Have you ever thrown a shopping cart out of a convertible at 80 mph to observe the sparks? How about hooking up a power amplifier to 115 volts AC just to see how it craters? Your fellow engineers have done these things and more! If you have a great picture of totally destroyed hardware, or a photo of a blown circuit, send it to webmaster@microwaves101. If it gets on the web site you will receive a free Microwaves101 key chain pocket knife! Impress your friends, if not your boss!

Note to mortuary contributors: please consider that your boss may not find your submission in the best interests of your Big Company. Lately we've been getting a lot of "please remove my submission" emails, try not to send us anything that you might regret.

Note to Big Companies: Don't blame us for posting your spectacular failures, we only post what your employees send us. On the other hand, don't rip off Microwaves101 pictures for presentations without permission. Maybe it's time for some training!

Note to mortuary fans: we put the new stuff here at the top, but the older disasters are still here too--just scroll down. Also, in many cases, if you click on the pictures on this page you can see higher resolution images.

New for March 2010! These came from David, who offers this simple explanation:

Someone took the back feet off a Tektronix scope and then used too-long screws to put the scope back together. And then sent it to our Calibration Lab to be calibrated.

Bummer! The perpetrator deserves to be promoted to manager so he/she can't do any more damage...

Also new for March 2010! These came from Colin, who had a little problem with heat. Come to think of it, without heat, this mortuary would be significantly reduced in scope...

I thought I'd share a couple pix taken of a voltage regulator that was dissipating a few more watts - into the highly ineffective circuit board instead of a heatsink. There was a 5A fused "protection circuit" that failed to blow the fuse, and of course, it didn't protect the regulator, or anything downstream. A 12V supply was feeding about 4A to equipment under microprocessor control, and a single fuse was used for the load as well as the 5V logic - so when the logic failed and drew about an amp, that was still fine with the fuse (4A + 1A = 5A) but the regulator was now dissipating 12-5=7V * 1A = 7W - without a heatsink. The microprocessor normally drew a few milliamps.

New for February 2010! You can thank John for these SEM images. They are very high-resolution, be sure to click on them to see destruction at the micron level!

Thought I'd send along a couple of pix of deceased IC's I took at work on our scanning electron microscope.

The first image is of an IC that had way too much current passed through it. We expected to see just the usual destroyed substrate but instead were surprised to see this nice little fiber of silicon oxide that grew out the unfortunate device. The device became hot enough to actually oxidize the silicon and cause the fiber to grow instead of just explode. Shades of the movie "Alien".

The second image shows a semiconductor from a power supply that we "unfortunately" managed to destroy by accident one day in the lab. We decided to cut open the deceased part to have a look. Not much left of it here, just this destroyed micro landscape. As I recall the part was something like a voltage regulator IC in a TO-3 package.

Also new for February 2010! This picture came from Mario, and illustrates what can go wrong when you don't "match the load to the generator..." Thanks! Hopefully we can all stay friends!

While on vacation in Europe a friend of mine plugged my battery charger into an outlet using an adapter but no transformer. Apparently 220V was more than the poor charger could take. After hearing a loud pop and opening the charger up, the damage was obvious. What wasn't as immediately obvious was the damage done to the house's electrical system...

New for January 2010! These images came from Karol, in Gdansk. Nice Christmas colors!

I'm an Electronics student from Poland and I have a few pics for your mortuary - please see the attached files.

The first one is a cratered FPGA chip. It was some kind of measurement equipment and this damage is caused by a thunder storm. The second picture shows a few other chips on the same PCB as the xilinx chip. In the blue circle you can see a laser burn on the CCD sensor of my camera :( Be careful when taking pictures of working lasers. On the third and fourth photos you can see a power supply module from the same equipment.

Fifth and sixth picture shows a self made probe for measuring something on a 3-phase power line. The resistors got hot and melted the caps. It was a design fail because that kind of resistors are supposed to get hot. It's quite surprising that the caps still work (I mean there's no short or open circuit and they still act as a capacitor).

The last picture shows a blown IGBT brick. I got it from trash so I don't know what has happened to it.

Thanks, and sorry for the delay caused by our overflowing spam bucket! - UE

New for December 2009! This came from Daniel:

I have an exhibit for the Microwave Mortuary. It's an AC input filter (220 volts) from a 300W Ku-band amplifier. The power connector turned out not as weatherproof as it claimed to be, so some rain water got inside and turned most of the filter into smoke, tar and charcoal.

Let guess... removing screws voids the manufacturer's warrantee! - UE

New for November 2009! These two video links came from Isaac:

I see this kind of work a lot, but this takes the cake. This is an $18,000 radio someone put up with zip ties and bailing wire. Makes me wonder why they are always having trouble with this path.

 

Ever get mad at your test equipment and wish you could do this with it??

New for November 2009! These photos also came from Ben:

There are lots of different antennas on the market for different applications ranging from simple Hertzian dipole, patch, Yagi, horn to expensive dish antennas. They can be made of a simple cable, build on a PCB, or even integrated on silicon.

We found a versatile antenna with reasonable efficiency and very low cost. The antenna was not found by 3D EM simulation or theoretical calculations, but by research in the field. We don't know the directional reception pattern yet, but we still are doing research on this topic. We called it the fork antenna, simply by its shape. I discovered it in Gori, Georgia, Central Asia. I actually don't know how it really works, but this is no fake, I watched television on this TV set with this antenna. Maybe somebody could check in the US if this type of antenna also works with the new digital TV.

New for October 2009! This photo came from Ben from Deutschland, who is becoming a regular contributor!

Hey, I killed a chip and made a photo of it. The chip is die and wire bonded to a PCB, some bondwires are gone because the board was on my desk for half a year and didn't work anymore. The photo was made with a Nokia Handycamera through a microscope, the quality is really good for that. What happened to the chip?

I designed a circuit including an amplifier and wanted to control the biasing current, which is a frequent concern in RFIC research. I made two current input pins. In the lab, I turned off my setup, the supply voltage and RF sources. But I forgot to turn off the current sources. When pulling the current plugs, the current source went to the voltage limit. Unfortunately this was 40V, a bad idea for an IC. The ESD protection and a part of the circuit exploded (I even could see a flash) and melted (both marked in red). The fallout is distributed all over the chip. So be advised: Never use current input if you want to tweak your biasing current. Always use voltage input biasing control pins!

Also new for October 2009! This picture came from Paul:

We had a path fading. No foul weather and we couldn't see a problem with the antenna or feed horn from the ground with your binoculars. See what we found on the feed horn when we had the cable crew closely inspect this feed horn. You had to get to within five feet before the corrosion was apparent. The pressure of the corrosion pushed the brass in till there was a short from side to side. Salt air, this was 100 yds from the ocean.

And also new for October 2009! This picture came from Benjamin from Slovenia. This is an example of what happens when you short out 230 volts at 60 amps.... yikes!

We accidentally connected the switch so it made a short when switched to ON. You wouldn't like to be there when it happened... We had a New Year's Eve. My co-worker (he switched it on) got minor burns and luckily that was everything.

New for September 2009! This photo came from Ben from Deutschland, who also provided us some advice on microwave career killers:

I've got something for the microwave mortuary. It's not so much microwave, but more power supply related.

What to do if your power cord is broken? Replace it? Not necessarily! Just cut the connector, skin the wires, plug them into the socket and fix it with the connector. Maybe you have to turn the power off before you plug the wires in! I think you also should not touch the top of the connector.

That's sage advice, especially when you consider that this is a European outlet providing 240 volts. Thanks, Ben!

New for May 2009! This picture came from Paul, showing the remains of a beefy amplifier. Thanks!

This is (what's left of) the input power filter network in a Techron 3020 300V, 200A power amplifier for a MRI system - one of a 3-channel set. All three channels were fed from a common DC rail fed by a full-wave rectifier directly from a 3-phase AC supply. Since it's a pulse amplifier with quite a lot of electrolytics, there's plenty of reserve to draw from. The only circuit protection was the panel breaker. This channel blew up with a shotgun-like blast when one of its output MOSFET modules shorted. The module itself disappeared into plasma, along with a good chunk of its heatsink structure. The date? 9/9/99. I shoulda guessed something bad was going to happen. We never did find the root cause, but figure it must have been poor heatsinking and device overheating. We ran these right at the hairy edge, so didn't even attempt a repair. We replaced them instead with much larger 700V/350A amplifiers.

New for April 2001! These pix came from Dave from Down Under, where beer does flow and men chunder! Thanks!

Some pix of a CCTV DVR board from a computer. I noticed that several power rail caps on the PC motherboard had also bulged but not blown. I'm not sure if it was the PC power supply that killed the DVR boards or the other way around. Unfortunately I couldn't keep the boards for a memento as the client wanted them for a insurance claim. Both DVR boards had blown in the same place around the V-regulator area.

New for March 2009! These images came from "anonymous ham enthusiast" who might have blown his cover as well as his amplifiers. Thanks!

Class C amplifiers and and black powder DO mix!

This is the product of a bet between myself and a radio geek friend of mine. I had an ailing, home-brew, ham fest special, 220 MHz amplifier that had an arcing trimmer cap. After cleaning up some of the "welds," the problem persisted.

To be honest with you, I didn't get much deeper into the troubleshooting process. A fellow radio geek came over and spotted a can of "powder" on my "other" bench. The devil's grin appeared on his face immediately after I told him what I was working on. He bet me that the spark from the cap. wouldn't be enough to trigger the mixture. He was wrong! Enjoy.

New for February 2009! Here a video of an acorn stash, inside a microwave antenna. Anyone know the dielectric constant of acorns? Sent to us by Kyle! Sorry we couldn't figure out a way to embed the video, just click the link below to view it.

http://www.ebaumsworld.com/video/watch/80540471/

New for January 2009! These photos came from Richard... the picture on the lower right would make a scary computer "desktop", click it twice to see the non-airbrushed details.

The first picture is a transfer panel that was missing a 4 inch seal which the birds took advantage of. The flood pictures were from this year in Franklin, In. during the great flood of aught8. There was 25 inches of water inside the building and it sits on a 30" high pad. The last and certainly least is one of my fellow technicians when we restored traffic.

Also new for January 2009! This image came from Mike:

You know how the data sheets on TO-case power resistors strongly suggest the use of high quality thermal grease? Ever wonder what happens when your tech doesn't bother? Bustication!

Mike also contributed an example of an unstable amplifier (example #4), thanks Big Guy!

New for December 2008: Here's a destroyed radome from a National Weather Service radar in Reno Nevada that got caught in 140 mile per hour wind gusts on December 19, 2008. Sent in by Matthew (thanks!) For the full story, visit their web site, there are a lot more images than this one! If you are looking for weather radar images anywhere near Reno Nevada, for now you are out of luck! In the future, the CASA project will populate the US with thousands of low-cost weather radars, and this type of single-point failure will be a thing of the past.

New for December 2008! This burning dish came from Jim.

Here's one you might want to include on your site. This picture was taken sometime in the early 80's in Dalton, Georgia. A tech was inside the site doing an inspection when he smelled something burning and the radios started squelching. He just happened to have a 35mm camera with him. Click on image for a closer look!

Hot path

New for November 2008! These images of a toasted circulator came from Happy_Trails...

This circulator had 2 frequencies running through it —> 10.9925 GHz & 10.9915 GHz with 296W at a 50% duty cycle. The only problem is that nobody put heat sinking brackets on it, therefore causing it to go to a couple hundred degrees. I now call it the Technicolor circulator.

Click the images to see some giant, high-resolution photos!

New for October 2008! Here's a pdf document that shows some toasted transformers in Southern Cali, sent in by Bill. We're not sure where the photos originally came from, but we'll leave them in the original document and maybe someone will fill us in on the back story. Thanks, Bill!

New for October 2008! More pictures from Marcelo of Buenos Aires:

And some people worry about asteroids... Apparently, it fell down during a particularly windy day. I´m attaching a couple close-ups of the mounting pole's base, there you can see the tiny fisher anchors that were used to fix the radio. No wonder why it was blown away...

New for September 2008! These pictures came from Laszlo.

We had some technical difficulties and malfunctions, off-line links -- attached pictures show ice problems and one of them shows our hangman just for fun :)

Tower is located at Budapest, Hungary, 350m hilltop. Pictures taken in January, 2004.

By the way, these antennas worked just fine, after they were de-iced...

Also new for September 2008! these pictures came from Marcelo in Buenos Aires.

These are two pictures of an Ericsson Mini-Link E indoor unit that landed in my repair shop. The nasty looking coat you see spread evenly all over the mainboard is the good old mix of dirt and water commonly known as Mud.

Pity I don't have a picture of the repair guy's face when he removed the cover.
He had this "about to puke" look when he came asking "I guess this is beyond repair, right?"

Note from UE - why not send these boards for a ride in the dishwasher before attempting the "repair"...

New for August 2008! This story came from Adam...

I recently had a serious problem with my self-built main CAD machine (Athlon
64 x2 @ 3.02 GHz, 8 GB of DDR-2 800 RAM, 1 TB RAID 10 array... You get the picture). I lost a drive on the RAID array, and the rebuild of the array kept failing... Then lots of bluescreens... Then lockups if I ran more than 3 of the 4 sticks of RAM... Totally bizarre, and I kept replacing things and testing individual parts in other machines trying to find the source of the problem. Weeks later, frustrated and now weeks behind schedule on several custom motorcycle parts projects, I was no closer to a solution. Until I accidentally did something that shouldn't be done, but gave a result far out of proportion with what should have occurred...

Apparently what had happened was that one of the two +12v rails in the power supply had failed --it was now grounded rather than providing +12v, although I never opened the power supply to find out how the problem had come about, exactly. The way I "diagnosed" this was while I was still mucking with the RAID drives, trying to get my precious work out of the remainder of the array (one drive had a dead controller card and was known DOA -- don't know if this was the chicken or the egg in this failure). Well, one of the drives was a "crossover" drive, a SATA 300 drive that had both a SATA and a Molex power connector, with a strict admonition against connecting both at the same time. This shouldn't have caused anything catastrophic... but in my case, I apparently connected the grounded rail to one of the connectors, and the working rail to the other, causing the full rail current to pass through the drive's controller card and out the grounded rail. I caught the burning component smell immediately and yanked the power cable out of the back of the machine, but the deed was done. I gather this was a "sacrificial" power resistor (capacitor? Hard to tell...), but that didn't save the drive coils for the read/write heads... Hoping the drive was still good and only the controller card bad, I tried a known good controller from an identical drive... And ended up with no smoke, but no joy, and two dead controller cards. At least I found the source of my problem (as well as the drain, no gate included!)

Ice damage on an air conditioner housing . . . And, not broken, but have to say, "here is a photo of a big pair of cans."

Saftey notiss

Useless antenna

Here are a few pictures of some unfortunate pieces of hardware. Although, in their life, they were not, strictly speaking, microwave hardware, but they were close support. High voltage power supplies for TWTs, to be exact.

 

This photo shows what happens to a 10 Amp 115V relay, attempting to interrupt about 500 Amps of (short circuit) current.	The scarred area in the middle used to be as green as the grass on the other side of the proverbial fence. Look at the far right side for a clue.... The cause – an X7R ceramic capacitor hooked across two prime power phases. X7Rs like to have a DC bias on them; real AC makes them crack and short out.
A flashover between 115V 3F rectified bus (makes 270 Vdc and, practically, as much current as the wiring survives). Apparently, due to some moisture condensation upon the (removed) connector.	The mating connector, which was plugged into the one that flashed over (which you can't see 'cause it's been removed…), in the previous picture. The pins that are missing were not broken off. They literally evaporated into thin air! (Some of the vapor can be seen to have condensed back onto the PCB in the previous picture.)

 

Here's what happens to a static discharging style lightning arrestor when it is in-line with a poorly tuned antenna, Hi-SWR! Burnt enamel wire and toroid core fractured.

I thought I had properly adjusted my vertical antenna for 80M, then hit it with 1 KW from the linear....arc popping and spitting in the amp and no output until I removed this arrestor.

 

Apparently ‘school bus’ exceeds the mil spec of this portable radio.

Below are two pictures I took when I worked on one SNG truck (SNG means Satellite News Gathering, it is from the Digital Broadcasting Standards of Europe.) It was interesting , it is a dual frequency band system both C and Ku, (See PIC 1), but it seems like they forgot something: when the antenna rotated, the output flange of the C waveguide will be obstructed by the Ku waveguide. (See PIC2).

Finally they made the system work--good job. Well I think it is still a good experience to remember.

 

2. An ESD inspector had kittens when he found this high-isolation ESD lead attached to a bench grounding point!

I work at a C-band shuttle tracking facility at KSC (Kennedy Space Center). After a launch and several weeks of downtime, we went back to the site to bring up the system. Everything booted up nicely but the pull-out keyboard and monitor did not work. We unscrewed the top of the flip-up keyboard and monitor assembly and WHOA!

A real nice stench and a whole lot of dead mice… not to mention some circuit board problems.

There was a note in our orientation papers about knives, saying something to the effect of, "you only need a pocket knife. If the enemy has gotten past the walls, concertina wire, armed guards and machine guns, and through all the armed personnel walking around base, a bowie knife isn't going to help." I don't care. I take an Applegate-Fairbairn with me in my travel bag every time I get on a helicopter and it makes me happy.I'm on Camp Liberty which is one of the several bases built around the Baghdad airport. It's fairly civilized, you wouldn't guess we were at war sometimes, at least until a rocket lands a couple hundred yards away. It's godawfully dusty, the dirt turns into talcum powder an inch thick in some places. A few of Saddam's old palaces are built out here, so we have a few lakes and some interesting architecture to look at, not all of them are bombed out so that's kind of cool, I need to get into some of them and crawl around. Rumor has it that if you go to the no-man's land between base and the public side of the airport, you can get all sorts of things -- cars, guns, etc. I need to find out more about this. Spare parts are interesting. Get what you can, when you can and preferably two of them is the rule of thumb. I work with Raytheon, and they do a pretty nice job of keeping us in stock with what we need. Praise the lord for e-commerce and shipping to APO addresses, we can mail order just about anything we want so that's pretty cool.

A friend of mine in Colonia NJ had a new phone line put in last weekend. Check out the photo, observe the perfect placement of the drill bit going right through the electric service cable. Since the installer shorted the line upstream of the breaker panel, the only thing that stopped the fireworks was the wire burning open. I wonder how fast the meter was spinning for the 20 seconds or so when the lights were flickering in the whole neighborhood? The technician is lucky to be alive, and the homeowner is also lucky that his house didn't burn down.

I guess this could have been done by Verizon, or the cable TV company, or the satellite company. They all show up at your house with the same 2-foot long drill bit, and their technicians are all about as well trained.A local electrician fixed the mess that Saturday evening. and Verizon paid it, somewhere north of $1200. There is also some obvious damage to the house which Verizon is going to have to pay for but has not yet done so. The electrician asked the Verizon guy if he would pick a lottery number for him, since his luck was so good.

We've got a feeling this picture is going to get more than 15 minutes fame, expect to see it on an Official Safety Poster soon (especially if you work for Verizon...). Think before you drill! If anyone wants to buy the original full sized picture (it has remarkable detail), make us an offer and we'll pass it on to the lucky homeowner...

We were testing TO-247 "300 watt" power mosfets to see how long they would last at 300 watts actual dissipation, pulsed load, bolted directly to a copper block. We ran them at 300 watts until they exploded, and many would not make it past 50 milliseconds. We wound up with an Ixys p-channel part and an IR n-channel. This is for an MRI gradient coil driver with about 17KW peak output, 32 fets on a huge heatsink with copper heat spreaders.

 

Ever wonder why digital cellphones suck? If this level of engineering is any example, it should be crystal clear. OK, so the link would only suffer when the wind was blowing in a certain range of azimuth. The impact on the PCS sectors might be less, due to the lower frequency but good grief. Do people not get basic concepts anymore? Did they ever know them in the first place? What are they teaching in schools these days? Am I the only one who thinks this is wrong?

OK, so if the blades are nonmetallic composite, its still a big slab of dielectric material, moving through the link path (or one or more sectors) at a time, varying in thickness...multipath? Doppler shift? If there is any metallic structure it gets even more stupefying. Names, companies and location withheld to protect the RF physics
impaired.

 

New for March 2007! This Agilent ESA-L1500A/E4411A 9 kHz-1.5 GHz spectrum analyzer was for recently sale on Ebay a "buy it now" for $1299. It's not worth 1299 cents! Seller "Effy6" has 100% positive feedback, and warns: Sold AS-IS, no returns and refunds. We wonder if it powers up? Contributed by Kerry from down under!

 

Here's a look at the destruction inside the package... no wonder they wanted to put a lid on it, look at all of those ugly jumpers!

Let's start with this photo of a burned Ku-band waveguide window, courtesy of "HP". He was running at 18.3 GHz doing some critical pressure testing when it rapidly started to lose vacuum in the TVAC chamber. It's a good guess that this W/G window didn't like drastic pressure changes while passing 100 watts! We'd guess that voids the warrantee. Nice photo, tell us the camera model and we'll look for one on Ebay...

NS6Y, next time we're up the coast, how about fixing us a little lunch?

Here at Microwaves101, we have an HP Pavilion computer that literally cost us 10 times more in killed productivity and repair charges than its purchase price, which was more than a comparable Dell (we were stupid). We are about to get even. We now have on hand a large bag of Mexican M-80 firecrackers (approximately 1/4 stick of dynamite) which we will use to blow up the HP box while we record it digitally for permanent display in the Mortuary. Below are some preview pictures of this impending catastrophe; stay tuned and come back soon to see the results... anyone have Carly Fiorini's email address so we can send her this data?

Keep those pictures coming in!