A basic shielding question
 

On an example like a simple homebrew 2 tube regen set it is typically
recommended to use a metal front panel to "shield" the ckt from hand
effects. However, you can still notice some effect by touching the
ground/panel.

Would one see any advantage by using double-sided PCB with the two sides
not electrically connected?

-Bill

[Hand capacitance detuning receiver]
double-sided PCB with the two sides
not electrically connected?

My gut feeling is that hand capacitance (maybe tens of pF) is small
compared to the capacitance between the two layers of PCB (hundreds of
pF) and that the shielding effect would be negligible.

Now, "guard ring" style shielding (as used in high-impedance PCB
layout) might be worthwhile, but would be more difficult to implement.

Tim.

Tim Shoppa wrote:

[Hand capacitance detuning receiver]
double-sided PCB with the two sides
not electrically connected?


My gut feeling is that hand capacitance (maybe tens of pF) is small
compared to the capacitance between the two layers of PCB (hundreds of
pF) and that the shielding effect would be negligible.

Thanks for the reply. Here's the logic that is making me ponder this.
The capacitance of the PCB would be in 'series' with the hand
capacitance resulting in a lower overall capacitance.

If I plug in numbers like 200pf for the board and 20pf for the hand I
see that the numerical improvement is minimal but I wonder if it can be
expected to work that way?

-Bill

-exray- wrote:

Tim Shoppa wrote:

[Hand capacitance detuning receiver]
double-sided PCB with the two sides
not electrically connected?



My gut feeling is that hand capacitance (maybe tens of pF) is small
compared to the capacitance between the two layers of PCB (hundreds of
pF) and that the shielding effect would be negligible.


Thanks for the reply. Here's the logic that is making me ponder this.
The capacitance of the PCB would be in 'series' with the hand
capacitance resulting in a lower overall capacitance.

If I plug in numbers like 200pf for the board and 20pf for the hand I
see that the numerical improvement is minimal but I wonder if it can be
expected to work that way?

-Bill

Yes, but there's always an outside chance that it'll do some good.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

-exray- wrote:

On an example like a simple homebrew 2 tube regen set it is typically
recommended to use a metal front panel to "shield" the ckt from hand
effects. However, you can still notice some effect by touching the
ground/panel.

Would one see any advantage by using double-sided PCB with the two sides
not electrically connected?

-Bill

If the chassis of the set is not referenced to the same ground that you
are then touching the chassis will change the effective capacitance of
the tank circuit. This effect will probably be greater if the antenna
is capacitively coupled rather than inductively. The two things that
I'd do before experimenting with the PC board would be _first_ to make
sure that the chassis of the receiver was brought to a good ground in
the station, and second to consider putting the entire receiver into a
shielded box.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

No. A single side is adequate.

What causes the change to occur when you touch the panel is that
oscillator current is flowing on the outside of the panel. The way to
avoid this is to completely enclose the oscillator with a shield.

Roy Lewallen, W7EL

-exray- wrote:
On an example like a simple homebrew 2 tube regen set it is typically
recommended to use a metal front panel to "shield" the ckt from hand
effects. However, you can still notice some effect by touching the
ground/panel.

Would one see any advantage by using double-sided PCB with the two sides
not electrically connected?

-Bill

Sorry Tim, I beg to differ....

True, the tank circuit frequency gets dragged around by "hand"
capacitance, but the place where ths effect is felt is almost
invariably the capacitance between ANTENNA, CHASSIS, BODY and GROUND
rather than direct hand-to-LC.

Better grounding is nearly useless, unless you bolt the chassis to a
very good, high capacitance ground - e.g. you bolt it to a steel bridge
on a ship. But even then, your hand and body may introduce a change
in antenna-to-ground capacitance - back to square one!

An extremely loose coupling will reduce this effect, but it's very
hard to pull off while keeping an efficient energy transfer.

It's better if you use an active buffer stage. Nowadays you rarely see
solid state regens without a buffer, and the reason is not signal
levels - it's that capacitance decoupling through an active device can
be very effective, making both frequency and regeneration more
controllable.

Another trick, less practical, is to move the antenna far away, and
feed it via coax. In this case, grounding the coax somewhere along the
run (at zero impedance if you route the coax by a good RF ground point)
is very effective at stabilizing antenna capacitance to ground. In
this case you will indeed see almost exclusively the effect of
hand-to-LC, which won't amount to much and can indeed be fully killed
by a shielded enclosure.

On 12 May 2005 15:41:28 -0700, SpamHog wrote:
Sorry Tim, I beg to differ....

About WHAT!?!? We see no quoted text here.

Oh, we understand: X-Trace: posting.google.com......
Just another sloppy groups.google posting.

SpamHog wrote:

Sorry Tim, I beg to differ....

True, the tank circuit frequency gets dragged around by "hand"
capacitance, but the place where ths effect is felt is almost
invariably the capacitance between ANTENNA, CHASSIS, BODY and GROUND
rather than direct hand-to-LC.

Better grounding is nearly useless, unless you bolt the chassis to a
very good, high capacitance ground - e.g. you bolt it to a steel bridge
on a ship. But even then, your hand and body may introduce a change
in antenna-to-ground capacitance - back to square one!

An extremely loose coupling will reduce this effect, but it's very
hard to pull off while keeping an efficient energy transfer.

It's better if you use an active buffer stage. Nowadays you rarely see
solid state regens without a buffer, and the reason is not signal
levels - it's that capacitance decoupling through an active device can
be very effective, making both frequency and regeneration more
controllable.

Another trick, less practical, is to move the antenna far away, and
feed it via coax. In this case, grounding the coax somewhere along the
run (at zero impedance if you route the coax by a good RF ground point)
is very effective at stabilizing antenna capacitance to ground. In
this case you will indeed see almost exclusively the effect of
hand-to-LC, which won't amount to much and can indeed be fully killed
by a shielded enclosure.

Goodness knows why, but I was assuming a nice coax feed from the
antenna. While that normally makes sense it can't be assumed with a
regen, I suppose.

If you're using the regen to receive CW or SSB you should be using an
amplifier anyway.

-------------------------------------------
Tim Wescott
Wescott Design Services
http://www.wescottdesign.com