Re: Interview Question (your Sunday ruined part 2)

On Wed, 23 Oct 2024 04:22:16 GMT, Jan Panteltje <alien@comet.invalid>
wrote:

On a sunny day (Tue, 22 Oct 2024 10:48:01 -0700) it happened john larkin
<jl@glen--canyon.com> wrote in <b2pfhjl4hbd1013mp1bkamudjm948jl7qb@4ax.com>:
>

On Mon, 21 Oct 2024 15:31:45 -0400, "Edward Rawde"
<invalid@invalid.invalid> wrote:
>

"john larkin" <jl@glen--canyon.com> wrote in message news:jgadhjl2uj77q3gdls2tfjrdu5qe8j9vgp@4ax.com...

On Mon, 21 Oct 2024 17:24:35 +0100, liz@poppyrecords.invalid.invalid
(Liz Tuddenham) wrote:
>

john larkin <JL@gct.com> wrote:
>

On Mon, 21 Oct 2024 11:57:33 +0100, liz@poppyrecords.invalid.invalid
(Liz Tuddenham) wrote:
>

Edward Rawde <invalid@invalid.invalid> wrote:
>

"john larkin" <JL@gct.com> wrote in message

news:risahjpn2kfue87dqm6t2461a966ncv9in@4ax.com...
>
[...] > >

Maybe european decimal points slide off drawings and fall on the
floor. Robust American decimal points and schematic connect dots don't
do that.

>
I wasn't aware that there was a difference between European and american
decimal points.

>
There is a difference between English-speking ones and French-speaking
ones: the French ones look like commas.

>
Do French engineers use the same convention on their schematics?

>
I don't know, but the comma is in everyday use as a decimal point.
>

I can't recall ever seeing a French schematic, actually.

>
I can - and it was dreadful !  It was of a laboratory R.F. signal
generator and they had put the earthing point in the centre with the
various sections of the circuit radiating like the spokes of a wheel
around it.  The HT line ran around the outside.
>
I don't think they do it like that nowadays.

>
There are people, even gigabuck organizations, who are obsessed with
single-point grounding.

>
Decoupling capacitors such as 0.1uF everywhere can also be an obsession.
Remove them from many (not all) circuits with no change in performance.
>

>
I knew one guy that claimed that bypass caps weren't necessary on
multilayer boards. His stuff worked.
>
Most electronics is way over-bypassed. Every data sheet assumes that
their part is the center of the universe and deserves a string of
stepped-value bypass caps on every supply.

>
I have also read about a case where they left out some bypass caps
and then those boards caused big troubles.
https://electronics.stackexchange.com/questions/630354/what-may-happen-if-i-leave-out-bypass-capacitors
Be very careful, it may work once (leaving them out),
but your products may fail..

Some fast parts, like CPUs and FPGAs, have substantial internal bypass
caps. I've measured as much as 1 uF on the core power of an FPGA, and
numbers like 100n on aux supplies. It seemed to be on-chip,
monolithic.

I wonder if DRAMs chips typically have internal bypass caps. Somebody
here could measure a few.

It's worth measuring a chip, to help plan bypassing. Some ugly Intel
CPU may need sheer joules of chunky caps for wakeup power surges.

I've done TDR experiments on multilayer boards. A proper power pour
looks like a perfect cap, and adding more caps, anywhere on the pour,
just makes it look like a bigger perfect cap.