Re: HP EliteBoard G1a: A computer inside a keyboard - YouTube

On Tue, 1/13/2026 5:36 AM, Mr. Man-wai Chang wrote:
> On 9/1/2026 2:31 am, Paul wrote:
>>
>> The keyboard is not pressed against your head.
>>
>> I think you're safe.
> 
> It's still better to do things the good old way, that's a computer in a grounded metal chassis a feet or more away from me. It's MUCH MUCH MUCH safer! :)
> 
> Can glass block EMF radiation?? The industry is pushing glassed computer chassis.
> 

For systems not intended to emit an RF signal, just the
PCB is sufficient to block interference. The clock signal
is buried on an inner layer and there is a copper plane on
the outside at ground potential.

Computer case designs have been "leaky" for a lot of years,
and we do not rely on the computer case as a shield. Containment
starts at the PCB level. The Y caps and other filter components
on the inlet side of the power supply, they're intended to prevent
some noise-on-mains from escaping.

And the products are tested for emissions.

The USB3 cable can emit under some circumstances, because it's
a 5GHz signal, it emits a broad spectrum centered at 2.5Ghz.
This can wipe out a Bluetooth transceiver. If your Bluetooth
can send files to a second Bluetooth equipped computer, this is
a good indication that your USB3 cables are NOT emitting :-)
You can then unplug your Bluetooth nano transceiver and stop using it.
My Bluetooth nano transceivers are unplugged right now, but they
sit next to their computer and can be plugged in when needed.

*******

If you want to see a test instrument for electronics, where the casing
is a necessary part of the filtering, that is the Spectrum Analyzer.
It uses a honeycomb cover on the air path, to block external RF from
getting inside the box.

   https://www.uaf.com/solutions/dual-emi-air-filters/

It uses beryllium copper gaskets between cover plates and chassis,
to make a Faraday cage around the electronics. The noise floor inside
the Spectrum Analyzer is -135db to -150db or so.

This allows the instrument to make measurements of external signals
on a scope cable or similar, and any other stray noises do not upset
measurements the instrument makes. It does allow the instrument to be
air cooled, but it is pretty hard to push air through the physical
packaging when so much effort goes into sealing up holes where interference
can get in.

Computers, they only have to meet FCC part 15, and they don't pass with
exceptional margin. The computer companies expect their engineers to make
good PCBs, such that a finished product only has to be sent to the
third party testing lab one time. That's the main reason the PCBs filter
as well as they do, is the manufacturer does not intend to waste a lot of
money "forcing" the design to meet the spec. What we used to do, many years
ago, is "improve the case design to stop emissions" which is expensive.
If you can stop emissions at the PCB level, it reduces the need for
"RF tight" packaging on the outside. That is why a "glass" computer
case works.

   Paul