# How yo detect capacitive power coupling

Started by March 10, 2016
```Could some electronics guru please help ?
Suppose a high frequency(approx. 30 - 40
Mhz) oscillator connected across a aimple
1 cm x 1 cm plate area parallel plate
capacitor, with a 1 mm thick dielectric
plate in between, material FR4. Now how
do I detect that indeed the signals are
being transmitted between the plates,
given that a capacitor is a short to AC.

Also, if I connect a simple rectifier
diode e.g., 1N4007 after the capacitor,
and then a small electrolytic capacitor,
after that in the signal return path,
would the capacitor indicate a DC voltage
across it ?

I have done some experiments regarding
this idea, but am confused by the results.

Any hints, suggestions would be welcome.
```
```On Thursday, 10 March 2016 05:54:25 UTC, daku...@gmail.com  wrote:

> Suppose a high frequency(approx. 30 - 40
> Mhz) oscillator connected across a aimple
> 1 cm x 1 cm plate area parallel plate
> capacitor, with a 1 mm thick dielectric
> plate in between, material FR4. Now how
> do I detect that indeed the signals are
> being transmitted between the plates,
> given that a capacitor is a short to AC.

detect current flow by detecting V drop across an R
detect low impedance of the FR4 cap, eg by feeding it frmo a 50ohm source and noting the voltage drops low
etc

> Also, if I connect a simple rectifier
> diode e.g., 1N4007 after the capacitor,
> and then a small electrolytic capacitor,
> after that in the signal return path,
> would the capacitor indicate a DC voltage
> across it ?

You'd need a rectifier that lets current flow both ways in the FR4. So minimum 2 diodes.

> I have done some experiments regarding
> this idea, but am confused by the results.
>
> Any hints, suggestions would be welcome.

NT
```
```First of all, if you just hook a diode to a cap like that, it will only put out any voltage until the cap charges up. If you ever have to rectify off a cap you need to either use a doubler, or either a coil or resistor to ground, and the latter methods are not efficient but might be useful for instrumentation.

Since there was a first of all, second of all, to what end are you doing this ? Building some secret weapon ? If so count me in.

Or are you interested in the dilectric properties of FR-4 ?

Might help if we know what you're shooting for here.
```
```daku...@gmail.com wrote:
>
>
> Suppose a high frequency(approx. 30 - 40
> Mhz) oscillator connected across a aimple
> 1 cm x 1 cm plate area parallel plate
> capacitor, with a 1 mm thick dielectric
> plate in between, material FR4.

** Makes a 4pF cap.

http://www.daycounter.com/Calculators/Plate-Capacitor-Calculator.phtml

> Now how
> do I detect that indeed the signals are
> being transmitted between the plates,
> given that a capacitor is a short to AC.
>

** It's a 1070 ohm impedance at 40MHz.

>
> Also, if I connect a simple rectifier
> diode e.g., 1N4007 after the capacitor,

** Not much use at RF.

You are not improving.

...  Phil
```
```On Wednesday, March 9, 2016 at 9:54:25 PM UTC-8, daku...@gmail.com wrote:

> Suppose a high frequency(approx. 30 - 40
> Mhz) oscillator connected across a aimple
> 1 cm x 1 cm plate area parallel plate
> capacitor, with a 1 mm thick dielectric
> plate in between, material FR4. Now how
> do I detect that indeed the signals are
> being transmitted between the plates...

Are you looking for transmission of power?  Or, for
a signal of some sort?  Or multiple signals?

Power metering requires both a magnetic and electric
field sensor, and access to the rest of the circuit, not just one
two-terminal component, unless you want to know the
power that's being dumped into heat in the dielectric "FR4".
```
```dakupoto@gmail.com wrote:

> Suppose a high frequency(approx. 30 - 40
> Mhz) oscillator connected across a aimple
> 1 cm x 1 cm plate area parallel plate
> capacitor, with a 1 mm thick dielectric
> plate in between, material FR4. Now how
> do I detect that indeed the signals are
> being transmitted between the plates,
> given that a capacitor is a short to AC.
>
> Also, if I connect a simple rectifier
> diode e.g., 1N4007 after the capacitor,
> and then a small electrolytic capacitor,
> after that in the signal return path,
> would the capacitor indicate a DC voltage
> across it ?
A 1N4007 will not perform well at 30+ MHz.  You might try a 1N4148, or even
a Schottky diode.

Jon
```
```>"A 1N4007 will not perform well at 30+ MHz.  You might try a 1N4148, >or even
>a Schottky diode. "

Now that you mention it... In fact if he tries to put any power through it, he will get smoke likely even way below its design limits.

For measuring, he is probably best off with a pair of Schotkeys configured as a doubler. And of course a small correction factor for their Vf, which will be lower than other high speed diodes. And we don't even know how much power he wants to try to push through this thing. If he does not know the 1N4007 won't work, he might not have much of an idea about this - yet.

If this is a purely academic experiment that is one thing. but if there is an application involved it really helps to know what that is. He might be trying to build a capacitively coupled battery charger, which is not the worst idea in the world. If I had to do that I would lean toward an inductively coupled system though. Trying to get power through a capacitor like that implies he wants a high degree of isolation, which may well be better provided by a transformer. Some electric cars charge by that method IIRC. I was thinking, years ago, of a cordless phone that would be totally waterproof. Some sort of shrink wrap around the whole thing but with the coil embedded. Might have been able to even make a kit, or retrofit or something for existing models. But at this stage of the game they got them pretty much done. A buddy of mine has a cellphone like that, says you can just dump it in a bucket of water and nothing happens.

I do believe the capacitive approach to that would result in enough EMI that the FCC sends the SWAT team to your door though. They DO have a SWAT team.

Now a guy starts a thread bitching about me asking about the purpose of this whole endevour. HA. (they are tearing him up...)
```
```On Friday, March 11, 2016 at 9:24:19 AM UTC+5:30, jurb...@gmail.com wrote:
> >"A 1N4007 will not perform well at 30+ MHz.  You might try a 1N4148, >or even
> >a Schottky diode. "
>
> Now that you mention it... In fact if he tries to put any power through it, he will get smoke likely even way below its design limits.
>
> For measuring, he is probably best off with a pair of Schotkeys configured as a doubler. And of course a small correction factor for their Vf, which will be lower than other high speed diodes. And we don't even know how much power he wants to try to push through this thing. If he does not know the 1N4007 won't work, he might not have much of an idea about this - yet.
>
> If this is a purely academic experiment that is one thing. but if there is an application involved it really helps to know what that is. He might be trying to build a capacitively coupled battery charger, which is not the worst idea in the world. If I had to do that I would lean toward an inductively coupled system though. Trying to get power through a capacitor like that implies he wants a high degree of isolation, which may well be better provided by a transformer. Some electric cars charge by that method IIRC. I was thinking, years ago, of a cordless phone that would be totally waterproof. Some sort of shrink wrap around the whole thing but with the coil embedded. Might have been able to even make a kit, or retrofit or something for existing models. But at this stage of the game they got them pretty much done. A buddy of mine has a cellphone like that, says you can just dump it in a bucket of water and nothing happens.
>
> I do believe the capacitive approach to that would result in enough EMI that the FCC sends the SWAT team to your door though. They DO have a SWAT team.
>
> Now a guy starts a thread bitching about me asking about the purpose of this whole endevour. HA. (they are tearing him up...)

Thanks to each of you for your feedback. I have worked with inductive power coupling, and now I am experimenting with
capacitively coupled power transfer, for example, for battery
charging. I am aware that 1N4007 is not a good chpice, but at
this stage, I am more interested in ensuring that capacitive
coupling works, before going to the next step.
```
```On 3/10/2016 11:53 PM, dakupoto@gmail.com wrote:
> On Friday, March 11, 2016 at 9:24:19 AM UTC+5:30, jurb...@gmail.com wrote:
>>> "A 1N4007 will not perform well at 30+ MHz.  You might try a 1N4148, >or even
>>> a Schottky diode. "
>>
>> Now that you mention it... In fact if he tries to put any power through it, he will get smoke likely even way below its design limits.
>>
>> For measuring, he is probably best off with a pair of Schotkeys configured as a doubler. And of course a small correction factor for their Vf, which will be lower than other high speed diodes. And we don't even know how much power he wants to try to push through this thing. If he does not know the 1N4007 won't work, he might not have much of an idea about this - yet.
>>
>> If this is a purely academic experiment that is one thing. but if there is an application involved it really helps to know what that is. He might be trying to build a capacitively coupled battery charger, which is not the worst idea in the world. If I had to do that I would lean toward an inductively coupled system though. Trying to get power through a capacitor like that implies he wants a high degree of isolation, which may well be better provided by a transformer. Some electric cars charge by that method IIRC. I was thinking, years ago, of a cordless phone that would be totally waterproof. Some sort of shrink wrap around the whole thing but with the coil embedded. Might have been able to even make a kit, or retrofit or something for existing models. But at this stage of the game they got them pretty much done. A buddy of mine has a cellphone like that, says you can just dump it in a bucket of water and nothing happens.
>>
>> I do believe the capacitive approach to that would result in enough EMI that the FCC sends the SWAT team to your door though. They DO have a SWAT team.
>>
>> Now a guy starts a thread bitching about me asking about the purpose of this whole endevour. HA. (they are tearing him up...)
>
> Thanks to each of you for your feedback. I have worked with inductive power coupling, and now I am experimenting with
> capacitively coupled power transfer, for example, for battery
> charging. I am aware that 1N4007 is not a good choice, but at
> this stage, I am more interested in ensuring that capacitive
> coupling works, before going to the next step.
>

At this stage it is worth picking a different diode.
100mm x 100mm gets you down to 10 ohms series Xc,

this might get you some decent current for charging.

FR4 has about 80 times the loss of polystyrene.

Rogers 5880 has 1/16 the loss of FR4.

Mikek
```