[sdiy] Converting rack mount to DC input?

[cheater cheater]

One way to have DC input, but retain the capability of using the
internal power supply, is to have a 4P2T switch for configuration.

It seems that 4P2T switches are super expensive, especially if they're
supposed to withstand, say, 1A.

So I started looking around. TLDR: classic pcie 8-pin power connectors
are probably the best, and also the cheapest.

My first thought was to look at PCIE x1 slots. The cheapest advanced
connector on earth, for 0.4 Euro at unit price, you get a connector
that handles 1.1A per pin as per eg this document, page 6, 4.4

https://cdn.amphenol-cs.com/media/wysiwyg/files/documentation/gs-12-233.pdf

> 4.4 CONTACT CURRENT RATING
> 1.1 amp per contact minimum per EIA-364—70, method 2 and PCI Express Connector High Speed Electrical Test Procedure. The temperature rise shall not exceed 30 degree C. Ambient condition is still air at 25°C.

The connector has 36 pins, so while I don't think it would handle 36A,
I'm sure it would handle something like 2A, especially if you share
pins. You can even leave pins empty to prevent shorts during
insertion.

You could put the connector out the back, and either insert a plug-in
"card" (really just a small edge connector with wires soldered on)
that provides DC, or a pass-through "card" that shorts some pins
together to carry power from the internal power supply. And now
instead of $20-30 per unit, this costs $1 per unit. My main question
is how I would fix the connector, but maybe a simple screw hole in the
connector that mates with a threaded hole in the case could do the
trick.

A pcie x1 port is 25mm long, so it can fit upright in the back of a 1U
rack unit, which is 45mm, so it doesn't take up much space either.

The cheapest one that can be found at Mouser currently and can be
bought in low volumes is roughly 0.4 Euro.

https://www.mouser.at/ProductDetail/Amphenol-FCI/10018783-10200TLF?qs=V%252BXmToedwojeZUI4fPwmPA%3D%3D

Qty.  Unit Price  Ext. Price
1     € 0,439     € 0,44
10    € 0,372     € 3,72
25    € 0,332     € 8,30
100   € 0,316     € 31,60

By making the connector require a dummy plug to connect the internal
power supply into the circuit, it makes it impossible to connect both
DC power and mains AC, so that makes the design intrinsically safe
without using switches.

Alternatively to a PCIE connector I could use some panel mount plug
with 8 pins, have 4 pins for DC input, and have the other 4 pins carry
power from the internal power supply, and similarly use a shorting
plug to use the internal power supply.

Circular DIN connectors start at about 8 Euro per pair of socket and plug.
Circular metric aren't better.
MIL Spec connectors seem to be cheaper. But the mouser search sucks,
so I can't find them by number of pins.

PCIE 8-pin power connectors are real, real cheap - 10 cents a piece.
They're latching (no need for screws and stuff). There doesn't seem to
be a panel mount version, but one could mount a through-hole connector
to a pcb, and have the pcb have screw holes for mounting to the rear
panel. They handle high power - 150W meaning 3A per pin (half the pins
are return pins). Seems like a winner to me. They're also tiny so they
will easily fit in the back of a 1U unit even vertically. The biggest
pain here might be making the rectangular hole and having it look any
good. might be a case for custom die, maybe a small steel job that
uses two bolts to screw together the two cutting parts through the
sheet metal.

"Pin and socket" connectors are the same kind of thing, just not
specifically PCIE 8-pin. Still cheap at roughly 20 cents a piece...
useful alternative if more than 4 rails are necessary.

D-sub connectors carry up to 3A... and are very cheap and can do panel
mount of some sort. But I'll be damned if I use one of those cursed
things for power.

Automotive connectors seem inexpensive too, but I haven't really
looked into them much other than a quick parametric search.

I wonder what everyone thinks of this.

On Tue, Jul 7, 2026 at 6:39 AM cheater cheater
<cheater00social at gmail.com> wrote:
>
> Don't you think this may have had something to do with you
> distributing AC power, rather than DC with 0V potential?
>
> On Sat, Jul 4, 2026 at 11:33 PM Mattias Rickardsson <mr at analogue.org> wrote:
> >
> > Den lör 4 juli 2026 14:23Florian Anwander via Synth-diy <synth-diy at synth-diy.org> skrev:
> >>
> >> Maybe, it's a stupid thought, but...: could it be that a device relies on beeing galvanical separated from other devices. I this case the common supply might clash with the devices concept.
> >
> >
> > This reminds me of a somewhat similar issue I had many years ago:
> >
> > I wanted a Korg MS-20 and a Korg KR-55 to share one external power puck. (Swedish-sold old Korgs had a big external power transformer to step down from 220 V AC to 16 V AC, which is then what the instrument then accepts via a plug with 2 flat prongs that doesn't seem roadworthy with European electrical standards.)
> >
> > So I made a short splitter Y cable on the 16 V AC side and attached both instruments and turned them on. But then I connected an audio cable between them (or audio cables from both units to the same mixer) and I had a blown fuse! It was apparently not okay to connect those two grounds together when running off the same AC supply.
> >
> > /mr
> >
> > ________________________________________________________
> > This is the Synth-diy mailing list
> > Submit email to: Synth-diy at synth-diy.org
> > View archive at: https://synth-diy.org/pipermail/synth-diy/
> > Check your settings at: https://synth-diy.org/mailman/listinfo/synth-diy
> > Selling or trading? Use marketplace at synth-diy.org