mutt you@domain.com -a /home/username/backups/archives/mysql_backup.$suffix.tar -s "MySQL Backup"

to get e-mail backups working. In reality, this command will merely open up mutt and create a new message but it will not automatically send it without additional input. Buried online I found someone else who said to simply:

mutt you@domain.com -s "MySQL Backup" < /home/username/backups/archives/mysql_backup.$suffix.tar

which does automatically e-mail you, but it just dumps the file into the e-mail body without attaching it as a document. What that statement is really doing is piping in body text and completing the e-mail prompts magically, which isn’t what I wanted to do. The solution is to:

mutt you@domain.com -a /home/username/backups/archives/mysql_backup.$suffix.tar -s "MySQL Backup" < emailbody.txt

What this gives you is an attachment of your file, and then it pipes in an e-mail body (put it in the same location that this script is running from) and sends the message. You would also use echo statements to “type” whatever additional input you wanted in mutt and script the rest of the message completion [untested but it conceptually should work, YMMV]

Full Script:

#!/bin/bash
cd /home/username/backups/
mkdir mysql
suffix=$(date +%y%m%d)
mysqldump --opt -uUser -ppass -h mysqlA.domain.com db_nameA > mysql/db_nameA.$suffix.sql
mysqldump --opt -uUser -ppass -h mysqlB.domain.com db_nameB > mysql/db_nameB.$suffix.sql
tar -cf archives/mysql_backup.$suffix.tar mysql/*
mutt you@domain.com -a /home/username/backups/archives/mysql_backup.$suffix.tar -s "MySQL Backup" < emailbody.txt
rm -r mysql/


Sector67 Forum Thread

I’ve gotten a few questions as to what you’re looking at, if you’re completely unfamiliar a visit to Wikipedia should clear things up.

It’s made out of 2″ x 2″ extruded aluminum channel, the railed looking stuff that makes up the top of the arm and the drop for the spindle (dremel) is called 80/20, it’s fairly expensive but is very precise and saves a lot of time for some projects.

The Z axis is the classic over constrained parallel rods, I pulled the rods out of a pair of identical printers and then drilled and reamed to size the end blocks and the slider block, which is just a chunk of delrin plastic. The bad part about using a plastic block for a sliding surface is that you have a drill/ream it over significantly to get it the right size, which in my case just means it’s sloppy and will be replaced by linear bearing when it becomes the weakest link in the machine for precision or stiffness.

The bottom end X-Y axes are THK precision linear bearing. I picked this out of discarded equipment at the university, which I understand it not available to “normal people”. I found out a few years after I picked it up that it’s $0.43/millimeter (yep, that’s a half of a dollar per the thickness of your fingernail, it’s expensive). The bearing carriages are ~$130/each. I was fortunate enough to find a pair of 18″ linear track and a pair of bearings still in their well oiled bags! They do have lash in them and should certainly be run in parallel or with a precision guide track for stiffness/rigidity but for my application (milling light plastics and depositing plastic in an additive process) it’s not a major concern this early in the project.

The other common question is how much does this thing cost. So far I have $199 for the full generation 3 Makerbot electronics (thanks Zach et al., they work great!) and $4.11 for 3 x 1″ long 1/4-20 nuts that are used on each of the axes (True Value). The rest has been collected or given to me by others.

Got to go out water skiing on the PerBlue corporate yacht, for my first time I wasn’t too bad:

Electronics vendors:
Digikey (Excellent search function, expensive but very good pictures, lots of products too)
Mouser (lots of what digikey has but no pictures! Great for Delphi / Amphenol connectors)
Newark (bigger weirder stuff, careful some stuff ships from Farnell in the UK)
Jameco (Search is terrible, but if you can’t find it anywhere else. . .)
Allied Electronics (Cheap amphenol. . .)
Waytek (cheap for bulk connectors and wire, next day shipping)

Arduino:
Adafruit (cheap and usually has what I need, I’d rather pay a little more to support a still growing startup)
Sparkfun (expensive but a good one-stop-shop)

Electronics Surplus:
All Electronics

China Direct Electronics:
Futurlec (Cheap, good customer support, responsive, shipping takes a while – as indicated up front)
Deal Extreme (holy cheap, lots of people on eBay/discount stores are buying and drop shipping from them)
Sure Electronics (Lots of great cheap parts, you have to add them to your cart and check shipping after each item though, they’re making up some of their costs by charging shipping for each item – I received everything 3rd day after I ordered it though, which is amazing)

Materials:
McMaster-Carr (Expensive, ships quick, stocks a lot, great pictures/details on products, search rocks)
Grainger (Search sucks, plumbing stuff/etc much cheaper than McMaster, if you can find anything. . .)
MSC Direct (Tooling, industrial consumables)

Tools
Harbor Freight (If you’re on a budget you can’t get better than HF tools, but make sure you read the reviews)

Turns out The Brothers Brick finally found my post on converting plastic/modern Lego tracks into the 9V tracks that I posted about a while ago.  Thanks for the traffic and enjoy your stay

reset – grn
tx
rx
vcc
gnd
blank – blk

The words grn and blk are silkscreened on the board to denote which pins are which.  You could of course use a RS232->TLL converter to program this as well, but using the Duemilanove is really convenient as it pulls the reset pin low for you.

On the topic of this Pro Mini board, I had a few e-mails back and forth with Sparkfun trying to figure out what the actual maximum input voltage for the board is, Sparkfun’s Website lists it at 12VDC max, the EagleCAD schematics list it as 16VDC, and the datasheet for the voltage regulator list it as operational maximum 16VDC, with an absolute maximum of +/-20VDC before damage occurs.

Also noted is that the oscillator on this board is absolute crap, I have a timing critical application that I intended to use this board on, using a stopwatch and a Duemilanove and this board, both running identical software and hardware configurations:

15hours 37 minutes elapses, the Duemilanove is exactly in sync with the stopwatch still, the Pro Mini has gained 1 minute and 1 second. I ran this test several more times with identical results, the stop watch and the Duemilanove are in sync, with the Pro Mini gaining time each time, always about 20 second per 5 hours elapsed time, which is not acceptable in my application.

I swapped the Sparkfun crystal/resonator for one of my own resonators and ended up with the same crappy timing. I finally switched out the resonator for a crystal and a pair of capacitors tied to ground and was able to clean up the timing mostly, it’s still gaining a second for every 40 minutes it runs for, it was previously 15 so it’s improved, but not ideal. I’m going to ditch using the Sparkfun board for my own layout on the next iteration.

DIAdem helpfully gives you a list of available commands, but doesn’t tell you what they display or how to use them.
Program Command: @@ChnPropGet(No,Property name)@@
Formatting: @@ChnPropGet(1,”Description”)@@
Output: The 1st channel’s Description, other input could be Name, Unit, etc
Notes: Make sure to use quotes around the Property name and remember that the channels are counted from the very first channel onwards and found as “Number” in the channel information.

Command: @@CD()@@
Formatting: @@CD(1)@@
Output: The 1st channel’s Units

Command: @@ChD(Line,No)@@
Formatting: @@ChD(2,1)@@
Output: The 1st channel’s second data point value

Command: @@CL(No)@@
Formatting: @@CL(1)@@
Output: The 1st channel’s number of points (length)

Command: @@CNoXGet(Channel group number,No)@@
Formatting: @@CNoXGet(2,1)@@
Output: Finds the 2nd group’s first channel and returns it’s channel number (counting from the first channel.
Notes: Could be considered translating from relative position of a channel (in a group) to absolute position in the list of channels.

Couple of general notes, you can’t nest lookups (at least not in the text boxes, so it would be invalid to:

@@ChnPropGet(794,"Description")@@

and turn it into a lookup:
@@ChnPropGet(@@CNoXGet(15,1)@@,"Description")@@

Otherwise it’s pretty helpful to use these when labeling graphs/etc so you don’t have to constantly correct things.  DIAdem also builds in lots of string functions like trim/etc, so you can clean things up if necessary!

<WordPress>  – I had all these formatted nicely in a table but WordPress ate itself trying to render it, hence the oddball list, sorry. </WordPress>

MAX6675 TC Board Schematic and Board (and Panelized)

While we were walking through a storage area they pointed out a real version of the Hunt for Red October’s caterpillar drive, in reality called a “Magentohydrodynamic drive” which uses massive magnetic fields to induce flow in water.

Pictures of the advanced beamline that they put materials/experiments in the way of to blast them with x-rays:

Here’s a picture of the accelerator used to get the photons up to warp speed. (99+% c )

And I managed to find evidence of the infamous government’s Sector 7 as noted on this bicycle:

Finally, for being one of the most advanced research facilities in the world, they sure managed to keep things looking fresh out of WWII on the exterior, here we can see some of the outdoor exposed infrastructure leaking like a sieve, good thing it’s just steam:

First key bit of information is that HP / Agilent are absolutely spectacular companies for providing service manuals, product manuals, and datasheets well beyond the retail life of their instruments.  You can still pick up documentation directly from Agilent here.

In digging through the datasheet you can see that a FAIL 023 means we’re having a problem with the dc offset, so I printed off the relevant service information (groups K,J) and prepared for the worst.  After powering the machine on at my desk I was again met with the FAIL 023 when requesting a sine wave, and a new error was a FAIL021 when requesting a square wave.  I then out of chance cycled through the other 3 settings (triangle, etc) and returned to the square wave and sine waves to discover that the errors had went away!  It’s likely that I had a stuck/hung up relay and switching away from square and sine waves cycled something into place again.  We’ll see how long it lasts but hopefully I’ve just picked up a nice piece of equipment.

Additionally while I was reassembling the covers, turning the instrument from side to side resulted in a nice pinball sound, I swung it back in forth until I was greeted with 3 washers on my desk, I can’t imagine that these were helping with things but they’re out now.  Looks like they came off the main power transformer which someone had helpfully duct-taped into place at some point.