SRC Forum - Message Replies
Forum: Reliability & Maintainability Questions and Answers
Topic: Reliability & Maintainability Questions and Answers
Topic Posted by: Reliability & Maintainability Forum
Organization: System Reliability Center
Date Posted: Mon Aug 31 12:47:36 US/Eastern 1998
Posted by: Hei-Ruey Jen
Date posted: Mon Jul 26 15:47:41 US/Eastern 1999
Subject: microphonics interference
I have received a request to do reliability testing of microphonics' effect on microwave devices. From a website I found, someone claimed that there was no standard on how to test the microphonics sensitivity. Screwdriver tapping, simulated hail, etc. are few examples people have used. Does anyone know anything on how to perform this test correctly and reliably? This test is to simulate the rain drops and hail hitting an outdoor device, which will create sound to interfere microwave device performance. Thanks.
Reply Posted by: Bob Fitzgibbon
Organization: Reliability Analysis Center
Date Posted: Wed Jul 28 15:49:54 US/Eastern 1999
Some clarifications are in order. The type of microwave device must be specified - is this a VCO, PLL, entire transmitting system, or what, exactly? The term 'microphonics' normally refers to audio frequency modulation of an RF signal being generated by the device being considered. This modulation may be amplitude (AM) or frequency (FM), although it is usually the latter. A case in point would be a varactor-tuned VCO in a PLL arrangement, ultimately locked to a crystal oscillator. Mechanical shock to the crystal will cause its effective frequency to shift. Depending on the PLL bandwidth, this will cause the VCO to shift also. The result will be a frequency deviation. For a single shock (e.g., 10g's from dropping on a hard surface) the output RF signal will deviate from nominal with a 'ringing' pattern, settling out to zero eventually. A person listening to an FM receiver tuned to the VCO frequency would hear a sound like the sound of the impact, with volume depending on the peak deviation. For repetitive vibration, the deviation will be continuous in nature.
The microphonic effects must be compared to the desired modulation characteristics of the system. In an FM transmitter, for example, the peak and average (or rms) deviation under normal conditions is specified. Any microphonic modulation must be 'small' in comparison. How 'small' it has to be depends on the detailed nature of the demodulation process. For data transmission, microphonics are simply one more unwanted noise component and their effect can be analyzed accordingly.
The usual way of specifying microphonic performance is in terms of peak deviation at a specified shock level (say 10g's in any axis), or rms deviation for a random vibration spectrum (say .04g^2/Hz from 80 to 400 Hz). If AM is the issue, deviation is replaced by modulation index. If you have a specific real-world situation, such as rain or hail, the best thing is to try it out in an actual rainstorm (or else simulate it somehow in the lab) and measure the resulting rms deviation. This has to be 'small' compared to the desired deviation.
As to how to design for minimal microphonics, there are a few rules of thumb. All frequency determining elements (coils, caps, mechanical resonators, crystals) should be securely staked to the board, and the board itself should be cushioned (i.e., mechanically isolated) from external impacts to the housing. Avoid mechanical tuning (such as trimmer caps) if possible. The circuits generating the tuning voltage should be made minimally sensitive to any vibration or shock, as should the main power supply.