I have a number of small motors whose bearings need occasional lubrication. Paper shredders, impact printers, and tape libraries contain mechanisms that require lubrication with a light oil. Hunter original ceiling fans, assorted locks, hinges, and small electric motors require SAE 10W straight weight, non-detergent oil. Good luck finding it locally! Home Depot, Lowes, and Sears stores have driven out the small hardware stores that used to carry this stuff.
The popular “3-In-One” brand oil in tbe red, black, and white squeeze can is about 20W, but it contains detergent, which is bad news in most applications in which you can’t flush, filter, or drain the old oil with the dirt particles held in suspension. Those particles will abrade bearing surfaces in simple machines.
I found exactly what I needed on Amazon. It’s La-Co 4 ounce Zoom spout oiler 79704G. It contains four ounces of pure 10W non-detergent oil. It’s packaged in an easy to use squeeze bottle with a tiny extendable tube for precise application. Cost is about $4.00. It’s sold by Pandora’s OEM Appliance Parts. It was delivered to my door within a couple days of my order.
I’ve worked with integrated circuits (I.C.s) since the 1960s, but haven’t been involved in their manufacture — only their application.
Today’s integrated circuit manufacture is a high stakes capital intensive business whose players use trade secrets to maintain their market advantage. I’ve never been inside an I.C. “fab” (factory), so it was a treat to find an hour-long presentation by an industry manufacturing engineer on YouTube. The technologies used at nano dimensions are mind-boggling.
Here’s the excellent presentation, in full:
The speaker mentions that lithographic imaging of the mask is now being done at 193 nanometer (nm). As you can see, we’re well above visible light and on our way to x-rays(!). Here’s the electromagnetic spectrum in that region:
The presentation is aimed at the layperson and is filled with surprises. For instance, one gigabyte of semiconductor memory can be produced on a flat substrate within the diameter of a human hair. I give it two (gloved) thumbs up.
I’ve retrofitted solid state drives (SSD) with SATA interfaces to a few Windows computers recently. Since SSDs use flash memory, which have a finite number of write cycles, we want to reduce the number of disk writes to our new SSD.
Within the computer’s BIOS, change the disk interface from IDE to AHCI. (In Windows XP, edit the registry first to use the AHCI disk driver.) Disk reads and writes will be faster with AHCI.
Within Control Panel / System / Hardware / Advanced / Performance Options / Advanced / Virtual Memory stop writes to the swap file by disabling virtual memory. Click the “No paging file” radio button, followed by the “Ok” button. Make sure that the host PC has plenty of memory — at least 4GB and preferably 8GB — as the operating system won’t be able to swap code/data out to disk anymore.
I’ve read recommendations that with SSDs, the System Restore facility be disabled. I’d rather retain the security of system restore points. I’ve also read that on SSDs, FAT32 partitions be used rather than NTFS partitions, the theory being that the FAT32 file system entails less writing per file. True, but NTFS partitions are more robust than FAT32 partitions, so I’ve retained the NTFS partitions.
I recently replaced the hard drive in a three-year-old mini-tower PC that had spent its life on an office floor. It was a routine job with the expected dirt and dust inside the case. No vacuum was available, so I used the shop’s compressed air and blow gun to blow out the dirt, then reassembled everything. When I started the PC, it howled a loud screeching/ticking noise. I opened the case, removed power from the hard drive, powered on the PC, and listened with a stethoscope. The noise was coming from the power supply — probably its fan.
I removed the power supply, disassembled it, and discovered that the power supply fan’s 12 Volt DC supply wires — maybe 28 gauge, were rubbing against the spinning fan blades. A jet of compressed air through the power supply case perforations must have nudged the wires into the fan blades. I dressed the wires so that they were a quarter-inch from the blades, reassembled everything, and the PC worked quietly. Time lost to troubleshoot and repair this dumb problem which I had caused: 1 to 1.5 hours.
Use a vacuum — not compressed air — to remove dirt from inside PCs.
In the late 1960s I was part of a small product development team whose goal was the design of a frequency-synthesized radio receiver. This included a phase-locked frequency synthesizer whose output could be set to anything from 156 to 186 Mhz in 100 Hz steps. It used two VCOs (Voltage Controlled Oscillators — their output frequencies vary as a function of steering voltage). The output frequency of each VCO was determined by a phase-locked loop. It was the first time that our employer had developed a phase-locked loop frequency synthesized product. The frequency synthesizer’s high-speed VCO covered 30 MHz in steps of 1 MHz. We distributed the 30 MHz range amongst three oscillators, each of which covered a 10 MHz range. While prototyping, we discovered low-frequency noise sidebands on the VCO’s output signal. We needed to remove these noise sidebands, or they’d appear in the receiver’s output. On a spectrum analyzer, the noise sidebands appeared to be 20 to 400 Hz each side of the output signal and 15 dB below output signal level.
After selectively freezing circuit components while observing the noise sidebands, we learned that the carbon composition resistors in the loop filter were (one of) the culprits. That was a surprise: I had always thought of resistors as discrete inert lumps. I learned that carbon composition resistors consist of tiny grains of carbon bound to tiny grains of ceramic. We were seeing noise caused by random molecular motion. These voltage spikes, though mere microvolts in amplitude, were modulating the VCOs, resulting in noisy sidebands on the VCO’s output signal.
We substituted carbon film resistors, and discovered that the noise sidebands dropped substantially. I learned that carbon film resistors can dissipate less power, but offer a more contiguous medium than ordinary carbon composition resistors. Once we’d peeled back this layer, we discovered that the ceramic feedthru filter capacitors that fed each oscillator with DC power were also modulating the high-speed oscillator.
Instrumenting this circuit was interesting: we couldn’t load the phase locked loop itself with instrumentation or the phase lock would fail; we could only examine the noise sidebands that were an effect of noisy phase locked loop components. Since those days, I’ve learned that often we can’t directly examine phenomena — we can only see indirect effects. I suppose that this is how astrophysicists have discovered dark matter. Dark matter doesn’t reflect or refract electromagnetic energy; it’s detectable only indirectly because it’s affected by gravity.
What do we ever know about anything?
B.F. Skinner was alternately praised and despised because his brand of psychology, which became known as behaviorism, admitted that it’s impossible to know the inner workings and hidden mechanisms of humans; we can only observe behavior. He had a point: are humans less mysterious than electron movements or dark matter? In the end, what do we know about anything or anybody other than its behavior?
I just watched the PBS American Experience documentary, Silicon Valley. I loved it! It begins at legendary Bell Labs in New Jersey where William Shockley, John Bardeen, and Walter Brattain invented the transistor in 1947. Eighty-two minutes later, the film concludes at Intel in California, with the introduction of the first microprocessor, the 4-bit 4004 CPU in 1971.
Here’s how PBS describes it:
An eye-opening look at the birthplace of the modern technological era told by the people who shaped it, Silicon Valley is a fascinating reminder of how Robert Noyce and his team of trailblazers led the way in transforming California’s Santa Clara Valley into a worldwide hub of industry and innovation, and laid the bedrock for modern technology.
While working at a customer site, I needed to quickly join a laptop PC to their LAN (local area network). The customer produced a yellow Category 5 (“Cat 5”) patch cable to connect the laptop PC to an open port on a switch. After a few minutes of wondering why the laptop wouldn’t join the LAN, I closely examined the cable’s ends. It was a crossover cable. It wasn’t labeled “crossover”; I guess that its manufacturer thought that its yellow jacket color was sufficient to identify it as a crossover cable. (Yellow has become the de facto standard jacket color for crossover cables.)
I recommend that organizations keep no yellow-jacketed straight-through Cat 5 patch cables on site. Reserve yellow jackets for Cat 5 crossover cables only. You’ll reduce troubleshooting time.
In 1945 John von Neumann reported on the electronic digital computer in his famous First Draft of a Report on the EDVAC. His architecture is still used today. It included a shared “store” for both instructions and data. A shared store makes efficient use of memory, which was an expensive resource. However, it allows poorly written programs to halt unexpectedly when the CPU’s program counter (often called the “instruction pointer”) lands on a word of data, instead of a valid instruction. Most CPUs halt when they attempt to execute an undefined instruction; to a user, the computer has “frozen” or “locked up”.
Since today’s memory is cheap, why not break the shared store into two stores — one for instructions and the other for data? That will prevent the CPU’s program counter from pointing to data, thus eliminating one source of computer crashes. This segregated memory computer architecture might be appropriate for mission-critical applications involving human life. One downside is that it will consume more power than the von Neumann model. How many crashes would it prevent? 10 percent? 90 percent? I don’t know.
Of course programmers will continue to write buggy code that will result in other forms of crashes.
Update, 3 December 2012: I’ve learned that a computer with separate stores for data and instructions is said to use the Harvard Architecture.
The Internet knits today’s society together, yet most of us aren’t familiar with exactly what comprises “The Cloud”. Last month, NPR’s Terry Gross interviewed Andrew Blum, the author of a new book, Tubes, A Journey to the Center of the Internet. The book explores the hardware infrastructure that instantly transports data across the globe. Ms. Gross asks the questions that any layperson would ask, and the author replies with amusing stories of his adventures inside the Internet’s data centers, points of presence, repeater huts, and cross-connect centers.
I found Mr. Blum’s descriptions to be easy to understand. I’ve worked inside similar facilities and don’t think that I could describe their components so clearly.
Much of the Internet is built atop older telephone and telegraph infrastructure. (Likewise, American highways are built atop the trails that were blazed by Indians a thousand years ago.) Fiberoptic cable often shares the conduit, cable trays, and trenches where 100-year old lead-sheathed oil-impregnated paper-insulated copper cable still resides.
Click here to read the brief article and/or listen to the 25 minute interview.
Simply connecting my cellphone wirelessly to my netbook proved to be much harder than it should be.
Last week I bought my third Samsung SGH-T439 cellphone. It’s a 2007-vintage product: a very small flip-phone. I’m fond of the product, so when I found a great deal on eBay, I pounced. I learned on howardforums.com that Samsung has a program (called PC Studio 3) that allows a Windows PC to exchange files with cellphones such as mine.
Since I was using a Windows XP netbook with a Bluetooth adapter, I thought that I’d try connecting the netbook to the cellphone via Bluetooth. It sounds easy . . . and it could be, except that multiple hardware and software vendors’ products need to communicate nicely and securely with each other.
After failing several times to get this multi-layer system to work, I carefully documented the interface of each layer: function, ports, protocols, passwords, and eventually got it to work. The task required at least an hour . . . and I understand most of the technology and vocabulary! The layperson would have little or no chance of making this setup work.
It’s at times like these that I appreciate the rationale of Apple’s single-vendor approach.
The supply of hard drives has been dramatically affected by Thailand’s recent floods.
Flooding in Bangkok, Thailand-October 24, 2011: This is a satellite image showing flooding near Bangkok, Thailand. Locals are parking their vehicles loaded with belongings on overpasses on highway ramps to protect them from being affected by flood waters. This stretch of highway is the Sapan Nonthaburi Ban Bua Thong Highway # 3215 where it intersects with the Soi Chai Uea Highway #307.
The recent flooding in Thailand has caused dramatic shortages in the hard drive market. Many hard drive manufacturers have major manufacturing plants in Thailand. Western Digital and Seagate facilities are especially hard-hit. The shortages have resulted in big jumps in hard drive prices. Hard drive prices have risen 20 to 40 percent recently. I noticed that one Western Digital 1.5 TB drive jumped from about $120 retail to $180 retail in just one week.
This is not a good time to upgrade or replace your hard drives. Wait a few months for the pipeline to fill up with new drives, and we’ll see hard drive prices fall again.
Thailand is not the only country that manufactures hard drives. Malaysia, Singapore, and China are major producers as well.
Industry observers predict that hard drive shortages will persist into 2012 and cause the prices of finished systems to rise. It’s expected that as the floods subside and Thai hard drive production resumes, prices will fall and stabilize roughly where they were a month ago. When will this happen? Maybe 2nd quarter, 2012.
StorageMojo, in Google’s Disk Failure Experience by Robin Harris, did a great job of summarizing the results. The article points out that when a disk’s SMART reports
the disk is failing and should be replaced before it dies completely.
For most users, the most relevant conclusion is that a disk is more likely to fail as its power-on time reaches 3 years and more. I usually recommend replacing working disks when they have 40,000 hours or more of power-on time. (There are 8760 hours in one year.) As usual, your mileage may vary.
If your computer won’t start, try removing USB devices.
I was engaged by a New York-based VAR (Value Added Reseller) to replace a server in a busy Miami environment. The symptom was that the server (only a few months old) wouldn’t boot. The VAR’s tech support personnel had remotely diagnosed the server and concluded that it needed to be replaced. The new server had been drop-shipped to the customer site, so I just needed to copy the data from the old server to the new server. I removed the drive from the old server and copied the data to the new server within an
hour; the new server started fine. Then the VAR’s technician logged on remotely to the new server and fine-tuned the VAR’s application.
The VAR had provided an external hard drive, connected via a USB cable, for routine data backup. We plugged it into the new server, and while it seemed to run okay, it made some odd squeaking noises for about 30 seconds. I tidied up the site as the VAR’s technician tidied up the software. When he tried to restart the new server, it wouldn’t boot. Neither of us liked the sound from the external hard drive. Turning off the external hard drive and removing its USB cable from the server allowed the new server to boot just fine. The USB-attached external hard drive wasn’t playing nice with the server, so the new server refused to boot. The technician asked the customer to return the old external hard drive and ordered a replacement external hard drive.
Moral of the story: When trouble-shooting a computer that won’t boot, don’t assume that all of its USB-connected devices are okay. Temporarily unplug any unnecessary USB devices and then try restarting the computer. It may just work!
p.s. I’ll bet that the old server was fine. We just needed to replace the external hard drive. D’oh!
Additional 1 GB of memory is worthwhile addition to MSI Wind U100 netbook.
The MSI Wind U100 netbook is a handy little computer. Mine was delivered with Microsoft Windows XP Home Edition and 1 GB (gigabyte) of memory. I’ve used it for at least six months and recently added 1 GB of 200-pin SODIMM DDR2 memory, bringing the netbook to its maximum of 2 GB of memory. I bought the memory module from Crucial Technology for about $24.00. I like Crucial’s ability to examine your computer, determine its memory configuration, and recommend upgrades. When you use Crucial’s Memory Advisor, Crucial guarantees that its memory will work in your computer.
The procedure to add memory to this netbook is just a little more involved than adding memory to most laptops: remove the battery charger, the battery, and the nine tiny Phillips head screws that secure the bottom half of the case to the top (don’t lose the screws! They tend to fly off into space.) There is at least one video on Youtube that illustrates the procedure.The netbook now seems to run a little smoother, with less hard drive activity. Its tiny Atom CPU ensures that this netbook will never be a rocket ship, but it’s now an even better bang per cubic inch portable tool.
On many site visits, I carry a MSI Wind U100 netbook. It’s a great tool for this task: it gives me a tiny portable Windows XP-based computer in known clean condition. I use it to scan disks, check wired and wireless connectivity. etc.
Two weeks ago, it began to shut down its display backlight, seemingly at random. I gathered my watchmaker’s tools and opened it to have a look. A video clip on Youtube demonstrated a procedure to replace the dubious heat conductive tape on the CPU and GPU with heat conductive paste. I followed the procedure; the netbook runs MUCH longer now. I seem to be moving in the right direction. My next step is to install an aftermarket heat sink (product review).
Use software to cool your MSI netbook In the meantime, the tiny WindFan utility displays the CPU temperature and how fast the fan is spinning. You can press the “Emergency Cooling” button to increase the fan’s speed. The CPU temperature varies by 20 degrees Fahrenheit, depending upon how hard it’s working.
Once this temperature issue has settled down, I’ll probably install Backtrack 3 on a new Linux partition, together with a boot menu such as LiLo. It’ll give me a full suite of network troubleshooting tools.