MEMS Microphones Test and Evaluation Technician Interview tips

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Aniol1349

Well-known member
Joined
Jan 16, 2012
Messages
103
Location
Deep forest
Hello guys!

I have an interview at a  reputable company for a MEMS Microphones Test Technician and I'm looking for some good info on MEMS microphone testing or anything related to microphone lab testing in general.

Any tips would be much appreciated.

Below I'm attaching the job listing if anyone would like to have a look.

Regards and wish me well :)


This position plays an important role within our engineering team and supports development of leading-edges MEMs microphones. This is an opportunity to work with a cross functional team of design, validation and test engineers.


Responsibilities

Help MEMS Engineers with test setups.  Soldering components and modifying engineering boards.
Install and maintaining lab equipment and engineering boards
Setup and execute automated tests in the lab using various scripting environments
Debug test setup issues and debug/develop test scripts
Collect and analyze data, then prepare and present summary reports
The tasks will be varied and differ day to day
Required Knowledge, Skills and Experience

HNC or HND in Electronics or closely related discipline
Ability to read/interpret circuit board schematics, specification docs, rework instructions
Surface Mount (SMT) Solder/Unsolder/rework Skills
Experience working with lab equipment such as  Audio Precision, Oscilloscopes, Multi-meters, Power Supplies etc
Experience working with lab test setups such as evaluation boards, cables/connectors, connecting boards to test equipment
Data collection and analysis skills (Excel skills a must, JMP experience also nice).
Organizational skills: record, organize, and report validation setups and results in meaningful methods
Quick learning and careful attention to detail are critical skills


Preferred Knowledge, Skills and Experience

Wafer fab experience
SEM analysis skills
Previous experience in Yield Enhancement and Failure Analysis.
Experience making and/or reviewing mask layouts
Experience interacting with off-site fabrication services (foundries).
Understanding of DOE theory and methodology.
Experience with high volume manufacturing of MEMS or IC
Basic software knowledge in text based programming such as C or scripting language is desirable
Windows: experience with installing/configuring OS and software, driver installation/debug, general OS debug, Excel macros is desired
Experience developing or modifying test programs in Labview
 
> no idea about  MEMS

Oh, come on. Get past 1999.

MEMS is micro-"machining" using Silicon foundry techniques: masks, etchants.

If you etch Silicon into a microscopic "diving board", you get an accelerometer. When you move your cell-phone to track the Pokemon, that's how it reacts to your swing quickly. Also how car-navigation detects that the car turned long before the GPS signals can settle to the new course.

An obvious variant is to tie a paper diaphragm to the diving board and sense pressure or sound. SEMS pressure sensors must have become common in cars. The "diving board" is expanded to a diaphragm over a sealed cavity. (No, I don't know how that etch that.) Taking pressure ports on both sides allows differential measurement, as for engine venturi sensing or for directional microphones.

The job description is written very wide. You might just run canned scripts on automated gear and put numbers in Excel, which many workers could do. But they might be happy if you could go to the wafer fab and talk MEMS etchant time/chemistry, a rather unique worker. Interesting that they also would like Windows OS de-bug, meaning they may not have a heavy PC Support staff on site. Taking the widest reading, they would like somebody who could test and outsource MEMS on his/her own, doesn't need this job.
 
I have done MEMS microphone design.
The job requirements look like they are looking for superman, as usual.
But the core of it is some soldering skills, ability to write simple scripts for test programs. and ability to write reports.

Other things like electron microscope operation skills you're not likely to have, but it's not very hard...easily learned.

I'd read up on basic wafer fab techniques before the interview.

If you have any other questions I'll be glad to answer.

Les
http://lmwattstechnology.com/
 
And for anyone who haven't seen one of these - Adafruit usually tells the brand of the Star Component on these breakout boards, but I don't see that here. A little googling of the part number shows the manufacturer is Knowles:
https://www.adafruit.com/product/2716
 
> Knowles:

Who are the Big Name in hearing aids.

Their stuff is so good it gets into cellphones and laptops, all small toys.

Which points to why we "haven't seen one of these" here. Hearing aids only have to hear better than a deaf person. A lot of other small toys don't have to be much better. That particular job is linear up to 129dB SPL, but the noise level is around 35dB SPL. Compare to 14dB SPL for the better studio mikes, 25-30dB SPL for 1-buck electrets.

The "100Hz-10KHz" quoted by AdaFruit needs qualification. It's only down 0.5dB at 100Hz, the droop should be simple first-order, the -3dB point may be 25Hz. OTOH it is up a good 2dB at 10KHz, leading to a resonance of unspecified height and frequency. Some of these mikes are un-damped and ring badly, far above where a hearing aid is expected to work.

http://www.knowles.com/index.php/eng/content/download/5642/89383/version/7/file/SPW2430HR5H-B.pdf
 
PRR,

There's a direct tradeoff between noise and max SPL. To get to the  noise level of the one buck electrets we have to
set mechanical overload to around the 110 dB region.

Pressure noise is also increased by damping the diaphragm resonance, so it's typically not done.

You can get the noise down with arrays, but that doesn't help the low overload point that most have.

Typically the diaphragm is 300 micron diameter polysilicon with 2 micron backplate spacing. Polarization is only a few volts.

Les
http://lmwattstechnology.com/
 
> tradeoff between noise and max SPL

The range is limited by sensing technique.

But the noise level is also limited by diaphragm size against transducer effectiveness and electronic amplifier noise level. Small capsules don't catch much air, can't deliver large output, whereas amplifier hiss is more or less all the same.

That Knowles is roughly 0.1" all over. Compare to a 1" large studio diaphragm. The Knowles is tapping 1/100 as much acoustic power, 1/10th the voltage. Adjusted 20dB, the 35dBSPL spec would be 15dBSPL, a fine spec for a 1" condenser.

There's lots of other techniques. The hearing aid can use a resonator to boost 3KHz for speech intelligibility, actually raising the acoustic level before the mike, and thus reaching further down into room noise. Of course the studio mike has resonances, but usually managed to be low-Q above 10KHz for flat response.
 
Hey guys,

Did my interview on Wednesday, It was actually simpler than I though it's gonna be. Some physics theory, simple questions about circuits and components, later I had two technical tasks, first to measure the sensitivity of the microphone using calibrated reference mic and reference signal, easy enough. Second task was to take a measurement under a light emitting microscope(not sure if I remembered the name right) measuring the positioning of the diaphragm in relation to the acoustic ports.

It was a really enjoyable experience and I did really good, however they decided to take someone with more experience.
Apparently they were so impressed that they will try to get me in into another position, so fingers crossed.

Cheers!
 
Back
Top