bodger: xkcd android girlfriend arc weld cherry stem (Default)
I got an email hint today reading "The LM741 is useful for many beginner circuits but if you are having a lot of trouble getting it to behave try upgrading to a more modern TL072." This makes sense, as far as it goes, but the 741 dates from 49 years ago, and the TL072 is only slightly more modern, introduced 42 years ago.

These parts are known as "op-amps", short for "operational amplifiers". The concept has been around for even longer than these parts, and is an integral* part of a wide variety of circuits, even today.

It is true that the (slightly) newer TL072 is better in pretty much every aspect than the 741, which was one of the early single-chip implementations, and had fairly poor performance in most categories. However, for many years, it was good enough, and well-known enough that many manufacturers offered it, it was made in enormous quantities, and was dirt cheap.

For a while, people would use the cheap ubiquitous 741 wherever it would do, and the nicer and more expensive TL072 where the better specs were required. Later, the TL072 was the go-to chip, and the more expensive 5532 (1977) would be used where a better chip was required.

Time continued to march on, and now the 5532 is cheaper than the older TL072, and a good choice for whenever you need an op-amp. It's overkill for a lot of uses, but these days it's cheap and ubiquitous. The audiophile crowd pooh-poohs it, and likes to swap in fancier chips to get the sonic character they want.

However, in moderate bandwidth analogue electronics, things are simply not obsoleted as rapidly as they are in the digital arena. All of the chips I mentioned are still made, and easily obtainable. There are, of course, more modern chips out there. Overlooking the bewildering variety of special-purpose chips, there are some newer general purpose op-amps to be had. The OPA627 came out in 1989 (a mere 28 years back), and has some truly impressive specifications. However, it costs nearly fifty times as much money as a 5532. The AD823 appeared in 1995 and is a fine chip, and more affordable than the OPA627. However, perhaps you want to use a chip introduced in this century for some reason. And sure enough, a successor to the OPA627, the OPA827 was introduced in 2006, only 11 years ago.

* pun intended: one common use for an op-amp is a lash-up known as an "integrator"

bodger: xkcd android girlfriend arc weld cherry stem (Default)
My fondness for vacuum tubes extends to cathode ray tubes (CRTs), which were ubiquitous for decades in television sets, as well as technical equipment like oscilloscopes. I like the look of them in operation, an electron beam painting glowing lines on a phosphor screen. Accordingly, I'd like to make a CRT-based art project. To do so, I need to come up with an appropriate power supply. Like other vacuum tubes, CRTs need both a heater supply and a high voltage supply. However, CRTs need higher voltage than most vacuum tubes. An ordinary small CRT generally needs a thousand volts to operate. Additionally, CRTs need several different voltages to operate their various electrodes.

The first order of business is to come up with the high voltage itself. When vacuum tubes were common, high voltage power supplies were too. These days, electronics are solid state, and run on low voltage. For a long time, 5 volts was the norm, and now voltages are getting even lower to support both devices with smaller geometry and reducing power consumption and its attendent heat generation. 3.3 volts was popular for a while, and newer devices run on 1.8 volts. In this arena, parts to produce high voltages are uncommon. As I like to share my designs, I'd prefer to use current production parts that other people can obtain fairly easily and cheaply. Happily, there is a current source for high voltage transformers. LCD screens need backlights, and one popular technology for backlights is the "cold cathode fluorescent light" (CCFL). These are long thin tubes that are lit by (aha!) high voltage. Since they're so common, the associated high voltage power supplies and the parts they're built from are also common.

I found a nice design at tubetime that used a CCFL transformer in a lashup using a voltage divider to sample the high voltage, a voltage reference for comparison, an operational amplifer (op-amp) to compare them, and a power transistor to control the CCFL circuitry. I figured I'd breadboard it and see how it performed. The CCFL transformer itself, while common, is still a specialty part. However, complete CCFL modules are, inexplicably, cheaper. Eyeballing the schematic showed that the core of the design was the same lashup used in the CCFL supply. All I had to do was add a rectifier and reservoir capacitor to convert the high voltage AC output to the high voltage DC I wanted, just as was done in the tubetime design. A quick breadboard showed that it worked as desired.

However, the tubetime documentation explained that the design was derived from a Jim Williams application note, so I read it and saw that Williams' original design didn't use an linear analog feedback loop like the tubetime version, but employed a switching voltage regulator chip instead. This was appealing, as it would be both more efficient and have a smaller parts count. The switching regulator chip replaced the voltage reference, op-amp, and power transistor with a single part, and added some nice protection circuitry as a bonus. I decided to build that version. In the process, I modified the CCFL supply slightly by cutting a trace to separate the low voltage ground from the high voltage ground. I did this because the low voltage "ground" wasn't actually ground, but the switching transistor, and I didn't want to ground an 800 volt signal through the low voltage switching power supply transistor.

The rest of the writeup (with pictures)... )

bodger: xkcd android girlfriend arc weld cherry stem (arc weld)
I decided to buy SparkFun's $10 Heaterizer XL 3000 Heat Gun to try for heat stink and desoldering duties. I'm getting more and more opportunities to remove surface mount parts from boards, and this looks like a more cost-effective approach than a surface mount rework station with a bunch of nozzles for different size and shape parts. Hey, it's only $10 and the instruction manual is a hoot! Seriously, give it a read.

I gave it a try today )

I am quite satisfied with this little beast. It does a lovely job at melting whatever I point it at.

bodger: (Kim Possible)
A worry with the laser cutter is fumes from the cutting condensing on the focussing lense, leading to all sorts of badness. Additionally, cutting materials like acrylic is improved by having an air stream to blow the melted plastic away. Accordingly, I decided to add air assist to the laser cutter. In short, this consists of adding a nozzle to the cutting head, and running an air line to it. Lindsay Wilson did a similar modification.

First, I had some shopping to do )

Since the new head will require a new mirror bracket, and another method to hold objects to be cut at the correct height, I haven't yet given the new air assist system a tryout. I'll need to realign the optics once the new mirror mount is fabricated and installed, then I can give the new air assist system a test drive.

bodger: girls in photo booth (photobooth girls)
Now that I have the EiBot board running the steppers, I'd like to have it turn the laser on and off too. Normally, the Egg-Bot this board was designed for has a servo to raise and lower the pen. This is controlled with a pulse width modulation (PWM) output. But I needed an on/off signal to feed to the laser. Looking at the documentation, I noticed references to a solenoid output, which sounded like just what I needed. Looking at the earlier versions, I could see the circuitry to drive the solenoid, but that circuitry isn't present on the board I have. Looking at the schematics, the I/O pin that drives it is now used for another purpose. I downloaded the firmware source code and tried to dope out what the current code did, but I wasn't able to figure it out. So I sent a note to the designer and asked him about it. He sent back a prompt and helpful reply, stating that the RB4 output should have the solenoid signal. I soldered a header on the appropriate pins and hooked up an LED (and current limiting resistor). At first, I had the LED hooked from +5V to RB4, but that voltage isn't supplied by the board when it's running from just USB, so no joy. So I hooked the LED between RB4 and ground, and sure enough, it followed the pen up/pen down state.

Sweet! )

bodger: xkcd android girlfriend arc weld cherry stem (arc weld)
The controller board that the laser cutter came with is designed to connect to a PC with a parallel port. I planned from the outset to replace this. The board I chose to use instead is the EiBot Board, a derivative of the UBW (USB Bit Whacker) board. This board was designed by Brian Schmalz for the Egg-Bot, a little robot for plotting on eggs, lightbulbs, and other small round objects. It's available from Evil Mad Science and SparkFun.

My interest in it was as a stepper motor driver I could control via USB from any modern computer (it appears as a serial device, and it is controlled by sending it simple ASCII commands).

The trick is to connect the EiBot board to the laser cutter )

bodger: (What terrible problems you have)
I didn't get a reply to my email for a couple of days, I assume because they were celebrating the New Year. Then I got a note reading:

1. NOw we comfirm your machine is 110V

2. Please check the laser tube has the broken or not inside.  and then tell us, because we think it is mostly the laser tube broken

Much apprecaite

I explained that the laser tube was intact, and I was convinced that the laser power supply was broken. They replied:

Can you please return the problem laser power supply and we will send you a new one


So I removed laser power supply from the unit, and figured it was worth a look before sending it halfway around the world for replacement. Since the supply appeared totally dead, not even the fan running, I suspected the problem was something pretty basic, and perhaps something I could fix.

So I took the cover off the power supply, and found this:

It's unplugged!

The power supply had an internal connector that was unplugged! It is a locking connector too, so it had never been fully engaged at the factory. I plugged it in firmly, buttoned up the supply, re-installed it in the cutter, powered it on, and the fan spun! Thus encouraged, I fired up the cooling pump and hit the test button. Nothing. Maybe the tube doesn't strike at minimum power. I turned up the power a little and tried again.

It worked! )

I did a few other tests, and determined that the laser was emitting plenty of power (even unfocussed, the beam burns through a sheet of paper immediately).

I wrote the manufacturer and explained that I had repaired the power supply and wouldn't need a replacement.

Next step is to replace the control electronics with something more useful. The existing controller only works with nasty ancient software on an MS-DOS PC with a parallel port.

bodger: (Bap)
I decided to buy a cheap Chinese laser cutter, with the intention of replacing its electronics so I could use it with a Macintosh. It is intended for use with a DOS PC with a parallel port, and comes with DOS-only software. I asked the vendor if it was available without the software, and found it would be only $40 cheaper. The vendor also stated the unit would not operate without the software. Yeah, right.

I elected to buy it from the eBay listing, even though it cost me $40 more, so I'd have the eBay feedback process to use as leverage if the unit was DOA. I told them I preferred the engraving software, as that seemed like something I would be able to resell. I also told them I wanted a 110V unit.

When the unit arrived, it showed up in two boxes completely covered in yellow tape, pretty banged up by EMS/DHL. The small box contained the smoke fan, which was a 220V, 50Hz unit (oops). Its motor was also loose on its mounting. It also contained the software and some clear plastic tubing.

The large box contined the laser cutter itself, double boxed with styrofoam spacers. It also contained the (crushed) smoke vent pipe, an aquarium pump (for cooling the laser), a UK style power cord, two UK-US adapters, and a combination "surge suppressor"/ adapter with no ground prong.

The laser cutter was in fairly good shape after its journey halfway around the world. The orange plastic viewing window was cracked (no big deal, I planned to replace it with a clear one, the orange plastic is of no benefit for a 10.6µm laser, it just looks cool or something). There was a loose screw rattling around, and the fuse had fallen out.

The laser tube was intact, but the water cooling channels had some cruft in them (so much for "use purified water"). The water tubing also had random cruft in it.

I hauled it down to my lab, rounded up a bucket and US power cord, hooked everything up, and turned it on. The fan spun up, and the steppers zipped to their home positions. So far, so good. Then I pressed the "laser enable" button, and the "laser test" button. No result. No glow in the tube, no current on the current meter, nothing.

So I had a good look at the innards. The unit has two power supplies, one for the logic (and fan), and one for the laser. The logic supply had a sticker warning "make sure switch is set for the right voltage", and pointing to one end, where there was no switch.

I tried powering it up with the electronics access door open, and could see that the laser power supply fan wasn't turning. I got a voltmeter and determined that the laser power supply was indeed getting power. I also measured its 5V control output, but that only yielded 0.69V. Looks like either the laser power supply is a dud, or perhaps it's a 220V one.

The unit was packed with a polite note requesting that I contact them if there's anything wrong, not to leave negative feedback. I'd do this anyway, naturally. So I sent them email detailing what I'd found. We shall see.

I got back a response:

Dear friend

1. YOu machine is 110V verison and also all the accessary is 110V. please do not plug to 220V or it will damage

2. can you please check the back of the machine and see it is mark as 110V or the 220V

I replied that the machine is marked 110V, the laser power supply still doesn't work, and asking what to do next. No reply yet. I have found the power supply manufacturer's website.

bodger: (Kim Possible)
With all the snow on the ground, I decided not to go in to work. I tried working from home, but there was a limited amount of stuff I could really do. After that, I figured I may as well get something useful done with the free time.

When I bought my cell phone, it was the latest and greatest. But it didn't support tethering (using the phone as an internet connection). There was a software update for all the other phones to support tethering, but none had appeared for mine. There was some payware stuff, and I'd tried the trail versions on the train in Massachusetts, but they didn't work very well.

So yesterday, I decided to try one of the open-source ones. There wasn't a working binary for my phone, but I had the Android SDK up and running and figured I'd compile it myself for my phone. This took a fair amount of research and tweaking, but I finally got a clean compile. I rounded up the appropriate cables and my phone and uploaded it. While I was uploading, the power failed (this tends to happen when there's interesting weather afoot). I didn't know if the process had completed or not.

I rounded up the laptop and tried pairing the phone with it and connecting to the internet. I didn't know all the right parameters for this lashup, and things had changed somewhat since the last time I'd done this (I have no idea what Bluetooth DUM and PAN are, for instance). But I tried various things until I got a connection. It took a while to negotiate, but hadn't failed, so I let it be. I went to make a sandwich, and then curled up with a book by the window to eat.

When I got back to the laptop, it had made a connection. So I tried firing up AIM (a utility that doesn't use too much bandwidth) to test it out. It connected, slowly, but I didn't see anybody I knew on. I figured my ISP might be having connectivity issues too, and didn't worry about it. Then I got a message. I didn't recognize the sender, and figured it was just a scammer, but had nothing better to do, so I accepted it. Yeah, looks like junk. In Korean. But a few seconds later, I got another message. Usually the garbage messages send one URL then disconnect. This wasn't a URL, and didn't disconnect immediately. Probably someone trying to talk to someone with a similar AIM ID.

I tried using the Systran widget in my Dashboard (much less bandwidth than firing up a web browser and using Google Translate or somesuch) to make a short message in Korean, explaining that I was probably not the person they wanted. I cut and pasted the result into the chat window and sent it. I immediately got back several lines of text I couldn't read. So I cut and pasted that into Systran, and asked it to translate Korean into English. No joy, the symbols stayed unchanged. Maybe it wasn't Korean? I can recognize many of the scripts used around the world, and it did have that blocky, phonetic look like Korean, but closer examination revealed it was something else.

Oho! A mystery! The sort of mystery my language-loving friends like [profile] ayasdollz and [personal profile] dcseain would enjoy. Too bad I couldn't reach them. I tried launching the Facebook app on my phone, but it crashed as usual (I really wish Froyo could come out for this phone, I suspect a lot of things would work better).

I sent some English text, and received a few more lines back that I couldn't read. Maybe it was someone's secret code?

I cut and pasted some into some cryptography software I'd written and did the basic analysis on it. Yup, looked like language, character distributions were consistent with a phonetic language, and several short words occurred more than once. Didn't look like English, however.

With nothing better to do, I chose a few of the short words, pasted them in the chat window and sent them.

There was a bit of a delay, then a brief response. Also short words. Probably useful, but I was out of my element. So I sent a few short words in English.

Another pause, then a very brief response that looked to me like punctuation. Were they asking for more of my text to analyze?

I cut and pasted a couple of simple children's stories and some stuff I'd written, chunk by chunk.

Eventually, I got two foreign words back. I guessed it was "hold on" or "that's enough" or somesuch and waited.

I got back "what is". Okay, sure looks like they're trying to use my language, but don't have a lot to go on. Are they asking what language I'm using, or what? I sent back "I am speaking English."

After a little while, I got back "send more". So I started grabbing whatever text files looked likely and sending them, piece by piece.

Then I got back "Where speak English". English is pretty well known around the world, but maybe they just want to know where I'm sending from. "I am in the United States."

Another long pause, then "Not hear of", followed by "Sun up".

Were they telling me the sun is up where they are, or asking if the sun is up where I am? Sort of like the old Star Trek ep where the Horta writes "NO KILL I", with similar ambiguity.

I replied "The sun is up here. It is snowing. Where are you?" There was no reply to this, and I realized that while they can apparently puzzle out what the words mean, they don't have a workable phonetic mapping, and can't tell me something like "We are in uhzbekeestan" or whatever.

So I sent "What time is it where you are? It is 2:30PM here."

No reply. And the phone battery is running low. Apparently, the local cell site is down and I'm burning transmitter power talking to a more distant site.

I had no idea how distant.

bodger: xkcd android girlfriend arc weld cherry stem (arc weld)
My latest idea to discourage (in the sense of extirpate) stinkbugs is to deploy UV-C lamps in the attic. These are low-pressure mercury arc bulbs that emit a bunch of mercury lines from blue to the far ultraviolet, with special glass or quartz envelopes that don't block the shorter wavelengths (253.7nm and 184.9nm). I'm hoping the bugs are attracted to the light and are then attacked by a combination of nasty UV (which wrecks biological molecules) and ozone (which also wrecks biological molecules, and will vanquish the residual stink). If this doesn't attract them, I'll add some phosphor (BaSi2O5:Pb or SrB4O7:Eu) to downconvert some of the UV-C to UV-A, which is known to attract the little buggers (it's the same phosphor used in bug zapper bulbs).

pic of apparatus )


Jan. 9th, 2011 10:24 pm
bodger: xkcd android girlfriend arc weld cherry stem (arc weld)
So it seems [personal profile] werewulf wanted an appropriate-looking custom sound player for her Cassandra setup.

She found a cheap toy at a thrift store that looked likely (pictures) )

bodger: Morbo demands an adorability scan! (Morbo adorability)
Let's set the way-back machine1 2-3 years ago. I had two TiVo units (a Series 2 and a DVD recorder) that were starting to act up. From the symptoms, it was very likely the disk drives (these run 24/7 in TiVo service, are the only moving part, and by far the most likely item to fail in a TiVo). I looked up the part numbers on them (both were 80GB units), found the model with the lowest price online, and bought a pair of them.

When they showed up, I grabbed a pair of USB-PATA2 adapters, pulled the drive out of the Series 2, and hooked it and one of the new drives to the computer. I used the Unix "dd"3 command to just make a verbatim copy from the old drive to the new one. It didn't fit. It turns out that not all "80GB" drives are alike, and the one in the TiVo had a little more storage than the new one. Grump. So I pulled the drive out of the DVD recorder (which was the identical model drive), and copied that one instead. The copy completed without issues, I popped the new drive in, and it worked fine.

but what about the other TiVo? )

1[profile] fizzygeek and I just re-watched my old laserdisc copy of the original Tron movie tonight. And it struck me that I'd just used that phrase from the film (I wrote that text earlier today). Hee-hee!
2PATA: Parallel ATA, as opposed to the more current SATA (serial ATA) drives. Also known as IDE or EIDE.
3The "dd" command is the "convert and copy" command. But since the C compiler was already called "cc", it got called "dd", in typical Unix harebrained naming style.

bodger: xkcd android girlfriend arc weld cherry stem (arc weld)
Given my history of dinking with electronics, I've burned every kind of component there is, many times over, both intentionally and accidentally. Because of this, I've acquired the ability1 to identify the burning component by scent. They all smell like burning something, but each has its own specific character. Freshly sharpened pencil scent points to an overloaded potentiometer. A more earthy version is burning Ohmite (resistors). A scorched plastic scent is burning semiconductors. A hot metal/varnish/paper odor is an overloaded transformer. The same, overlaid with oil, is a motor. And a sour, alkaline smell is an electrolytic capacitor giving up the ghost (and its electrolyte).

While playing with the computer tonight, I noticed a worrisome odor in the air. Shortly, I was able to identify it as the scent of a dying electrolytic capacitor. Sniffing around failed to localize it, but I figured it was the computer (which was warm and spinning its fans), its power supply (also warm), the monitor, its power supply, or the external disk drive. I wouldn't be pleased by any of these things popping a capacitor. I also checked the shelf of electronics nearby, but they didn't seem to be the source either. I fired off a backup and shut down all the other stuff, just in case.

Then I stepped out in the hall, and the odor was much stronger. Hmm, the CFL in the hall fixture had flamed out a few months ago, maybe it's the replacement. I unscrewed it and gave it a sniff. Smelt like warm plastic, but not much else. Probably not that, unless the heat was volatilizing spew from the earlier failure, but it seemed a bit strong and sudden for that. Checking [profile] fizzygeek's room, the craft room, the bathroom, and the office yielded nothing useful. But the stairway to the kitchen seemed to point the way.

I told [profile] fizzygeek what was up and she said she didn't smell anything. But when she came out into the hall, she sure did! She checked around too, and agreed that the other rooms probably weren't the source. But what was?

I had replaced the kitchen lightbulb earlier in the evening, but it wasn't a CFL, as they didn't live long in that enclosed fixture. The previous bulb had been a 100W halogen, but when I ordered replacements (they're oddballs that can fit), they sent me 230W frosted ones by mistake. So I had put one of those in. Sure enough, the globe was uncomfortably hot. Our theory is that the spew from the failed CFL had been vaporized by the heat from the monster halogen, tricking me into ignoring anything that didn't contain electrolytic capacitors!

1+1000 experience points!

bodger: xkcd android girlfriend arc weld cherry stem (arc weld)
I'm a fan of bright, high-quality light for reading and crafts, so I figured I'd try the fancy Microsun lamps. These are 68W metal halide lamps, and provide 5300-6000 lumens with a CRI of 72-90, depending on who you believe. The lamps are expensive, but nicely made. I bought two of 'em, a floor lamp (for crafts) and a table lamp (for reading). They work well, and several people admired the quality of light they produce. However, after two years of fairly light use, the reading lamp started to flicker and then died. I tried swapping bulbs, but the problem remained with the lamp.

People who know me will not be surprised by my next move. The ballast (called a "gear pack" by the seller) is clearly marked "DO NOT OPEN. NO USER SERVICEABLE PARTS INSIDE." [profile] maugorn and others might remember me reciting the litany "Do not open, no user serviceable parts inside, pilot lamps soldered in place, danger high voltage, warranty void if opened, etc." as I tore into things.

Well, I found out some things. The manufacturer claims that you can't just make your own lamp by buying a gear pack, because the gear pack doesn't include "the transformer". I had initially believed there was a transformer tucked into the base of the lamp, as the lamp was sufficiently large and heavy. But there isn't one. The AC leads go straight to the gear pack. You could easily retrofit any ordinary threaded-tube lamp by buying a $60 gear pack and screwing it on top.

The ballast is patented. This means I can download the schematic. It's not quite the same as the one used in the lamp, but it's pretty close. It's just a current-regulated DC power supply with a transformer-coupled starting pulse network, run off a voltage doubler. Armed with that info, and the failure mode, I deduced that it isn't the starting circuitry (the lamp failed when it was already started), and it probably isn't the current regulator (those generally work right or not at all). That leaves the voltage doubler. Not much to it, an inrush current limiter, a pair of diodes, and two big electrolytic capacitors. And electrolytic capacitors are famous for drying out and failing. So I read the specs off 'em, measured the relevant dimensions, and found equivalent units on Digikey. The inrush limiter was a house-numbered part, but a little research convinced me that it was a 10Ω, 1.7A unit or thereabouts, and such things aren't critical (it's just there to make sure that the discharged capacitors don't pull too much current when first turned on). I chose a similar unit that would physically fit and ordered it too.

While I waited for the parts to show up in the mail, I considered my other options. These days, low-wattage metal halide bulbs aren't too hard to find, and ballasts can be had from Fulham (a manufacturer whose products I like) and cheaper manufacturers. If it dies again, I can always re-use the lamp body and nice, high-temperature ceramic socket with a 70W (or more!) metal halide bulb and ballast.

May 2017

21222324 252627


RSS Atom

Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags
Page generated Jun. 26th, 2017 08:56 am
Powered by Dreamwidth Studios