When are ferrite beads the most approproate component to use?

ead.

=A0 =A0 ...Jim Thompson

=A0 =A0| =A0 =A0mens =A0 =A0 |

=A0 | =A0 =A0 et =A0 =A0 =A0|

=A0|

=A0 =A0 =A0 |

Technically the op amp is put in a voltage follower for the PSRR test. I suppose that is the same as input referred. I don't find that cheating since the voltage follower is a common application.

One problem with doing the PSRR test is you need to maintain a low impedance at the DUT pin while wiggling it. This generally done by passing the DC through a transformer. If the driving impedance is high, the test isn't valid. If you attach a generator to the DUT pin, the test is questionable since the generator impedance is probably 50 ohms. Even using op amps to drive the dut is questionable since their output impedance rises with frequency. Too bad the "damn fast buffer" is history.

ahd

=A0Most of

tion

a

s outside gets a

I bought some BAFNS (big ass ferrite noise suppressors) at the Livermore flea market from an audio company that slips my mind. Anyway, the ferrite was on a cable internal to the project. When shipped, the weight of the ferrite would yank the cable from the headers, causing the product to be DOA. They made the cable longer so it could be supported at both ends. I suppose you only make that mistake once. ;-)

It probably pays to have a ferrite kit. Either free if you ship enough product or very cheap. [Or cause product to be shipped if you are just a designer.] There are different beads for different uses.

I just did such a test for a small sensor amplifier. In my case I used a

50R cable termination which made the generator impedance 25R, but that's not mandatory. Gain-phase analyzers have a reference input, so you don't mind the 1/f AC voltage resulting from the bypass caps. There's no lack of signal either since PSRR decreases at the same rate the applied AC voltage does.

Also mind that PSRR is often misleading since it's (as often) well below the real circuit's PSRR which is (still as often) dominated by the opamp's PS rail to inputs parasitic caps. In my case it degraded from the opamp's 80dB down to a real 46dB.

--
Thanks,
Fred.

That effectively shorts out the coupling path from the rails to the input pins, about best-case for PSRR measurements.

Once a supply pin has the desired DC voltage and desired AC ripple, why does it matter how it got there?

John

[snip]

No. What is does is divide the expected output movement by the loop gain... so it looks good :-(

...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
                    Help save the environment!
             Please dispose of socialism responsibly!

ly

=A0 =A0 ...Jim Thompson

=A0 =A0| =A0 =A0mens =A0 =A0 |

=A0 | =A0 =A0 et =A0 =A0 =A0|

=A0|

=A0 =A0 =A0 |

But I still think the follower is a fair test circuit. What would you propose?

I did a quick survey yesterday regarding PSRR test setups. LTC didn't provide a test circuit. I'd have to double check, but I think Analog didn't either. Intersil (yeah, who cares about them) had a test circuit where they put a buffer at the output of the DUT, I guess just to drive the test equipment. National had a document on op amp testing and basically showed the voltage follower.

e bead.

y

=A0 =A0 =A0 ...Jim Thompson

=A0 =A0 =A0| =A0 =A0mens =A0 =A0 |

=A0 | =A0 =A0 et =A0 =A0 =A0|

=A0 =A0|

=A0 =A0 =A0 =A0 |

=A0|

|

Suppose the DUT has some capacitance to ground from a supply rail. Then the high impedance source would not be stimulating the circuit guts properly. Obviously, this become a problem at higher frequency. So not driving it hard could give a more favorable result.

Also not mentioned is the applied AC for such testing. I do 100mvpp. Maybe 250mvpp is the device is really good, i.e. the output is hard to measure.

Thinking about this a bit more, I can see why the voltage follower circuit is used. If you are measuring the output with a volt meter, you can't reject circuit noise from the test frequency. I use spectrum analyzers for PSRR, but this may not be a hard and fast rule in such testing. Now this makes me wonder if a circuit produces harmonics, those would be measured with a simple meter, but not a network analyzer. I could see that scenario making a device work worse in an application than the test data.

Thou should have squirreled away some of those before the last order bell was rung. I did :-)

Same with the SL6440 mixer from Plessey. One of the best things since the invention of pivot irrigation. So, before the company went phut I got myself a tube of these as well.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.

Joerg a écrit :

Ah, that was you...

And did the CEO kaboomed then?

--
Thanks,
Fred.

I have a dozen of the LH0033 and LH0063, as well as of the lesser-known LH4009, which was one of the weird outpouring of hybrid parts in National's 1989 catalogue. Almost all were cone in a couple of years.

Now I need to husband my MAT04s too. Hopefully they don't stop making BF862s! :(

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058
hobbs at electrooptical dot net
http://electrooptical.net

bead.

If the DUT V+ pin is +15 volts and has 250 mV p-p AC superimposed on it, how is the DUT going to know how the 250 mV was generated, or what the impedance of the generator is?

John