Re: brickwalled relays

On a sunny day (Sat, 07 Dec 2024 08:18:48 -0800) it happened john larkin
<JL@gct.com> wrote in <ses8ljd91oq73c7svqmfhr78d4cllcdk9v@4ax.com>:

On Sat, 07 Dec 2024 06:06:01 GMT, Jan Panteltje <alien@comet.invalid>
wrote:
>

On a sunny day (Fri, 06 Dec 2024 13:50:40 -0800) it happened john larkin
<jl@glen--canyon.com> wrote in <geq6ljdsa24e3kg8k1tk0p2hkvlcn7bnct@4ax.com>:
>

>
I put all the relays that would fit on the board.
>
https://www.dropbox.com/scl/fo/pz4axx3pmyl3a6fmyef8e/AOP8q-Z_KeMtiv8TgslL9CQ?rlkey=imu35x1qvt9j8j29if1b1vgm5&dl=0
>
The blue wires were my mistake. I put test points on both ends of two
signals on separate schematic sheets, and somehow that let it pass a
connectivity check. On the other hand, the test points make the
jumpering work nicely.
>
So far, that's all that seems to be wrong.
>
It's our FITS cable fault insertion module, for our P940 modular power
system.

>
What happens when they no longer make those relais and you have to replace once?

>
Certainly we considered that. Several people make the big black ones
and the small white ones and the SSRs. All are industry standards.
>
Some don't wash well in production, but there are workarounds for
that.
>
The poyfuses and some software protections are there to protect the
relay contacts and current shunts and stuff. We specify max voltages
and currents, but you never know if a customer will respect that.
>
We drop the coil voltages from 12 to 8 when we're not transitioning,
to keep the relays a bit cooler. There is guaranteed air flow too.
>
We spell it "relays" here in the developed world.

Zorry, I jam but a neutral net

The Efinix FPGA is cheap, about $10, so we could buy a zillion if it
goes EOL.
>
As one might imagine, the PCB layout was tough on that one.
>
Is there anything else you think we did wrong?

Use relays?