Parasitic capacitance of SMD resistors
Jeroen Belleman
capacitance of SMD resistors.
From the dimensions of a 1MOhm 1206 resistor, assuming
all the field is in the alumin dielectric, and pretending
it is otherwise a simple flat-plate capacitor, I found
it should be in the 35fF ballpark.
Measuring one using my HP8753D network analyzer, I got
50fF. Not too bad an agreement for a five-minute lash-up.
That's a factor of 6 below John's results.
You were right, Win, 300fF was way too high.
Jeroen Belleman
John Larkin
I calculated, roughly, 100 fF, based on the alumina between the end
caps. But the resistance element effctively shunts sections of the
alumina, and that varies with value and frequency, so maybe there's no
single correct value for C.
My AADE meter reported a lot higher than my Boonton, numbers like 0.3
and 0.1 pF respectively. I think the AADE is also affected by the
resistance itself.
How did you mount/fixture it?
John
George Herold
Hey, thanks for the measurement Jeron!
George H.
Jeroen
> On Fri, 03 Dec 2010 17:12:04 +0100, Jeroen Belleman
> <jer...@nospam.please> wrote:
>
>> A few days ago there was some discussion about parasitic
>> capacitance of SMD resistors.
>> Measuring one using my HP8753D network analyzer, I got
>> 50fF. Not too bad an agreement for a five-minute lash-up.
>
> alumina, and that varies with value and frequency, so maybe there's no
> single correct value for C.
>
> My AADE meter reported a lot higher than my Boonton, numbers like 0.3
> and 0.1 pF respectively. I think the AADE is also affected by the
> resistance itself.
>
> How did you mount/fixture it?
I mounted it in series between two SMA connectors. The NA S21
is basically a straight line with positive slope of 20dB/decade
over several decades, characteristic of a pure fixed capacitance.
I'm at home now, and the NA is in the lab. I'll post a picture on
Monday, if anyone wants to see the measurement.
Regards,
Jeroen Belleman
John Larkin
wrote:
>On 12/03/2010 06:32 PM, John Larkin wrote:
>> On Fri, 03 Dec 2010 17:12:04 +0100, Jeroen Belleman
>> <jer...@nospam.please> wrote:
>>
>>> A few days ago there was some discussion about parasitic
>>> capacitance of SMD resistors.
>>>[...]
>>> Measuring one using my HP8753D network analyzer, I got
>>> 50fF. Not too bad an agreement for a five-minute lash-up.
>>
>> I calculated, roughly, 100 fF, based on the alumina between the end
>> caps. But the resistance element effectively shunts sections of the
>> alumina, and that varies with value and frequency, so maybe there's no
>> single correct value for C.
>>
>> My AADE meter reported a lot higher than my Boonton, numbers like 0.3
>> and 0.1 pF respectively. I think the AADE is also affected by the
>> resistance itself.
>>
>> How did you mount/fixture it?
>
>I mounted it in series between two SMA connectors. The NA S21
>is basically a straight line with positive slope of 20dB/decade
>over several decades, characteristic of a pure fixed capacitance.
What was the frequency range? Any conjectured influence of the
resistance element would happen at relatively low frequencies. Things
should start to deviate from the ideal slope in the low MHz range,
roughly where Xc approaches 1 meg.
John
Jeroen
Agreed. I was looking well above that, in the 10 MHz to 1 GHz
range. I'll have another look on Monday.
Jeroen Belleman
George Herold
>
> - Show quoted text -
You might try a few lower R values, (If it's not too much to ask.) It
would be interesting, to see if you see anything like John.
George H.
Jeroen
1206 SMD resistors. Basically they all have about 50fF
of shunt capacitance, although at lowish resistance,
other effects start to take precedence.
There is some info about the setup and some plots of
the results at <http://cern.ch/jeroen/resistor/shuntC.html>.
I haven't really taken a detailed look around the
transition region between resistive and capacitive
impedance. There's a link to the frequency-magnitude
data that should allow anyone curious enough to see for
themselves.
Regards,
Jeroen Belleman
Jan Panteltje
<jer...@nospam.please> wrote in <idjcud$6ol$2...@speranza.aioe.org>:
That is very nicely done, thank you :)
mi...@sushi.com
I ran a SMD resistor on a network analyzer once for yucks. I don't
recall the LCR model, but I do recall the L was virtually that of the
electrical length of the resistor.
George Herold
That's wonderful, Thanks Jeroen. Do you work at Cern? (It must be
fun!)
George H.
Jeroen
> On Dec 6, 2:16 pm, Jeroen <jer...@nospam.please> wrote:
>> I have some measurement results for different values of
>
> That's wonderful, Thanks Jeroen. Do you work at Cern? (It must be
> fun!)
>
> George H.
Indeed I do. These are interesting times...
Jeroen Belleman
John Larkin
wrote:
>On 12/03/2010 06:32 PM, John Larkin wrote:
Here's another measurement.
Tek TDR/sampling head and fixture:
ftp://jjlarkin.lmi.net/0805_res_fix.JPG
This is the TDR (lower) and TDT (upper) with no resistor, just the gap
in the CPW trace:
ftp://jjlarkin.lmi.net/0805_nores.JPG
And here's the TDT blip, the signal that passes through a 1M 0805
resistor:
ftp://jjlarkin.lmi.net/0805_res_TDT.JPG
If you do the math on that, it works out to 47 fF.
So the AADE meter, and maybe the Boonton, don't measure this
capacitance very well. I still don't know if the resistance element
changes C as a function of frequency, but maybe it doesn't.
John