Without knowing whether you are measuring a cable ferrite or on-board SMT parts, here's my 2-cents comment: In order to achieve good results of the ferrite performance up to 500 MHz, we need to be careful: 1) build the test fixture such that the characteristic dimension of the overall test fixture is less than one-tenth of the wavelength ( < 6 cm) 2) if you are using network analyzer or an HP4291 impedance analyzer, carefully calibrate so we can exclude the parasitics of the test fixture 3) to measure the impedane response of your cable ferrite, chosse a cable that's representative of your real application. For example, if your application is to have a 14-pin flex cable go through the ferrite, use a small section of 14-pin flex cable and tie all the pins together on both ends ( rather than use a single wire ). You'll find the resonance frequence is significantly affected by such details. 4) Depends on your application, you might also want to check with your vendor for the impedance of the ferrite under small different DC CM current and low frequency AC CM current bias. 5) measure at least 3~5 samples to get an idea of tolerance on impedance and resonance frequency. Good vendors can control the impedance within +/- 20%. Some can be off to +/- 40%. Hope this helps.
Minjia Xu ( Catherine) <>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>< EMC Engineer , Ph.D Hardware Test & Regulations All-in-One Personal Printing - HP [email protected] <>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>< From: [email protected] [mailto:[email protected]]On Behalf Of Gary McInturff Sent: Thursday, December 18, 2003 12:59 PM To: 'Price, Ed'; 'Grasso, Charles'; '[email protected]' Subject: RE: Measuring a ferrite performance Ed, Steward ferrite discusses this question their website. Other than noting it was there I didn't read it so I can't speak to the quality of the article, but one would assume these guys should know how to make such a measurement. Gary From: Price, Ed [mailto:[email protected]] Sent: Thursday, December 18, 2003 11:43 AM To: 'Grasso, Charles'; '[email protected]' Subject: RE: Measuring a ferrite performance -----Original Message----- From: Grasso, Charles [ mailto:[email protected]] Sent: Thursday, December 18, 2003 9:40 AM To: '[email protected]' Subject: Measuring a ferrite performance Greetings all: I have a requirement to test the performance of ferrite bead (one that that can fit over a power cord) to 500MHz for qualification of alternate vendors. I have an impedance analyzer - but it only goes to 40MHz. Does anyone have a good way to verify ferrite impedance.?? Thanks Best Regards Charles Grasso Senior Compliance Engineer Echostar Communications Corp. Tel: 303-706-5467 Fax: 303-799-6222 Cell: 303-204-2974 Email: [email protected]; Email Alternate: [email protected] Chuck: Assuming that you have a signal generator and a spectrum analyzer, you could try the following: 1. Get a small RF project box. Mount two N or BNC or SMA coax connectors on opposite walls of the box. Solder a short wire directly from one coax connector center pin to the other. 2. Decide on system impedance. If 50 ohms is OK, then just connect a 3 dB attenuator to each of the coax connectors. If you want another impedance, then put appropriate resistors (carbon comp preferred) into the project box. 3. Connect the signal generator to one attenuator, and the spectrum analyzer to the other. 4. Now set the generator to 0 dBm and sweep or step the frequency range, noting the spectrum analyzer response. 5. Now open the box, break the wire, thread the bead onto the wire, and re-solder the wire. 6. Repeat step 4. The amplitude differences will be the insertion loss due to the bead performance. 7. Repeat steps 5 & 6 with any other bead samples. This isn't an exhaustive test, but it's a good A to B comparison of RF loss. Ed Ed Price [email protected] WB6WSN NARTE Certified EMC Engineer & Technician Electromagnetic Compatibility Lab Cubic Defense Applications San Diego, CA USA 858-505-2780 (Voice) 858-505-1583 (Fax) Military & Avionics EMC Is Our Specialty
