The Smell of Molten Projects in the Morning

Ed Nisley's Blog: Shop notes, electronics, firmware, machinery, 3D printing, laser cuttery, and curiosities. Contents: 100% human thinking, 0% AI slop.

Category: PC Tweakage

Remembering which tweaks worked

  • Capacity Test For USB Flash Drive Memory

    Centon 4 GB USB Flash Drives
    Centon 4 GB USB Flash Drives

    It’s always a good idea to verify that a USB flash drive works and has its rated capacity, even when you buy them from a reputable vendor.

    The easiest way to measure their capacity (quite different than measuring battery capacity):

    • Create a monster file of random data
    • Copy it to the drive
    • Verify that the copy matches the original
    • Delete the copy

    That doesn’t verify that you can successfully create a bazillion little files, but it’s a good rough-and-ready check that you haven’t gotten, say, a 2 GB drive mis-labeled as 4 GB. It could happen…

    Assuming you’ve deleted any shovelware (these were clean) and that the drives are now empty (as these were), find out how big they claim to be:

    df /media/ed/CENTON\ USB/
Filesystem     1K-blocks  Used Available Use% Mounted on
/dev/sdb1        4107284     4   4107280   1% /media/ed/CENTON USB

    Pour /dev/urandom into a file that will fill the available space (not the total space), which will take several minutes:

    time dd bs=1K count=4107280 if=/dev/urandom of=/tmp/test.dat
4107280+0 records in
4107280+0 records out
4205854720 bytes (4.2 GB) copied, 450.883 s, 9.3 MB/s

real	7m31.162s
user	0m0.712s
sys	6m54.166s

    Copy it to the drive, using rsync with a progress indicator:

    time rsync --progress /tmp/test.dat /media/ed/CENTON\ USB/
test.dat
  4205854720 100%    8.45MB/s    0:07:54 (xfer#1, to-check=0/1)

sent 4206368202 bytes  received 31 bytes  8772405.07 bytes/sec
total size is 4205854720  speedup is 1.00

real	7m59.035s
user	0m24.490s
sys	0m17.433s

    Verify that the two files match:

    time diff /tmp/test.dat /media/ed/CENTON\ USB
real	3m32.576s
user	0m0.588s
sys	0m6.268s

    Then delete the file:

    rm /media/ed/CENTON\ USB/test.dat

    Repeat as needed for the other flash drives, using the same test.dat file. All these drives worked; one subsequently caught a disease at the library.

    And, yes, one of them is noticeably darker; four of the others seem lighter and five darker gray. Most likely, the cases came from three different anodizing batches and, I suppose, if I were to pry them apart, the innards could be radically different. Ya never know!

  • Capacity Test For New UPS Batteries

    Just got a quartet of 12 V 7 A·h lead batteries, prompted by a big Belkin UPS that instantly shut down during a power blink. It needs only two batteries, but the shipping was the same for two or four and I’m sure the spares will come in handy.

    A stiff 2 A discharge test shows that SLA batteries really don’t like high currents, which is exactly what they must provide in a UPS:

    Rhino SLA - 2013-01
    Rhino SLA – 2013-01

    The capacity is barely 4 A·h at 2 A, not to mention that I’m using a conservative 11.4 V cutoff.

    The two batteries with the highest capacity also were the closest matches, so they’re now in the UPS.

  • Epson R380 Printer: Cutting In a Continuous Ink Supply System

    The continuous ink supply system on the Epson R380 printer developed a slow air leak in one cartridge, which may have contributed to the nozzle problems, so I just installed another system from the usual eBay supplier: prefilled with ink and $30 delivered.

    As nearly as I can tell, Epson designed a number of features into the R380 specifically to thwart CISS installation, including the awkward bridge across the middle of the printer that interferes with the flat tube feeding ink to the flying cartridges. I managed to route the previous CISS tubing around the bridge, but this time I figured enough was enough.

    So I tucked a shop rag inside the printer, put a vacuum cleaner nozzle near the operation, and applied a fine-tooth pull saw to the bridge:

    Epson R380 - bridge removed
    Epson R380 – bridge removed

    That certainly simplified the rest of the installation…

  • Why Friends Don’t Let Friends Run Windows: Conficker

    Mary gave a gardening presentation at the local library, popping a 4 GB USB memory stick with the presentation into a library computer connected to the display projector. Back home, she deleted the presentations and was about to add more files, when she noticed something interesting:

    drwx------  4 ed   ed    4096 Dec 31  1969 ./
drwxr-x---+ 3 root root  4096 Jan 31 19:21 ../
-r--r--r--  1 ed   ed   59288 Mar 21  2009 autorun.inf
drwx------  3 ed   ed    4096 Jan 30 19:31 RECYCLER/
drwx------  4 ed   ed    4096 Jan 31 19:10 .Trash-1001/

    Ubuntu 12.10 automagically mounts FAT filesystems with the current user as owner and group. The .Trash-1001 directory is the Linux trash heap, but where did all that other stuff come from? The autorun.inf definitely looks Window-y, doesn’t it?

    Perforce, the library runs Windows, but that shouldn’t add files to a USB memory stick that just was plugged in and used for a read-only presentation, should it?

    Huh. You know where this is going…

    Let’s hand autorun.inf to VirusTotal for a second opinion. The first three results from their long list confirm my suspicion:

    Antivirus Result Update
    Agnitum INF.Conficker.F 20130131
    AhnLab-V3 Win32/Conficker.worm 20130131
    AntiVir Worm/Kido.IH.40 20130131

    The executable file containing the actual payload is, of course, buried in a subdirectory that might look more innocent on a Windows box:
    /RECYCLER/S-5-3-42-2819952290-8240758988-879315005-3665/

    It sports a randomized name to evade a really stupid malware detector:
    jwgkvsq.vmx

    Here’s what VirusTotal reports from some heavy hitters in the AV field:

    Kaspersky Net-Worm.Win32.Kido.ih 20130131
    Kingsoft Worm.Kido.ih.(kcloud) 20130131
    Malwarebytes Worm.Conficker 20130131
    McAfee W32/Conficker.worm 20130201
    McAfee-GW-Edition W32/Conficker.worm 20130131
    Microsoft Worm:Win32/Conficker.B 20130131

    The Wikipedia article gives the details. I suppose that PC got it from somebody else’s USB stick, but the library really should be running some defensive software; Conficker dates back to 2008, so it’s not new news these days.

    That kind of Windows Genuine Advantage makes up for all the hassles of running Linux, right there. Mary reported the problem to the library; we’ll never know the rest of the story.

    [Update: We got an update!]

  • Anonymous 5 Axis Parallel Port Breakout Board Pinout

    Parallel port breakout boards of this ilk run about $14, complete with cable, on eBay:

    5 axis parallel port breakout board
    5 axis parallel port breakout board

    The PCB has no part number and the inferred URL isn’t productive. The “driver CD” accompanying it has doc for every possible board the vendor might sell and, absent a part number, the file names aren’t helpful. An exhaustive search suggests it corresponds to the HY-JK02-M 5-axis interface board manual.doc file.

    Despite any implication to the contrary, the board does not have optoisolators between the parallel port pins and the outside world. The stepper driver bricks should, but the input signals from limit switches and suchlike connect directly to the guts of your PC.

    This overview (from the manual) shows the physical pin layout (clicky for more dots) and reveals the hidden silkscreen legend:

    HY-JK02-M Breakout Board - overview
    HY-JK02-M Breakout Board – overview

    It looks like the board I got added a spindle relay driver transistor, plus a few resistors over by the manual control connector on the right.

    Notice that the fourth terminal on each axis is GND, not the positive supply required for the optoisolators on the 2M415-oid driver bricks, which means you can’t just run a section of ribbon cable from the breakout board to the brick. You’ll need a separate +5 V (or whatever) power supply wire for each brick, with a common return to the system ground for this board. Those terminals are firmly bonded to the top and bottom ground planes on the board, so there’s no practical way to re-route them.

    The small switch in the upper left, just to the right of the parallel port connector, selects +5 V power from the USB port (which has no data lines) or the power connector in the lower left. The LED near the switch won’t light up until you have both the parallel port cable and the USB cable plugged in.

    The doc includes a timing diagram with no numeric values. I established that it can’t keep up with a 500 kHz pulse train and seems content at 100 kHz, but that’s conjecture. Setting the timing to match whatever the stepper driver bricks prefer will probably work. The diagram suggests the setup and hold times for direction changes are whatever you use for the minimum time between step pulses.

    This shows the functional labels:

    HY-JK02-M Breakout Board - function labels
    HY-JK02-M Breakout Board – function labels

    The parallel port connector output pins, sorted by function:

    Pin 9 1 2 14 16 3 7 8 6 5 4 17
    Function Spindle
    motor    		 Enabled X step X dir Y step Y dir Z step Z dir A step A dir B step B dir

    The parallel port connector input functions, sorted by pin:

    X -Limit Y- Limit Z- Limit A- Limit Emerg Stop
    10 11 12 13 15

    The table uses Chinese for Pin 15: 急停.

    It’s not clear whether the pins on the manual control connector are inputs or outputs, nor what the three separate Enabled lines do:

    P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 P11 P12 P13 P14 P15
    B step B dir A dir Z step Y step X step X dir Enabled 5V/VDD 5V/GND A step Z dir Y dir Enabled Enabled

    The three white connectors in the middle drive an LED readout board that’s probably most useful as a DRO for CNC-converted manual mills using the pendant for positioning.

    The small white connectors duplicate the functions of the green screw terminals. They’re probably useful in a small machine that I’m not building.

    This isn’t the board I intend to use in the final setup, because I need far more I/O pins, but it’ll serve for the short term.

  • Kill A Watt: IEC Adapters

    I should have done this a long time ago:

    Kill-A-Watt - IEC plug and socket
    Kill-A-Watt – IEC plug and socket

    It makes measuring PC power consumption much easier!

    I picked up some cheap AC plugs and sockets, cut a short IEC extender cable in half, and wired ’em up. If the IEC extender link breaks again, search amazon.com for something like “computer power cord extension” and rummage around.

    IEC color code hint: brown = hot, blue = neutral (unless they cheat).

    US NEMA 5 plug / socket hint: the blade marked W is neutral. More expensive hardware will have dark brass = hot, light brass = neutral, but don’t bet your life on it.

  • LibreOffice 3.6: Fixing Font Selection Problems

    This may not be a LibreOffice problem, but that’s where it shows up: the font selection dialog won’t display fonts with nonstandard Style names. There is, of course, no documentation anywhere (that I can find, anyway) on what Style names are permitted, so you discover this only when a font style that’s properly installed and accessible by other programs (like, say, Inkscape or Scribus) doesn’t render properly and doesn’t appear in the list.

    In Xubuntu 12.10, LibreOffice 3.6.2.2 can’t handle the American Typewriter font style called Medium, which is what I’ve been using for the return address field on my (very few, these days) mail envelopes. Over the years, various versions of OpenOffice and LibreOffice have alternately accepted and rejected the Medium style, so this isn’t exactly a regression. It is, however, Yet Another Annoyance.

    The solution, hinted at in that thread, involves using FontForge to rename the offending Style to, say, Regular, then saving the font. It’s actually the Weight property, hidden in Element → Font Info → PS Names tab. In this case, I changed the word “Medium” in the Fontname, Name for Humans, and Weight fields to “Regular”, which also updates the values in the TTF Names tab.

    I save the modified font files in ~/.local/share/fonts using TrueType format, just to be sure I don’t confuse them with the original Postscript version in /usr/share/fonts/custom, delete the original, and then run fc-cache -v -f to update the caches. This surely isn’t the cleanest way to make it happen and almost certainly isn’t allowed by the Adobe EULA I agreed to, back when I actually bought the fonts, but so it goes.

    And then It Just Works…

    LibreOffice vs American Typewriter font
    LibreOffice vs American Typewriter font