Microwave Mortuary

Note to Mortuary contributors: please consider that your boss may not find your submission in the best interests of your Big Company. Once in a while we get a "please remove my submission" email, 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, please don't rip off Microwaves101 pictures for presentations without permission, that's bad manners. Maybe it's time for some training!

Note to Mortuary fans: In many cases, if you click on pictures on this page you can see higher resolution images.

May 2020:  Here's a fried FET, from the collection of the Unknown Editor. From the individual source vias and the way the nitride is pulled back, you can tell it is a millimeterwave effort.  The way the busbars are so far from the active channels, you know a microwave guy did not argue enough over the layout design rules. 

August 2019:  The following pictures come in from Bernd: 

It is the result of connecting a log-periodic antenna system based on a FR4 substrate to a 50 watt amplifier at 1.7GHz. I think it's interesting to see what branches of the array had burnt. With this photo everyone is seeing how these antennas work and which are the elements which are used for a given frequency. In a log periodic system only a few elements contribute gain as it can be clearly seen.

Thank you, Bernd!



July 2019: We love attending IMS, and we love to meet with our friends and fans.  And it's great that some suppliers offer free beer on the exhibit floor, so long as everyone remembers to be careful. Here's a picture of an accident at a booth that could have been prevented. Pro tip: don't assume you can use the display table-top as a place to set your beer when you fawn over the sample hardware...

October 2018: This picture came in from Dave, and we admit, it took us a minute to find the problem.  Look closely - can you see where they went wrong?

September 2018: From "Anonymous", this is a photo of the clean room at an unnamed facility...ick.

Here's a submission from David, who received this from a customer for calibration.  We aren't sure calibrating will fix that needle!

New for May 2018: From Tom.  Why is it soldered this way?

Close up of FET in an amplifier

New for April 2018: This came from Dave, title is "Coax Dielectric Loss".

Located problem at approximately 263 feet...opened the line 9 sections up from the base of the tower and confirmed that this was the location of the failure.  It was necessary to clean six transmission line sections above the failure due to contaminants from the burned Teflon. Nine transmission line sections below the failure were also cleaned due to the soot contaminants from the burn.

During the cleaning the crew found a black and charred ty-rap inside the transmission line above the failure point. We suspect that the original failure was due to a ty-rap that has now been completely burned up. There was no damage to the inner conductor or the expansion joint and the only point of failure was a break down of the Teflon insulator. The insulator had to be contaminated with some foreign object or material in order to breakdown and burn as is in this case.

Another ty-rap was found approximately three transmission line sections below the failure point.


Editor's note: just because something is white and plastic, does not mean it has excellent microwave properties like Teflon.  If a ty-wrap has poor loss tangent and is exposed to high electric field strength, it can heat up and eventually burn. Foreign object/debris (FOD) is a real problem in manufacturing hand and maintaining hardware.

Update from Tom

Here's what the remaining amps look like inside:


New for December 2017:   Some problems with a high-power combiner, from Tom...

So is this Wilkinson or Wontkinson*?

This is a 2500 watt FM broadcast band setup from a manufacturer of very poor and well deserved repute. The upper unit is the exciter, which is a combined exciter and 150 watt PA unit. In normal operation it runs about 45 watts which goes to the power divider in the center unit, which divides it out at 0 degrees to each amplifier. Those feed a Wilkinson combiner, which sends power to the final output. Each amplifier is 500 watts output. Remember this for later.

IMG 20171207 094759

Sounds good, right? Brace yourself.

The power divider is complete poppycock. I've never seen anything like it, and it doesn't work well -- a few watts reflect back to the exciter. But then you hit the Wilkinson and the quarter wave transformers made of coax cable. Very very thin coax. I forget what the part number is but the center conductor is maybe 28 gauge. For 500 watts.

So now, the resistors. Their purpose is to dissipate energy and provide isolation in the event of a mismatch, which shouldn't be sustained normally because the amplifier should ideally have some safety features that will trigger foldback or shutdown in the event of mismatch.

This one has none.

The amps don't have ANY interlocks or unified power control at all, and simply switching one of the amps off with the system running will thus trigger meltdown. The resistors also can't handle a serious mismatch dissipating the full power of an amp. I don't know which amp(s?) went down first but I got the message the signal was "unlistenable". As I drove to the site I heard a horrible chunking noise full of 60 cycle ac growl. It sounded like the transmitter was going down a garbage disposal.... Which wouldn't be far from what actually happened. Three of the five amps seem to have blown up and the combiner is WELL DONE, it even melted the fan partially.

It should be self explanatory which photos are before and after.  The copper plate turned blue!!

 IMG 20171031 091434 20171207210338456 400 Capture 2017 12 07 11.29.57 400
 IMG 20171031 085659 400  Capture 2017 12 07 11.28.15 400
 IMG 20171031 085901 20171207210255404 400  Capture 2017 12 07 11.28.31 400
 IMG 20171031 085035 20171207210156821 400R  Capture 2017 12 07 11.28.41 400


These images come from Jose. Editor's note... Jose is not alone in the GaN-blows-up-when-you-look-at-it-funny experience. Send in your own non-ITAR/non-proprietary failure photos please!

In the process of testing some GaN PAs we ended up blowing up one of them, and the blown up transistor is so magnificent that I felt compelled to share it with the world. Brief story behind it, we needed to do some measurements without using off-chip capacitors and this unstable PA ended up drawing too much current. Surprisingly it still had gain after being left with one less transistor.

Before 400 After 400

Thanks to Zach:

I hope this iphone charging brick suits the hardware graveyard, my girlfriend's sister had an extension cord next to her closet and somehow a metal clothes hanger got in between the outlet and the prongs and tripped the breaker thankfully but not without doing some damage...

By the way, this is precisely the reason that grounded outlets are now mounted with the ground on top (see here). Too bad phone chargers have to be made so cheaply they leave off the safety ground...

Iphone charge cord

New fro July 2017: here is a classic needle probe crash from Robert.  Thanks!  This is a DC probe, with integrated bypass capacitors used to stabilize amplifiers when they are RF rested on-wafer. Believe it or not, this type of accident can be repaired, either by a patient technician, or by returning it to the manufacturer.

ProbePing 800


This cautionary tale came from Darrell. 

This is an Anritsu 2.92 mm airline from Ebay. The female end was stuck between the wall and center pin of the end cap (short circuit) when it arrived. It's going back for a refund.  Ebay for anything with microwave connectors seems like gambling. I've sent back about 25%.

IMG 8924.resized 400 IMG 8924 crop 400

Also from Tom, titled "C-band by the Sea".  The original picture are very high-res, unfortunately they are too big to upload.... we'll figure it out and add them later.

This C-band satellite feed horn and polarotor lived by the sea. Can you tell?

Link margin and EB/N0 just gradually shrank away to Davy Jones Locker.

IMG 20170105 150717 400 IMG 20170105 150749 400


New for January 2017: from Jacob, in two installments:

This week I bought twenty surplus Nokia RRIC boards. When I bought them, it was very clear the warranty was void... I bought the board very cheap, primarily for salvaging the N-connectors for ham radio use. And there are some nice other components like several xtal oscillators and smd chips for practicing my smd soldering skills. I knew the condition of the board when I bought them and for the price I couldn't complain. But I wondered what happened to the board before I bought them... Are they been run over by a truck?! Not only parts are missing/split in half, but parts of the pcb are cracked or even missing, including half of the connector!

amateurtelecom RRIC MM01 400 amateurtelecom RRIC MM03 400
amateurtelecom RRIC MM04 400 amateurtelecom RRIC MM05 400
amateurtelecom RRIC MM06 400

amateurtelecom RRIC MM07 400

Want more? Check out the Microwave Mortuary Archives!

Author : Unknown Editor