[time-nuts] 10 MHz low pass filter

John Lofgren jlofgren at lsr.com
Thu Jun 21 13:39:01 UTC 2012


The inductors used in this board look like multilayer ceramic chip types.  They actually have a fairly low stray field around them and they're "wound" around an axis that's perpendicular to the PCB.  Most of the solenoidal coupling will be in the axis normal to the board.  While rotating inductors relative to each other is a good general practice it doesn't help much in this instance.

One thing we have found that helps, though, is voiding the ground underneath the inductor.  In filters where you're trying to get more than about 40 dB of rejection it's possible to have coupling through induced ground currents that will just end up bypassing the filter by coupling from one stage to the next.  We had a case in the past where we were fixing a customer design that used a 2 stage crystal filter.  The filter nulls were off by something like 20 dB from the data sheets and the measured data from the filter vendor.  Standing the in/out matching inductors up vertically and wiring them back down to the board made the response look like it should.  Voiding the ground under the inductors on the next board spin fixed the problem.

Self inductance and self resonance of the capacitors is always something to watch out for.  The general rule of thumb we use for generic NPO multilayer chip capacitors is an inductance on the order of 1.0 to 1.3 nH for 0402 or 0603 size parts.  The better RF specific parts from MuRata, ATC, and Johanson will have lower inductance and higher SRF.  Typically a good bypass capacitor for 900 MHz is 18 to 22 pF and for 2.4 GHz it's 6.8 or 8.2 pF.  At 10 Mhz and the first few harmonics values on the order of 1 to 20 nF would be below SRF.

-John



-----Original Message-----
From: time-nuts-bounces at febo.com [mailto:time-nuts-bounces at febo.com] On Behalf Of Thomas S. Knutsen
Sent: Wednesday, June 20, 2012 6:57 PM
To: lists at lazygranch.com
Cc: Discussion of precise time and frequency measurement
Subject: Re: [time-nuts] 10 MHz low pass filter

I don't have any problems with rotating the inductors, after all, that is
one of the best way to avoid coupling between them, but the main problem as
I see with that board is that there are 2 caps that would become an series
resonance with the inductance in the via to reach the ground plane.

Of course, at 10MHz this is just theoretical, since the problem most
probably would appear above 500MHz, and the 50'th harmonic of an OXCO
should be low.

My experience says that the inductance in the capacitor it self should be
low, specialy if NP0 or such capacitors caps are used.

An 10MHz sallen key lowpass may be interesting to build,and with the GHz
bandwith op-amps avaible today, it should work great.

Thomas.

2012/6/21 <lists at lazygranch.com>

> **
> In the days when I had access to a network analyzer with a chip component
> fixture (all calibrated of course), I tested components on hand just to see
> how ideal they were. Chip resistors are quite good. The inductance is
> basically the electrical length of the device. Caps can be decent. My
> recollection is Johanson had some really good (low parasitic) caps.
> Inductors basically suck.
>
> You will note in most RF board design with lumped elements, they rotate
> adjacent inductors to reduce mutual inductance.
>
> 10MHz is probably too low in frequency for practical stripline. You could
> probably do active filters these days, but the power budget would not be
> trivial.
>
> ------------------------------
> *From: * "Thomas S. Knutsen" <la3pna at gmail.com>
> *Date: *Thu, 21 Jun 2012 01:17:02 +0200
> *To: *<lists at lazygranch.com>; Discussion of precise time and frequency
> measurement<time-nuts at febo.com>
> *Subject: *Re: [time-nuts] 10 MHz low pass filter
>
> Based on that PCB, I want to see an sweep to at least 1GHz.
> The reason is that experience have shown that the inductance (perhaps
> 10nH) in series with C2 and C6 would damage the stop-band rejection at UHF.
>
> Used with an OXCO this would not matter, but the desire to make the
> ultimate filter is still there.
>
> Thomas.
>
> 2012/6/21 <lists at lazygranch.com>
>
>> If the output is buffered, there really shouldn't be a problem.
>>
>> Incidentally, I can crank out high order LCR filters all day just by
>> transforming prototypes out of Zverev. But it has been my experience at
>> even 10MHz the parasitics of the elements will throw off the design.
>>
>>
>> -----Original Message-----
>> From: Bob Camp <lists at rtty.us>
>> Sender: time-nuts-bounces at febo.com
>> Date: Wed, 20 Jun 2012 17:49:37
>> To: Discussion of precise time and frequency measurement<
>> time-nuts at febo.com>
>> Reply-To: Discussion of precise time and frequency measurement
>>        <time-nuts at febo.com>
>> Subject: Re: [time-nuts] 10 MHz low pass filter
>>
>> Hi
>>
>> That's a pretty high order filter. The output of most OCXO's already has
>> a filter on it. Combining two filters (especially high order ones) without
>> isolation between them is not a good idea. The resulting transfer function
>> will not be what you expect it to be....
>>
>> Bob
>>
>> On Jun 20, 2012, at 4:54 PM, Joseph Gray wrote:
>>
>> > I came across this filter design recently and thought I'd build a few.
>> >
>> > http://www.jwmeng.com/AppNote/AppNote003.html
>> >
>> > I was about to place a Mouser order when I came across someone selling
>> > filters, based on the above design.
>> >
>> >
>> http://www.ebay.com/itm/10-MHz-filter-kit-for-tcxo-gps-pll-/110893777470
>> >
>> > The price with a board seemed low enough that I ordered a few. I'll
>> > let everyone know how they work out.
>> >
>> > Joe Gray
>> > W5JG
>> >
>> > _______________________________________________
>> > time-nuts mailing list -- time-nuts at febo.com
>> > To unsubscribe, go to
>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>> > and follow the instructions there.
>>
>>
>> _______________________________________________
>> time-nuts mailing list -- time-nuts at febo.com
>> To unsubscribe, go to
>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>> and follow the instructions there.
>> _______________________________________________
>> time-nuts mailing list -- time-nuts at febo.com
>> To unsubscribe, go to
>> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>> and follow the instructions there.
>>
>
>
>
> --
>
>  Please  avoid sending  me  Word  or  PowerPoint  attachments.
>  See  <http://www.gnu.org/philosophy/no-word-attachments.html>
> PDF is an better alternative and there are always LaTeX!
>
>


-- 

 Please  avoid sending  me  Word  or  PowerPoint  attachments.
 See  <http://www.gnu.org/philosophy/no-word-attachments.html>
PDF is an better alternative and there are always LaTeX!
_______________________________________________
time-nuts mailing list -- time-nuts at febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.



More information about the time-nuts mailing list