USB microscopes for very small SMT

I've been surprised that the lower limit on many-an-Intel-CPU is 1GHz... seems like 100MHz would be plenty for an idling PC. (I realize that modern CPUs often have the registers implemented as dynamic memory, so you can't go down to DC... but still... 100MHz should be plenty...)

Was that coach, business, or 1st class? Are you tall or short or inbetween? :-) I've always found that in coach you're packed in tightly enough that it's kinda difficult to get productive work done... although I do see people trying.

At least a netbook is a good size for a computer if you are going to try to get work done...

---Joel

Back then the good ones would go down to DC. But the RAM didn't and that's why they either had to have an extra arbiter or keep them running at a few MHz. AFAIK the Contura didn't have an arbiter. But hey, 6h out of legacy NiCd ain't bad. With today's LiIon it would probably run for two days. And this wasn't even the Aero version (kind of the first netbook back when netbooks weren't "invented" yet).

Those were the golden days. International business class and sometimes first. Blini, caviar and champagne. So comfy that once I did not wake up until the flight attendants tapped me on the shoulder and the whole cabin was empty. I am over 6ft and some coach class seat arrangements give me a back pain.

Yep. Much bigger than my first laptop. A Wang with 200 vertical pixels, no backlight, no mouse, yet I did large designs on it using OrCad SDT. Almost until the arrow keys fell off. This thing had a built-in modem and printer (!).

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

On a sunny day (Mon, 08 Dec 2008 13:06:07 -0800) it happened Joerg wrote in :

Although that is true, the real question is: How much current does it save to clock down the processor? There is a test from some guy for the eeePC, clocked down all the way to 160MHz IIRC or there about, and it hardly made any difference in power consumption. I installed the clock mod (very simple: unload a module and run something), and the thing became unstable at those low clock speeds. Makes you wonder if the Celeron can be static clocked anyways... So, because of that instability (I like my data) I removed the clock mod again. It had no real advantage, maybe because it is already under clocked at 630MHz.. There are other big power guzzlers in the thing, display, WiFi, switching WiFi off adds an hour or so? And the USB GPRS-HSDPA modem.

Camera people sometimes wear a belt with batteries, maybe some idea?

may have some of them for free download.

--
http://improve-usenet.org/index.html

aioe.org, Goggle Groups, and Web TV users must request to be white
listed, or I will not see your messages.

If you have broadband, your ISP may have a NNTP news server included in
your account: http://www.usenettools.net/ISP.htm


There are two kinds of people on this earth:
The crazy, and the insane.
The first sign of insanity is denying that you\'re crazy.

A lot. I measured it back then but that data is gone by now. The biggest difference was turning off the hard drive which was easy. They already had hibernate and suspend modes back then so you really only needed it shortly before turn-off (if hibernate was chosen) and then it came on automatically. I didn't care because DOS stuff just didn't crash. Data was either stored on a PCMCIA memory card or 3-1/4 floppy. Because I didn't trust hard drives for good reason, lost three in that laptop alone.

Next was the backlight IIRC, third was the processor clock. Now, backlight is easy to handle. The designers of this laptops were very smart, they put some programmable extra buttons on there. So if, say, I got stuck and had to think hard about some equation while writing a module spec I hit a button and the display would blitz off. Later another press of the button and it was back instantly. This feature and the turned-off HD made the processor the number one consumer and that was fixed by their nifty under-clock scheme. A marvelous machine. And then, later, came Carly ...

I have never seen a laptop this nifty during the whole 15 years until a few weeks ago with the latest generation netbooks.

Try to get that past the security guys at the airport ;-)

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

Interesting that the whole netbook movement came from Asus, a Taiwanese company.

Although HP was relatively fast to jump on the bandwagon with the 2133... which wasn't very competitive, although the newer 1000 series are. Try one out at Costco next time you visit... very similar to your NC10... somewhat thinner and lighter, although in part that comes from not having a standard VGA connector, only 2 USB ports rather than 3, a 1.8" drive rather than 2.5", and a 3-cell battery rather than six. (Six-cell batteries and regular VGA cables supposedly show up in January...)

---Joel

A lot of innovative stuff comes from Asia. I remember when the traditional watchmakers laughed and scoffed at the first Japanese quartz wrist watches. They no longer laugh. I would not be surprised if some former execs of the Europa Uhrenwerk in Ulm are now wearing Seiko watches. I was in that building after it became a mini ghost town because a client rented rooms there, dirt cheap. Spooky. My car plus a dozen others in a parking lot designed for several hundred. A humongous array of buildings, 99% empty.

I always knew there was a market for netbooks but the industry slept. Now they'll have their lunch eaten. Reminds me of certain auto manufacturers ...

One company I called about the Samsung told me they can't keep up because these things fly off the shelves like hotcakes. And Samsung knows how to do proper production ramp-up. The only mistake they may have made is to build too many white ones.

3-cell? What were they thinking? The smaller drive and less USB is ok but not having a standard VGA is going to cost them. Those things will be used by presentation jockeys left and right.

Advice to HP and others: Talk to potential _customers_, _not_ your own marketeers. That's how we do it in medical (and caused HP-IVUS to throw in the towel ...) Get some better web designers. Oh, and the Costco web site doesn't even show it under netbooks. That ain't the way S&M works.

If there is no VGA at all via some cheap adapter then the product might fizzle in the marketplace altogether. Anyhow, January they'll have missed two major milestones, section 179 tax year sales and the whole Christmas shopping period. Not that I like the fact that Christmas is being commercialized but it's unfortunately a fact.

Sometimes I wish I could be VP of sales just for a year or so. In many companies that'll probably feel more like a salvage job.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

Multicore processors are better at this, because you can shut down cores that aren't being used. With sub-65 nm transistors, a lot of the power dissipation is due to leakage, so you don't go anywhere close to zero power at zero speed, unless you remove Vdd. (Sort of like NMOS.)

Cheers,

Phil Hobbs

Apparently their marketing guys decided it was, "Less than one inch thick or bust!" -- it's 0.99" thick, and you apparently can't get a regular VGA connector into a case than thin after allowing for some reasonable thickness for the "top" (the LCD panel). Actually, while this is what I've read, I'm not sure it's correct, since it does still have a regular RJ-45 Ethernet socket, which seems about as thick... I suppose the difference is that Ethernet cables fit within the RJ-45 socket completely, whereas VGA cable connectors are larger than the connector.

Anyway, HP chose to use a "DisplayPort" connector instead, so one can have a short pig-tail lead to a regular VGA connector... once they're available, supposedly in January. (I guess they took a page from Apple's playbook here -- their ultra-thin MacBook Air requires a pig-tailed RJ-45 Ethernet connector...) Purportedly the upside of DisplayPort is that -- since it's actually more of a general-purpose high-speed bus and not just a video connection standard -- they'll have docking stations as well that have USB, Ethernet, etc., all going through the same connector as the video. This is something I'll only believe when I see it though, since DisplayPort is still awfully new.

I had a Sony laptop once that had a pig-tailed lead to a VGA connector... I always lived in fear that one day I'd lose it and be completely unable to connect a regular monitor, as this was somewhat before eBay became so well-known and made it easy to obtain such obscure replacement parts in many cases.

I swear that the Costco web site tends to have less than half of what you find in the actual stores on-line... Here's Newegg's link:

---Joel

Heck, even if it comes out the top with a small lid that's better than not having VGA at all.

Personally I don't like proprietary solutions. This is also why I would never buy a netbook with a promise that a 6-cell battery will come out next month or so. Because most likely it will carry premium pricing. I hope they at least furnish a VGA pigtail with the netbook.

That'll be a real pain when you get to the client, all the suits are lined up in the board room to see your solution and you can't find that VGA adapter thang. Then you'd start spooling the PPT file onto somebody else's laptop while the suits with income levels of $10 per minute or more are twiddling their thumbs ;-)

Oh man. So S&M asleep at the wheels again? Our guys sure as hell watched all the major sales outfits that our products were advertized properly. If not it costs maybe 15 minutes to alert a vendor that a link is missing or broken. In pre-Christmas times this is paramount for any computer gear. The return on investment for those 15 minutes spent by a S&M assistant can easily exceed 1000:1. Uncle Leroy looks for a netbook for lil' Jeremy. Oh, they don't have HP. Ah, the MSI Wind looks good as well. Click, click, click, confirm order, done.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

Leveraging some insider information part of the issue is the balance between leakage current and switching current. They seem to need to keep enough switching going to keep leakage from corrupting the processor state. Appropriately mobile rated parts do better.

"iatrogenia dirty hospital", "god complex doctor", "confirmation bias"

It is not hot coffee suits.

This is interesting:

I've been preaching this point for years: write code as if it's meant to be read and criticized by the world, print it on paper, and READ it several times before you ever run it. The savings in time and bugs is an overwhelming payoff.

John

On a sunny day (Tue, 09 Dec 2008 08:36:10 -0800) it happened John Larkin wrote in :

On paper? In the 198X years I did that, had 500 pages of docs on asm I wrote for just _one_ project (all written with pencil). But even back then text editors were so nice. I'd stay away from paper these days. A text search in paper docs sucks.

grep xxxx *.txt

works anytime

You idea of 'being criticised by the world' is not new, that is how open source works, basically. The result is a good OS (Linux for example) and good applications, many people fixing things, improving things.

The problem is that many companies will not want to open their source, they think the better keep the advantage they have, and prefer not being sued by patent trolls.

hehe

Yes. It should be formatted and paged to look good when printed on fanfold paper. During the read-throughs, one makes hand notes regarding code, commenting, and formatting, and edits as a result.

It's simple: the more times your eyes pass over the code, the better it gets. So even a pass that tweaks "cosmetics" is a valuable debugging and optimization function.

The more you automate, the more you adbicate responsibility. Sure, keep the source file open in Crimson Editor, for searches, but review on paper.

The point is to SLOW DOWN and do it right. Programmers keep telling me how much faster their methods are, and then they are still debugging, and shipping bugs, months later.

SLOW DOWN AND GET IT RIGHT.

John

On a sunny day (Tue, 09 Dec 2008 09:09:54 -0800) it happened John Larkin wrote in :

I dunno, there always is a time limit, if not the end of week, then Monday when things should be ready .... But your paper idea has some shortcomings. One big one is interactive developing. For example some project I am working on is not precisely defined, because it has to do with counter measures against undefined attacks on servers. It works very well, but running it, and it outputs to a log file what it does (and it runs 24/7), shows sometimes for example it considers a 'good' IP an attacker, based on behaviour. So I added some stuff for that case. You can read all you want on paper, but that will never give you a real situation. In fact the code is so complicated that you cannot just follow it by reading. For example what a sort routine can do to a list.... of many megabytes.

Code is developed by testing small pieces, or using small tested pieces, and then letting it lose in the real world. Then corrections are made. No paper, there never will be paper docs, but you can run it in 'verbose' mode and it will tell you all functions with their arguments as it runs.

BTW I hear that word 'crash' so many times, my programs do not crash. Last time I had one out there that crashed was many many years ago, asked the person to run it with 'verbose' enabled and dump output to a file. Looked at that file, took 10 minutes to find the NULL pointer and fix it. C is a beautiful language, printf is your friend.

struct posting *lookup_posting(char *name) { struct posting *pa;

if(debug_flag) fprintf(stderr, "lookup_posting(): arg=%s\\n", name);

/* argument check */ if(! name) return 0;

for(pa = postingtab[0]; pa != 0; pa = pa -> nxtentr) { if(strcmp(pa -> name, name) == 0) return pa; } return 0; /* not found */ } /* end function lookup_posting */

You'll like the slogan these guys have:

situation.

This being s.e.d., I refer to embedded apps that run on hardware, not so much higher-level pc/server things. I test code using a background debug pod. Sometimes my first-rev code crashes, usually because of a dumb mistake. If it does, the laptop screen (the pc running the BDM program) shows me all the registers and the contents of the stack; the offending opcode address is right there in plain sight. I can also run code to a breakpoint, and single-step, and see all the registers and selected data (in little windows) at every step.

John

the use

moisture, a

teach

so the

Which is precisely my point.

People don't want to have to decide which computer to use based on what they think they might be doing. They want it all. What happens if the job changes *after* you've left home? Do you buy multiple copies of all your tools? Keep multiple copies of all your documentation? ...up to date? No thanks (decided not to buy an eeePC for this reason), I'll carry my one full size laptop with me even though I'm limited to 2-3 hours on the battery.

You're going to let engineers in the back room make SKU and inventory decisions?

There is very good reason. It doesn't save much to have it running at half that. If you want to save significant power you have to turn things *OFF*. Then it takes time to get them back.

Utter nonsense.

His was a desktop, with legs. Yours a "lite". Different design point.

Intel will screw up a wet dream.