Drilling Holes In Server PSU Output Tabs vs C-Clamp Binding Posts(?)

[Trurl]

There are those that drill holes on the output tabs of server PSUs to attach binding posts(or bolts & nuts) and have no problems.
* Example below of HP HSTNS-PL11/PL30/PL42 BMW Load(50A) Test
()

Then there are those that advise against this because in some PSUs with multi-layered PCBs drilling holes can cause shorting between the layers leading to unexpected faults.
* Example below of a HP DPS-800GB A mod:
( https://pa0fri.home.xs4all.nl/Diversen/DPS-800GB%20A%20Server%20sypply/DPS-800GB%20A%20PSU%20eng.htm )


[ Questions ]
1. Is there a "non-destructive" way to determine if drilling holes is possible (without issues)?

2. Are there miniature c-clamps that can be used instead of binding posts? Most examples I've seen are large c-clamps with cabling for grounding welding machines, but I'm wondering if there are tiny c-clamps for electronics.

My preference is to make a mechanical connection instead of soldering on cables, so that the cable can be removed & adjusted without soldering.
If there are no such "mini c-clamps" on the market, I'm considering just making some myself, by cutting slits in brass rod sections so the PCB slides into the slit, and drill/thread one side of the slit for a screw with a copper plate tip.

[coromonadalix]

for me  i would have soldered a bracket  for the entire width of the pcb trace

or wires instead,  it depend how the company made some servers psu, sometimes under some positives traces you have ground planes

for sure drilling holes may short theses layers,   but if you do clean and sharp edge holes (wich is not easy)   you may still be fully functional,  i did on some pcb  and i saw all the layers (non shorted out)


if this server psu worked fine ... goood  but could not be so lucky for other brand / models,   

but on some  you can find breakout pcbs  to get all the power lines out, even some of them have the power switch etc .... a smal dc meter etc ...


NON DESTRUCTIVE WAYS

not   because unless you go xrays or else,  you dont have the pcb gerbers to verify that, and for sure companies wont give you that

[inse]

Even if you drill a hole without shorting the planes, you would only be able to put a plastic screw through, which will not help a lot.
Why not mount a panel mount version of XT90 connector to the power supply and solder it with decent wires to the terminals.
No problem with strain relief or whatever cabling you need outside.

[Trurl]

I've been eye-ballin' XT90 connectors indeed as they would be good for easy/quick connection & disconnection between the PSU and the hefty cable for the car's engine bay terminals.

I think I'll leave the PSU's output pads as they are, as I think I've found an easier and safer solution (c-clamp alternative), which I realized in the shower!

The small glass shelf where the shampoo sits is mounted to the wall with "glass shelf brackets". As I'd sandwich the PSU's output pads with copper plates anyway when clamping, the gap 5~8mm is fine. And the holes for the wall mount screws can be used instead for bolting on cable lugs.

[inse]

Anodized Aluminum has an isolating surface

[Trurl]

Many of these brackets are apparently made of cast zinc. At any rate, whether made of zinc or anodized aluminum, the bracket would only serve as the mechanical "c-clamp", and the conduction would be through the copper plates (sandwiching the output pads) that will be connected to the cable lug bolts.

The PSU would be secured in a plastic tool case and the clamps on the outputs would be low stress connections. The XT90 connectors would be secured on the tool case* and take the brunt of any cable stress due to connection with the hefty outer cables.

*There are apparently inexpensive panel mount XT90 connectors! There are female versions and flap cover options too (not pictured).


I think I'd use a panel connector for each cable (i.e. + male / - female) as this would ensure proper connection of the cables(both black), and as the max amperage of the PSU would be just under 100amps (The XT90 connectors are rated for 40amps continuous/90amps peak. I would likely never need to use the PSU's max amperage, but for peace of mind... and these connectors are so inexpensive).

[Monkeh]

With a little measurement and hunting around (perhaps check on the backplane these things plug into, for example..) you can just buy the connector which fits on the end.

[Trurl]

I think you mean something like the following?

https://colintd.blogspot.com/2016/11/busbars-on-output.html?view=magazine

https://colintd.blogspot.com/2016/11/completed-psu.html?view=magazine

I considered the idea which IS nice. But I need to minimize the overall length as the PSU will be fitted into a plastic tool case and also minimize stress on the protruding PCB.

[enut11]

I have use several server PSU and most of them I drilled the PCB and fitted metal bolts. If you drill a slightly oversized hole and clean it up with a fine round file you can then use a small piece of heatshrink over part of the bolt to isolate it from any internal PCB tracks.
enut11

[inse]

This maybe kinda works, but heatshrink is not pressure or piercing resistant.
You would need a hard plastic sleeve for a reliable isolation.

[Trurl]

Thank ya'll for your input. Have a great New Year!

[pienari]

I have soldered 4mm bras hex nuts good results.

[Trurl]

That's another great idea!

The reason I've been seeking mechanical connections (that sandwich/compress the output tabs) is concerns of the output tabs(copper lands) lifting off the PCB especially if considerable heat is needed for soldering bulky cables or binding hardware. Perhaps this concern is not relevant for such large copper lands...
I could of course preheat such relatively bulky brass nut(away from the output tabs/PCB) with my oxy-butane torch before placing it on "pre-tinned" output tabs... hmmm...

I'll keep the idea in mind... Thanks!

[pienari]

[Trurl]

Cool~! I  was thinking vertical(short stubby nut & bolts to minimize blocking of the air inlets), but horizontal seems fine too!

I've read about folks using a capacitor on the output to reduce ripple & noise down to 10s of mV or lower...

What is the effect of the capacitor on your PSU's output and what is the logic behind selection of appropriate or optimal capacitance (and any other pertinent spec)? 

[pienari]

This model ovp is kicking at 13.6v and voltage is set near that point.
Cap helps a bit with high current spikes but not a problem.