he mentioned that some people he knows were running the Marathon.

Power-Inside

07:40

Thank you for all the information. About caps without polarity, IDK exactly what those orange small ones are called. They don't have any indication of polarity anyway, so they're probably non polarised.

About regulated supply, I'm considering an old USB cable to connect off from my laptop instead of an adapter.. that ought to give it a nice regulated 5 Volts atleast for testing, right?

Anindo Ghosh

07:53

@Power-Inside The USB cable will provide a maximum of 500 mA, which may or may not be enough.

Power-Inside

Hmm.. true.. how do I know how much current my circuit might need? Meanwhile, I'll just try this now and see how it goes. It's probably safe I guess.

Anindo Ghosh

08:33

@Power-Inside Yes, USB ports generally have overcurrent protection.

Power-Inside

09:02

Got my hands on a 7808 IC! :D

weird though.. input pin reads 5V from usb and the output vs ground pin shows 0.21 Volts for the 7808..

does it need above 8Volts input?

Anindo Ghosh

@Power-Inside Any 78xx series IC requires input at least a couple of volts over its designed output, and preferably 3 volts or more.

@Power-Inside It also needs the capacitors for stability, else it could oscillate.

Power-Inside

09:26

I ditched the regulators for now and I'm feeding it 5V directly..

Power-Inside

10:56

@AnindoGhosh I updated my question with new findings.. electronics.stackexchange.com/questions/65478/…

Anindo Ghosh

@Power-Inside Listening to the sound clip, it sure shows progress. I have two inputs: 1. If nudging changes things, you have a loose contact. While this could be something not making contact till nudged, which is easy to solve by wiggling all leads around a bit on the breadboard, it could also be some component being pulled out of the circuit on a nudge, the likely culprit being some capacitor which is either not functional, or is leaking too much current.

Identify and remove that cap, and you should get stable behavior, however noisy.

2. From the sound itself, it feels like the sound is going beyond the operational range of the amp on one of the two rails. If you are sure the input signal is small enough for the expected gain to not exceed the operating range (5 Volts minus 2 Volts), then there's some biasing away from ground level. Check with a multimeter set to DC mode (yes, strange), that should show you if there is some DC bias in the input...

The same multimeter should also show the output ideally biased to precisely 2.5 Volts (or 0 volts if measured after the output cap).

If either of those reads not zero, then that's to be fixed.

From the sound, it's as though one half of the signal (upper half-cycle or lower) is getting clipped, or the power source is unable to provide sufficient current for the peaks. That's very subjective, though, from situations where I've generated the same kind of muffled sound.

Power-Inside

11:28

@AnindoGhosh Actually, nothing was loose. Only nudging the power cable to make the circuit go from open to closed, causes a short peak/spike in the sound.

Anindo Ghosh

@Power-Inside That might imply that the source is unable to sustain the current draw of the system. When there's a disconnect / reconnect, the source supplies a spike of current till it sags due to load.

Power-Inside

@AnindoGhosh DC volt reading shows around 0.26 volts at the input.

@AnindoGhosh After output cap, I get 0.35V

Anindo Ghosh

@Power-Inside Ouch. Umm ok hold on. You don't have an oscilloscope accessible by any chance, do you? Because some multimeters don't actually do an averaging to arrive at a DC voltage, they read junk when provided an AC input.

Power-Inside

Nope no oscillators.. :/

Anindo Ghosh

@Power-Inside This is something that needs a scope, I'm thinking.

Power-Inside

11:33

Probably.. since it involves signals

Anindo Ghosh

Not oscillator, oscilloscope. If you're the persistent sort, I could suggest work-around circuits such as I put together for such things.

For instance, to read the DC bias of an AC signal, I have a simple battery operated (that's important as you'll see) quad op-amp device I made: One op-amp is used as a voltage follower with a carefully adjusted preset for precisely midway between the supply rails. The next op-amp is a voltage follower for the input signal, to provide a very high-impedance input. The third is a low-pass filter at 1 Hz, gain of 1. The fourth is another low-pass at 1 Hz (because I had a spare), gain 1.

@Power-Inside The reason I emphasize battery operated is, the mid-rail I created with the first op-amp is used as the ground reference for input and output. That way, I can take any input, and its ground, reference them to my gadget's "ground", and get a reliable reading of the DC bias if any on the incoming signal.

Using a cheap LED voltmeter module from eBay.

Power-Inside

That's probably too much work for me.. lol

Ok I just played some 'chiptunes' over this and this time, I seem to be getting better output, though the low frequency sounds (bass) seem to go heavily clipped/distorted.. High frequency is perfect. Also I removed the 100uF cap from the power rails.

nvm I just reduced the volume at 60% and everything sounds smooth.. now I'll just try the 200 Gain. Hope it works.

Anindo Ghosh

11:48

@Power-Inside Nope, if 20 gain needs reduced volume to work for you, how would you expect 200 gain to not clip?

Power-Inside

Hmm.. I was expecting some loudness..

Wouldn't mind a few distortions if I can achieve some degree of loudness.. I guess I need to increase the voltage rather than gain?

Anindo Ghosh

12:10

@Power-Inside Yep. The clipping occurs because the output voltage cannot go beyond the power rails (Vcc and GND). There's a technique known as Bridge-Tied-Load, using two amplifiers, that can get you to double that, but beyond that, you'd need a 1:2 or 1:3 or whatever audio frequency transformer for further boosting.

Anindo Ghosh

12:38

@Power-Inside Low frequency sound consumes more power to amplify, what you are seeing is the USB supply hitting its current limits.

Anindo Ghosh

13:15

Nice reference for power saving on AVRs:

►

Kortuk

13:30

Morning guys.

Anindo Ghosh

14:02

@Kortuk Hiya

Kortuk

@AnindoGhosh Hope you are having a good day.

Anindo Ghosh

@Kortuk Going through Orcad tutorial videos. I'd called it "mixed" rather than "good" :-D

Kortuk

@AnindoGhosh haha.

Anindo Ghosh

14:34

@Kortuk are you open to very beginner level questions about Orcad?

Kortuk

@AnindoGhosh always :)

Anindo Ghosh

@Kortuk OK, so I'm starting with something trivial. I have placed a CON9 9-pin connector on the schematic, I would now like to assign a footprint to it, essentially a 0.1" pitch pin-header footprint of 9 pins. How do I do this?

Kortuk

@AnindoGhosh I thought you meant on the site, I have only used it a little and not in 4 years. Sorry.

Anindo Ghosh

@Kortuk No, these are too trivial to justify a question on the site. I'll google a bit.

Kortuk

@AnindoGhosh Well, if you need to google it, many others do also, sounds like what we want as google bait.

Anindo Ghosh

14:40

@Kortuk Not going for that one :-)

Holy cow! Orcad chokes on spaces in folder names! What year is it, 1990?

Kortuk

@AnindoGhosh but think about how many experienced good engineers have these questions and we could draw in!

@AnindoGhosh yeah, stuff like that kills me!

Anindo Ghosh

@Kortuk OK, I'm probably not going to stick with Orcad for more than a day. I have at least a basic level of expectation from a modern CAD product. Good thing I'm still experimenting on a friend's set-up, I just saved me some money and some disk space I think.

Kortuk

@AnindoGhosh Yes, agreed.

rawbrawb

15:02

@AnindoGhosh thanks for asking... RE: Friends. Turns out everyone that I know is fine.

Anindo Ghosh

@rawbrawb Good to hear

user61389

15:20

Hi all!

Kortuk

16:03

@rawbrawb glad to hear it.

W5VO

@AnindoGhosh why are your OLED display comments starred?

Anindo Ghosh

@W5VO I have no idea, please unstar them, thanks.

Kortuk

@AnindoGhosh We dont, someone decided to. How it happens.

morning @DavidKessner

Anindo Ghosh

@Kortuk Isn't there a way to unstar them?

W5VO

@AnindoGhosh yes

Kortuk

16:06

@W5VO but the user who starred them has the right to star them, I am not sure mod powers make sense there.

angelatlarge

@W5VO Rly? You can unstar?

Anindo Ghosh

Thanks. :-) While you're at it, everyone please add stars to the moral code comment, isn't there a badge for starred comments?

W5VO

@angelatlarge I have super star powers

angelatlarge

@AnindoGhosh ARE YOU BADGE BEGGING?!?!?!

W5VO

@angelatlarge stars don't give you rep

Anindo Ghosh

16:07

@Kortuk My thinking is, that user was unfamiliar with the purpose of starring comments.

Kortuk

@AnindoGhosh might be the case, but starred comments dont hurt anything, why not leave them :)

angelatlarge

@W5VO Noted

Anindo Ghosh

@angelatlarge Nope, badge hunting. No point rep begging, I'm as far up as I can go.

@Kortuk Because then it hides the nicer comments, like the brilliant sock puppet one.

Kortuk

@AnindoGhosh ha. I do love that one

Anindo Ghosh

which actually deserves more than the measly 2 stars

@W5VO Yep, you're a superstar.

David Kessner

16:09

Morning all.

Anindo Ghosh

That hangout link is outdated too

jippie

17:53

good morning

user61389

Hi @jippie

jippie

Hi @CamilStaps

David Kessner

@jippie

Hi

jippie

@DavidKessner :)

so did I miss anything good on the stack today?

user61389

I think the can undervoltage/current do any damage questions are fun, but nah, not really - that I know of

David Kessner

18:01

@jippie If you did, then I missed it too.

angelatlarge

Hallo all

@jippie Something good coming up for you from yours truly

user61389

Hi @angelatlarge

David Kessner

@angelatlarge Hi.

I'm playing around with some new thermoelectric coolers for testing at low-ish temps.

angelatlarge

@DavidKessner Ooohhh.. I've been trying to use TECs for beer temperature control...

jippie

@angelatlarge 16 degrees Celsius.

David Kessner

18:03

@angelatlarge I doubt that they work well enough to make a lager. They are tricky to use.

jippie

that is the perfect temperature to drink Westvleteren 12

angelatlarge

@DavidKessner Nah, I am into Ales

David Kessner

The thing I just found out is that higher power does not equal more cooling.

angelatlarge

@DavidKessner I thought it did, in the abstract, but the curve flattens quite a bit... and the device produces more heat, so yeah...

jippie

@DavidKessner peltier?

David Kessner

18:05

@jippie Yeah.

@angelatlarge It is a balance between more "inherent cooling" and getting rid of the heat generated.

angelatlarge

@DavidKessner Right!

David Kessner

My setup is more complex because I am using two stages. This requires more tweaking to get the balance right.

jippie

@DavidKessner did you ~~upvote~~ check this question

8

Q: How to drive a Peltier element?

Maybe the underlying question is what the Voltage-Current curve looks like. Can I drive it from a voltage source (like you drive a heater) or from a current source (like you drive an LED)? Or even different than those two options? ADDITIONAL1: Say (hypothetically) I have two commercially availab...

peltier

David Kessner

@jippie Actually, I did upvote it. Although I am having second thoughts. :)

jippie

@DavidKessner think about it for another 5 minutes

David Kessner

18:10

I'm driving my coolers with a benchtop power supply in constant-current mode.

jippie

constant current is a good start

David Kessner

@jippie I have two identical coolers stacked (with a giant heatsink on the top one). The top cooler is running 6 amps, the bottom is 3.5 amps, and the chip it's cooling is -4.2 deg C. I'm surprised that the 6 vs 3.5 difference is so big, but that is what appears to be giving me the best cooling.

jippie

?

they're not in series? or parallel?

David Kessner

@jippie Two channel benchtop supply. One channel for each stage.

angelatlarge

18:44

0

Q: Unexpected global variable read result in C++ using avr-gcc for (local variable access is as expected)

I am getting unexpected global variable read results when compiling the following code in avr-gcc 4.6.2 for ATmega328: #include <avr/io.h> #include <util/delay.h> #define LED_PORT PORTD #define LED_BIT 7 #define LED_DDR DDRD uint8_t latchingFlag; int main() ...

gcc assembly global-variables avr avr-gcc

@jippie Free beer for you!

@jippie I just can't shake the feeling that I am a total moron, missing something blindingly obvious

@DavidKessner How did you determine that? Experimentally?

David Kessner

@angelatlarge Yes. I'm still playing around with the whole thing, trying to tweak it even more.

jippie

19:25

@angelatlarge If I compile your code, my mainloop looks like:

00000090 <main>:
90: 8f ef ldi r24, 0xFF ; 255
92: 8a b9 out DDRB, r24 ; 0x0a
94: ff cf rjmp .-2 ; 0x94 <main+0x4>

angelatlarge

@jippie What processor are you compiling that for?

jippie

avr-gcc -g -DF_CPU=16000000 -Wall -Os -Werror -Wextra -mmcu=atmega328 -Wa,-ahlmns=project.lst -c -o project.o project.cpp

angelatlarge

Can you try -mmcu=atmega328p?

jippie

1 sec

angelatlarge

Just to give you my full command line:

(I should post that in the question, one sec)

jippie

19:29

I am curious for avr-objdump -C -d project.elf too

@angelatlarge same code. PORTD is not in the listing at all

angelatlarge

@jippie Can you pastebin your entire disassembly?

jippie

I have no clue how pastebin works

angelatlarge

@jippie Go to pastebin.com, paste the disassembly there, it will give you a link, and you paste the link here :) It's easy!

jippie

pastebin.com/bFzUhtzf

What happens when you change your -O0 to -Os?

angelatlarge

Good question.... let me see...

I think I tried -O1 before, but not -Os

With -Os seems to behave as expected (i.e. does not turn on)

jippie

19:43

@angelatlarge Do you get the error or warning on avr/delay with -S0?

angelatlarge

@jippie Yes (I updated the question)

With -Os there is no loop (it is optimized away, as it should be, one would think :)

If I declare the latchingFlag to be volatile, presumably even with Os the loop should be there... let's see....

Yes, the loop is back... and the behavior...

the strange behavior persists.

jippie

-O0 => error
-O[s12345] => no loop

angelatlarge

@jippie Unless latchingFlag is volatile, in which case => error

user834

I'm having trouble displaying math when enclosing them in single '$'s

user61389

@user834 You need \$

user834

19:47

For example, $$x$$ renders, but $x$ does not

ah, thank you

jippie

yes, changed it to volatile too

angelatlarge

@jippie Wait, was this you reporting your results, or asking about mine?

jippie

What if you remove the delay?

@angelatlarge mine

angelatlarge

@jippie No difference. Let is essentially PWM-ed at that point, but you can tell the difference between it being full on, full off, and PWM.

@jippie Thank you. So I am not insane, right?

user61389

@angelatlarge not because of this ;)

jippie

19:49

let's regard that as a retoric question

angelatlarge

@CamilStaps I appreciate your vote of confidence, but I want to see @jippie say it :)

@jippie But I am getting error behavior with -Os and volatile uint8_t latchingFlag too...

jippie

@angelatlarge Is it by any chance overwriting the PORTD when in -O0? In other words, the controller more or less by default sets the output low, then does some checks and 'realizes' it should have been high?

@angelatlarge ?

angelatlarge

@jippie Let's use another port?

@jippie Or should we try to reproduce the -Os error for you?

@jippie You sure you don't get that with +Os and volatile? I'll try again to be sure

jippie

wait a minute

are you switching on and off the pull up resistor?

I have to carefully analyze the code

never mind the pull up

angelatlarge

Yeah, just checked that. If DDxn is written logic one, Pxn is configured as an output pin.

And besides, even it was the pullup, the logic is still foobar.

Yeah, I just reproduced with -Os.

jippie

19:55

throw out more lines until it stops :)

you might want to try another pin

angelatlarge

@jippie pastebin.com/4TkP5A6Q

@jippie I was going to say "yes", but... the logic isn't about the pin per se, is it? The logic is about the damn variable.

jippie

unless the pin isn't being toggled by your program, but by some other mechanism

angelatlarge

@jippie But the fact that moving latchingFlag to be main()-local solves the problem suggests that the LED is being twiddled by my program, no?

jippie

@angelatlarge so you are saying the byte in memory is being changed, not the output pin?

angelatlarge

@jippie Is don't know if latchingFlag is being changed or is read incorrectly. But I think the fact that moving latchingFlag to be declared inside main() makes the LED not blink means that it is LED_PORT ^= 1<<LED_BIT; that is blinking the LED, not some random timer thing.

jippie

20:03

How are SPH and SPL set for your controller?

hmmm you don't use the stack, do you?

angelatlarge

@jippie What are SPH and SPL?

jippie

stack pointer

angelatlarge

@jippie When latchingFlag is declared local to main() it seems to be on the stack.

@jippie The code I posted is literally all there is. So to answer your question: I don't know. I could look in the disassembly.

jippie

did you do the obj-dump?

avr-objdump -C -d project.elf

are you on linux?

angelatlarge

@jippie cygwin

jippie

20:11

o. ... so no perl available?

angelatlarge

@jippie It's like unix. So yes, perl available.

So I am looking at avr-objdump -C -d project.elf, and...

jippie

here is a script I use to insert register names, but it is far from finished. Good enough for ATmega328 if you change line 11 to 'm328' git.linformatronics.nl/gitweb/…

@angelatlarge and?

angelatlarge

@jippie Can you explain what this is going to show us? (Granted it is cool, but...)

jippie

instead of showing out 0x0b, r24 it shows out PORTB, r24

oh just throw the dump on pastebin

I want to compare it with mine

nearly time for bed

angelatlarge

After ./avrRegisters.pl or before?

And which version you want? -Os + volatile? With delay or without?

BTW, did you see this?

0

A: Unexpected global variable read result in C++ using avr-gcc for (local variable access is as expected)

According to your disassembler listing, latchingFlag global variable is located at RAM address 0. This address corresponds to mirrored register r0 and is not a valid RAM address for global variable.

That doesn't seem exactly right, but could be on the right track. I don't think my listing shows where latchingFlag is allocated though.

Ah, wait, maybe it does....

jippie

20:22

r0 is usually 0

do -Os

Actually the variable is read from SRAM address 0x01 in my listing

no wait, I'm lying

angelatlarge

@jippie I am trying to read the address from the opcodes

@jippie You want to see gcc-type dump or avr-objdump?

jippie

avr

angelatlarge

@jippie pastebin.com/v3m2yuSa

@jippie Isn't pastebin awesome?

jippie

the variable is stored by std Y+1, r25 in my disassembly. Not sure what Y is though, it isn't initialized.

angelatlarge

@jippie That's the local variable version

@jippie Y is the stack pointer

jippie

20:29

?

angelatlarge

@jippie No, it isn't zero!

jippie

Y = stackpointer!?

angelatlarge

lds r24, 0x0060

@jippie Yes. First reference I can find (there are many): lists.gnu.org/archive/html/avr-gcc-list/2012-07/msg00007.html

jippie

Y = r29 << 8 + r28

angelatlarge

Hmmm... but 0x0060 is WDTCSR, it would seem.

WDTCSR – Watchdog Timer Control Register

jippie

20:32

the avr-dump you posted, what C source goes with that?

angelatlarge

@jippie pastebin.com/zve6dcZD

jippie

I am definitely storing the latchingFlag in different SRAM location

0x060 (yours) vs 0x0100

angelatlarge

@jippie What version of avr-gcc are you running?

@jippie 0x60 would be a problem, would it not?

jippie

avr-gcc --version

avr-gcc (GCC) 4.7.0
Copyright (C) 2012 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

@angelatlarge was just about to ask you that same question

but it is in your question

angelatlarge

Yes

4.6.2

jippie

20:42

is upgrading to 4.7.0 an option?

angelatlarge

Maybe I need to upgrade.

jippie

try to find 4.7.0 so we can compare results

angelatlarge

Your -mmcu is atmega328? With p or without?

jippie

avr-gcc -g -DF_CPU=16000000 -Wall -Os -Werror -Wextra -mmcu=atmega328p -Wa,-ahlmns=project.lst -c -o project.o project.cpp

angelatlarge

@jippie Yeah, I'll do that later.

@jippie Yeah, OK.

@jippie Do you think if that's the case I should accept Egor Skriptunoff's answer?

@jippie I haven't upvoted it even at this point. I think it was very useful, but not exactly right, you know?

jippie

20:44

It is the best answer so far ;o)

it is useful, absolutely

angelatlarge

@jippie Agreed. But I can't tell if it is "accept"-type material.

OK, gotta go make food, otherwise I'll eat my ATmega328, LEDs and all

jippie

it is a solution?

(think not)

if upgrading avr-gcc is the solution, then someone should type that as an answer :o)

angelatlarge

@jippie Ha! Actually I think I probably have 4.7.0 on my linux, so I can check without having to do much.

@jippie I think you deserve that free beer :)

@jippie I'll be happy to accept your answer :)

@jippie OK, have a good night! Gonna go eat now.

jippie

time for bed

/me is out

angelatlarge

@jippie I think you should credit Egor in your answer though :)

goodnight

jippie

20:48

OK, I'll type an answer before I leave

David Kessner

I'm glad you folks are flooding the chat logs with this stuff. That way when he who shall not be named goes through it, he'll fall asleep before he gets to the part where I talk smack about him. :)

(Just so we're entirely clear. You're fine. I'm just giving you a hard time about it.)

angelatlarge

@DavidKessner Is there something better to flood chat logs with?

@DavidKessner Beer?

David Kessner

@angelatlarge No. But that's irrelevant! :)

angelatlarge

@DavidKessner Curious to hear if Bahstan Pahlice has anything new to say about the investigation in 2 minutes.

David Kessner

@angelatlarge I am not following.

angelatlarge

20:59

@DavidKessner Sorry. I am in Boston.

David Kessner

@angelatlarge Oh, you mean the cops!

angelatlarge

@DavidKessner And if you are wondering why I hit the reply button, I am not sure either.

@DavidKessner cahps

David Kessner

@angelatlarge Is there a press conf?

angelatlarge

@DavidKessner Yeah, a local one

@DavidKessner Friend of mine was in a restaurant right across the street from the second blast.

David Kessner

@angelatlarge Doing fine?

The daughter of a coworker had just finished the race and was standing around watching the finish line when it went off. Yikes!

She's fine.

angelatlarge

21:02

@DavidKessner Yeah, completely unhurt (except emotionally). Some friends of friends were very close too.

W5VO

@angelatlarge I swear I saw Bahistan... and was thinking "Where the hell is that?"

angelatlarge

@W5VO Bahston. Bahistan is your invention :)

@W5VO Bahstan, I meant.

W5VO

@angelatlarge Bahstan is where people from Boston claim they are from so others can't find them

David Kessner

I finally got those coolers adjusted and damn they work good! Cooling my chip to lower than -20 deg C.

W5VO

@DavidKessner That's some nice active cooling

angelatlarge

21:04

@DavidKessner Wow. Can you write up a question/answer about that? Please, pretty please, with sugar on top?

David Kessner

@angelatlarge What do you mean Q&A?

Kortuk

@DavidKessner He wants you to share by asking a question where you write the answer for others to read.

The Photon

@DavidKessner Hope you have a good dry air supply!

David Kessner

Title of the Q: How do you cool a chip down without using gallons of freeze-spray?

@ThePhoton I'm in Colorado. I have a good, dry air supply. :)

I have an FPGA on the PCB. On top of that is the thermoelectric cooler. Then heatsink+fan. Multimeter is measuring Deg C from a thermocouple on the top of the FPGA.

W5VO

@DavidKessner Liquid nitrogen

we actually have several environmental chambers for a whole bunch of temperature ranges

we have a "refrigerator" style probe station cooler for bare die that goes to -65C

and we have a LN2 cooler/oven and a LHe vessel as well

David Kessner

21:11

@W5VO This is a fairly large unit, and would require a large chamber (which we don't have the budget for). One day, a few months ago, I was doing testing outside when the temp was -11C (+12F).

W5VO

@DavidKessner Looking at it, your equipment is bigger than our box :)

Is this purely environmental verification?

David Kessner

@W5VO Basically. We had an FPGA issue that only popped up at low temps. It was easier to build this thing (cost about $25) than to buy a chamber big enough.

The Photon

Lab where I work has about a dozen of these:

W5VO

@DavidKessner Hold conditions and clock skew suck

The Photon

I don't think we can do -65, but we can get -40 or so.

rawbrawb

21:24

@ThePhoton That's not fighting fair! The poor dude is cobbling together a hair dryer running from a trained squirrel and getting results and you pull out your Ferrari !

And then complain about ONLY being able to go 150 MPH!

;)

David Kessner

@W5VO It wasn't a timing issue. It was a signal integrity/driver issue. Not sure why the cold mattered, but it does.

The Photon

@rawbrawb On the other hand, if you've ever been in a closed room with a dozen of those things running, you might rather use the squirrel-driven TEC.

W5VO

@DavidKessner Weird - like the internal termination resistance was shifting too much?

rawbrawb

@ThePhoton just because you're using 10 of the 12 units to cool your beer ....

W5VO

or the driver was suddenly giving too sharp edges?

rawbrawb

21:32

@W5VO That or the CMFB operating point shifting with temp. Eye diagram will change subtly and the Intersymbol interference will start to pop up - as a guess.

David Kessner

@W5VO I have no idea what was happening. It only did it on the latest batch of PCB's, where we've shipped 500+ other units without issues. Played with the TX Pre-emphasis and RX Equalization and it started working.

angelatlarge

@DavidKessner I mean post a question and answer it yourself, "In Q&A format" as SE suggests.

David Kessner

@angelatlarge Sounds good. Too bad I lack motivation. Currently, I'm perfectly content to sit and stare at a blank wall.

The Photon

@DavidKessner Don't worry, it's almost 5.

David Kessner

@ThePhoton Yup. I'm filing my expense report from my biz trip last week, then I'm (mostly) outta here.

jippie

21:47

@angelatlarge You write the question and @DavidKessner won't be able to resist answering it ;o)

I'll see the result in the morning

good night guys

David Kessner

@jippie Jerk! :)

Kortuk

@jippie Night

rawbrawb

there have been some seriously wacky questions today. We now have someone who wants to make his own IC's at home and is wondering were to start.

0

Q: Some general questions about electronic engineering

so I am interested in EE specifically for the desire to want to design and develop integrated circuits. I want to both design, develop, theorize, and implement the electrical pathways, components, and process of microfabricating, and creating ICs from scratch for hobbyist test printed circuit boa...

reverse-engineering

The Photon

@rawbrawb Does he have an adequate supply of HF and a vacuum pump?

rawbrawb

@ThePhoton my brita filter will stand in for DI water won't it?

David Kessner

21:54

@rawbrawb If only TS were here to answer that one! :)

rawbrawb

@ThePhoton I think I'll exercise the better part of valor and scram before I answer and get snarky.

David Kessner

@rawbrawb Answer it. I like snark.

rawbrawb

@DavidKessner on or off his meds?

David Kessner

Too bad the mods don't like snark. :(

@rawbrawb Off! Some questions deserve his particular brand of incoherent babbling.

rawbrawb

@DavidKessner True that! I never thought of it that way, but you're right!

Kortuk

21:57

seems off topic, right?

-1

Q: How to configure an off-grid generator & battery power-source

Background I have a construction trailer that does not have access to utility power. I generally use the trailer for 2-3 hours/day. Occasionally I'll spend the day working in it. Starting the generator is inconvenient when I only need to work for a short time. I'm considering augmenting it with ...

switches generator ups

rawbrawb

@Kortuk arduino?

David Kessner

@rawbrawb LOL

rawbrawb

no wait, no wait...... Gumstix!

David Kessner

@Kortuk Consumer electronics. Off topic.

Kortuk

@rawbrawb !!!!!!

@DavidKessner my thought, it is so long and detailed I felt like I might be missing something

David Kessner

21:59

@Kortuk The Unibombers manifesto was long and detailed too.

Kortuk

@DavidKessner Ha. It also seems not constructive, "Can I do this better" is pretty open ended.

rawbrawb

@DavidKessner looks like sneaky mod pants went and closed the home IC dude down before I could snark away.

Transcript for