On Thu, 04 Aug 2022 10:15:25 -0700, John Larkin
<jlarkin@highland_atwork_technology.com> wrote:
>On Thu, 04 Aug 2022 18:35:45 +0200, Jeroen Belleman
><jeroen@nospam.please> wrote:
>
>>On 2022-08-04 16:08, jlarkin@highlandsniptechnology.com wrote:
>>> On Thu, 4 Aug 2022 02:24:55 +0200, Klaus Vestergaard Kragelund
>>> <klauskvik@hotmail.com> wrote:
>>>
>>>> On 04/08/2022 00.14, John Larkin wrote:
>>>>> On Wed, 3 Aug 2022 14:54:38 -0700 (PDT), whit3rd <whit3rd@gmail.com>
>>>>> wrote:
>>>>>
>>>>>> On Wednesday, August 3, 2022 at 11:03:53 AM UTC-7, sea moss wrote:
>>>>>>> Two part question (my google skills have failed me for both):
>>>>>>>
>>>>>>> 1. Does anyone know approximate pad-to-pad capacitance for the common SMD packages (0201, 0402, 0603, 0805)? Ever measured it? I can't use the parallel plate capacitor formula, since the plates are adjacent, not parallel. (assume there is no copper below the pads)
>>>>>>
>>>>>> But, regardless of the pads, the component's mating metal surface has that
>>>>>> capacitance regardless, doesn't it? You don't need always to make the 'plates' larger
>>>>>> than that component feature.
>>>>>
>>>>> Most surface-mount part capacitance is in the FR4, not in air.
>>>>>
>>>>>>>
>>>>>>> 2. Given an SMD footprint as above, how close does a ground plane need to be (on the same layer) to decrease the effective pad-to-pad capacitance? Is there a formula which gives a good approximation?
>>>>>>
>>>>>> The use of guard rings can make such capacitance negligible, if that's required, but...
>>>>>> trying to simply treat the tiny elements as separable capacitances, isn't a good and useful
>>>>>> model.
>>>>>
>>>>> It's often important to know the three capacitances: each pad to the
>>>>> ground plane, and 3-terminal equivalent capacitance pad to pad.
>>>>>
>>>> turn PCB has a toolkit for a number of different calculations on PCB
>>>> structures:
>>>>
>>>> https://saturnpcb.com/saturn-pcb-toolkit/
>>>
>>> It's easy to measure attofarads with cheap instruments that a decent
>>> home lab should normally have.
>>>
>>
>>Femtofarads, sure, but attofarads? Please explain?
>>
>>Jeroen Belleman
>
>Femtos, just make a voltage divider from a sig gen, Cx, and the usual
>15 pF input of a scope. Signal average a bit or FFT at the low end. A
>few fF resolution is easy. To get to aF, add a jfet or an opamp.
>
>One could also add an inductor to resonate out most of that 15 pF.
>
>Lots of ways to do this.
The old HP selective (tuned) voltmeters were most cool for things like
this.
Reply by John Larkin●August 4, 20222022-08-04
On Thu, 04 Aug 2022 18:35:45 +0200, Jeroen Belleman
<jeroen@nospam.please> wrote:
>On 2022-08-04 16:08, jlarkin@highlandsniptechnology.com wrote:
>> On Thu, 4 Aug 2022 02:24:55 +0200, Klaus Vestergaard Kragelund
>> <klauskvik@hotmail.com> wrote:
>>
>>> On 04/08/2022 00.14, John Larkin wrote:
>>>> On Wed, 3 Aug 2022 14:54:38 -0700 (PDT), whit3rd <whit3rd@gmail.com>
>>>> wrote:
>>>>
>>>>> On Wednesday, August 3, 2022 at 11:03:53 AM UTC-7, sea moss wrote:
>>>>>> Two part question (my google skills have failed me for both):
>>>>>>
>>>>>> 1. Does anyone know approximate pad-to-pad capacitance for the common SMD packages (0201, 0402, 0603, 0805)? Ever measured it? I can't use the parallel plate capacitor formula, since the plates are adjacent, not parallel. (assume there is no copper below the pads)
>>>>>
>>>>> But, regardless of the pads, the component's mating metal surface has that
>>>>> capacitance regardless, doesn't it? You don't need always to make the 'plates' larger
>>>>> than that component feature.
>>>>
>>>> Most surface-mount part capacitance is in the FR4, not in air.
>>>>
>>>>>>
>>>>>> 2. Given an SMD footprint as above, how close does a ground plane need to be (on the same layer) to decrease the effective pad-to-pad capacitance? Is there a formula which gives a good approximation?
>>>>>
>>>>> The use of guard rings can make such capacitance negligible, if that's required, but...
>>>>> trying to simply treat the tiny elements as separable capacitances, isn't a good and useful
>>>>> model.
>>>>
>>>> It's often important to know the three capacitances: each pad to the
>>>> ground plane, and 3-terminal equivalent capacitance pad to pad.
>>>>
>>> turn PCB has a toolkit for a number of different calculations on PCB
>>> structures:
>>>
>>> https://saturnpcb.com/saturn-pcb-toolkit/
>>
>> It's easy to measure attofarads with cheap instruments that a decent
>> home lab should normally have.
>>
>
>Femtofarads, sure, but attofarads? Please explain?
>
>Jeroen Belleman
Femtos, just make a voltage divider from a sig gen, Cx, and the usual
15 pF input of a scope. Signal average a bit or FFT at the low end. A
few fF resolution is easy. To get to aF, add a jfet or an opamp.
One could also add an inductor to resonate out most of that 15 pF.
Lots of ways to do this.
Reply by Jeroen Belleman●August 4, 20222022-08-04
On 2022-08-04 16:08, jlarkin@highlandsniptechnology.com wrote:
> On Thu, 4 Aug 2022 02:24:55 +0200, Klaus Vestergaard Kragelund
> <klauskvik@hotmail.com> wrote:
>
>> On 04/08/2022 00.14, John Larkin wrote:
>>> On Wed, 3 Aug 2022 14:54:38 -0700 (PDT), whit3rd <whit3rd@gmail.com>
>>> wrote:
>>>
>>>> On Wednesday, August 3, 2022 at 11:03:53 AM UTC-7, sea moss wrote:
>>>>> Two part question (my google skills have failed me for both):
>>>>>
>>>>> 1. Does anyone know approximate pad-to-pad capacitance for the common SMD packages (0201, 0402, 0603, 0805)? Ever measured it? I can't use the parallel plate capacitor formula, since the plates are adjacent, not parallel. (assume there is no copper below the pads)
>>>>
>>>> But, regardless of the pads, the component's mating metal surface has that
>>>> capacitance regardless, doesn't it? You don't need always to make the 'plates' larger
>>>> than that component feature.
>>>
>>> Most surface-mount part capacitance is in the FR4, not in air.
>>>
>>>>>
>>>>> 2. Given an SMD footprint as above, how close does a ground plane need to be (on the same layer) to decrease the effective pad-to-pad capacitance? Is there a formula which gives a good approximation?
>>>>
>>>> The use of guard rings can make such capacitance negligible, if that's required, but...
>>>> trying to simply treat the tiny elements as separable capacitances, isn't a good and useful
>>>> model.
>>>
>>> It's often important to know the three capacitances: each pad to the
>>> ground plane, and 3-terminal equivalent capacitance pad to pad.
>>>
>> turn PCB has a toolkit for a number of different calculations on PCB
>> structures:
>>
>> https://saturnpcb.com/saturn-pcb-toolkit/
>
> It's easy to measure attofarads with cheap instruments that a decent
> home lab should normally have.
>
Femtofarads, sure, but attofarads? Please explain?
Jeroen Belleman
Reply by ●August 4, 20222022-08-04
On Thu, 4 Aug 2022 02:24:55 +0200, Klaus Vestergaard Kragelund
<klauskvik@hotmail.com> wrote:
>On 04/08/2022 00.14, John Larkin wrote:
>> On Wed, 3 Aug 2022 14:54:38 -0700 (PDT), whit3rd <whit3rd@gmail.com>
>> wrote:
>>
>>> On Wednesday, August 3, 2022 at 11:03:53 AM UTC-7, sea moss wrote:
>>>> Two part question (my google skills have failed me for both):
>>>>
>>>> 1. Does anyone know approximate pad-to-pad capacitance for the common SMD packages (0201, 0402, 0603, 0805)? Ever measured it? I can't use the parallel plate capacitor formula, since the plates are adjacent, not parallel. (assume there is no copper below the pads)
>>>
>>> But, regardless of the pads, the component's mating metal surface has that
>>> capacitance regardless, doesn't it? You don't need always to make the 'plates' larger
>>> than that component feature.
>>
>> Most surface-mount part capacitance is in the FR4, not in air.
>>
>>>>
>>>> 2. Given an SMD footprint as above, how close does a ground plane need to be (on the same layer) to decrease the effective pad-to-pad capacitance? Is there a formula which gives a good approximation?
>>>
>>> The use of guard rings can make such capacitance negligible, if that's required, but...
>>> trying to simply treat the tiny elements as separable capacitances, isn't a good and useful
>>> model.
>>
>> It's often important to know the three capacitances: each pad to the
>> ground plane, and 3-terminal equivalent capacitance pad to pad.
>>
>turn PCB has a toolkit for a number of different calculations on PCB
>structures:
>
>https://saturnpcb.com/saturn-pcb-toolkit/
It's easy to measure attofarads with cheap instruments that a decent
home lab should normally have.
Reply by sea moss●August 3, 20222022-08-03
On Wednesday, August 3, 2022 at 5:24:55 PM UTC-7, Klaus Kragelund wrote:
> On 04/08/2022 00.14, John Larkin wrote:
> > On Wed, 3 Aug 2022 14:54:38 -0700 (PDT), whit3rd <whi...@gmail.com>
> > wrote:
> >
> >> On Wednesday, August 3, 2022 at 11:03:53 AM UTC-7, sea moss wrote:
> >>> Two part question (my google skills have failed me for both):
> >>>
> >>> 1. Does anyone know approximate pad-to-pad capacitance for the common SMD packages (0201, 0402, 0603, 0805)? Ever measured it? I can't use the parallel plate capacitor formula, since the plates are adjacent, not parallel. (assume there is no copper below the pads)
> >>
> >> But, regardless of the pads, the component's mating metal surface has that
> >> capacitance regardless, doesn't it? You don't need always to make the 'plates' larger
> >> than that component feature.
> >
> > Most surface-mount part capacitance is in the FR4, not in air.
> >
> >>>
> >>> 2. Given an SMD footprint as above, how close does a ground plane need to be (on the same layer) to decrease the effective pad-to-pad capacitance? Is there a formula which gives a good approximation?
> >>
> >> The use of guard rings can make such capacitance negligible, if that's required, but...
> >> trying to simply treat the tiny elements as separable capacitances, isn't a good and useful
> >> model.
> >
> > It's often important to know the three capacitances: each pad to the
> > ground plane, and 3-terminal equivalent capacitance pad to pad.
> >
> turn PCB has a toolkit for a number of different calculations on PCB
> structures:
>
> https://saturnpcb.com/saturn-pcb-toolkit/
Unfortunately Saturn PCB does not cover this... I tried there first.
Reply by Klaus Vestergaard Kragelund●August 3, 20222022-08-03
On 04/08/2022 00.14, John Larkin wrote:
> On Wed, 3 Aug 2022 14:54:38 -0700 (PDT), whit3rd <whit3rd@gmail.com>
> wrote:
>
>> On Wednesday, August 3, 2022 at 11:03:53 AM UTC-7, sea moss wrote:
>>> Two part question (my google skills have failed me for both):
>>>
>>> 1. Does anyone know approximate pad-to-pad capacitance for the common SMD packages (0201, 0402, 0603, 0805)? Ever measured it? I can't use the parallel plate capacitor formula, since the plates are adjacent, not parallel. (assume there is no copper below the pads)
>>
>> But, regardless of the pads, the component's mating metal surface has that
>> capacitance regardless, doesn't it? You don't need always to make the 'plates' larger
>> than that component feature.
>
> Most surface-mount part capacitance is in the FR4, not in air.
>
>>>
>>> 2. Given an SMD footprint as above, how close does a ground plane need to be (on the same layer) to decrease the effective pad-to-pad capacitance? Is there a formula which gives a good approximation?
>>
>> The use of guard rings can make such capacitance negligible, if that's required, but...
>> trying to simply treat the tiny elements as separable capacitances, isn't a good and useful
>> model.
>
> It's often important to know the three capacitances: each pad to the
> ground plane, and 3-terminal equivalent capacitance pad to pad.
>
On Wed, 3 Aug 2022 14:54:38 -0700 (PDT), whit3rd <whit3rd@gmail.com>
wrote:
>On Wednesday, August 3, 2022 at 11:03:53 AM UTC-7, sea moss wrote:
>> Two part question (my google skills have failed me for both):
>>
>> 1. Does anyone know approximate pad-to-pad capacitance for the common SMD packages (0201, 0402, 0603, 0805)? Ever measured it? I can't use the parallel plate capacitor formula, since the plates are adjacent, not parallel. (assume there is no copper below the pads)
>
>But, regardless of the pads, the component's mating metal surface has that
>capacitance regardless, doesn't it? You don't need always to make the 'plates' larger
>than that component feature.
Most surface-mount part capacitance is in the FR4, not in air.
>>
>> 2. Given an SMD footprint as above, how close does a ground plane need to be (on the same layer) to decrease the effective pad-to-pad capacitance? Is there a formula which gives a good approximation?
>
>The use of guard rings can make such capacitance negligible, if that's required, but...
>trying to simply treat the tiny elements as separable capacitances, isn't a good and useful
>model.
It's often important to know the three capacitances: each pad to the
ground plane, and 3-terminal equivalent capacitance pad to pad.
Reply by whit3rd●August 3, 20222022-08-03
On Wednesday, August 3, 2022 at 11:03:53 AM UTC-7, sea moss wrote:
> Two part question (my google skills have failed me for both):
>
> 1. Does anyone know approximate pad-to-pad capacitance for the common SMD packages (0201, 0402, 0603, 0805)? Ever measured it? I can't use the parallel plate capacitor formula, since the plates are adjacent, not parallel. (assume there is no copper below the pads)
But, regardless of the pads, the component's mating metal surface has that
capacitance regardless, doesn't it? You don't need always to make the 'plates' larger
than that component feature.
>
> 2. Given an SMD footprint as above, how close does a ground plane need to be (on the same layer) to decrease the effective pad-to-pad capacitance? Is there a formula which gives a good approximation?
The use of guard rings can make such capacitance negligible, if that's required, but...
trying to simply treat the tiny elements as separable capacitances, isn't a good and useful
model.
Reply by sea moss●August 3, 20222022-08-03
Two part question (my google skills have failed me for both):
1. Does anyone know approximate pad-to-pad capacitance for the common SMD packages (0201, 0402, 0603, 0805)? Ever measured it? I can't use the parallel plate capacitor formula, since the plates are adjacent, not parallel. (assume there is no copper below the pads)
2. Given an SMD footprint as above, how close does a ground plane need to be (on the same layer) to decrease the effective pad-to-pad capacitance? Is there a formula which gives a good approximation?