PSU bracket remake ?

[exxos]

How would the switch and socket mount though on a flat sheet ?

[PhilC]

Good question but maybe I have the answer to that? You can bend acrylic sheet with a hot air gun, or as previously thought, we could 3d print the brackets for the switch and socket, which would be quicker and cheaper to print?

[exxos]

Good question but maybe I have the answer to that? You can bend acrylic sheet with a hot air gun, or as previously thought, we could 3d print the brackets for the switch and socket, which would be quicker and cheaper to print?

But doing that 100 times over? For me just printing them be best as I only have to keep a eye on it, not spend time messing about. In fact I would probably look to get them made, or people print their own.

[PhilC]

If I was you I'd use the print your own option. I'd bet that that bracket would take between 5 to 10 hours to print for each one.

Trying not to be a killjoy but at the same time being honest.

I can give you an estimate of print time and cost by loading the .stl into Cura? And when my workshop is no longer "upside down" would be happy to try printing it?

[sandord]

Just did a quick sketch of how I think a stronger leg could look like. I'm trying to take into account that there are parts underneath the bracket (perhaps even add-ons). I still have to check my ST how much space around the PSU we could additionally claim.

I'm not sure if this is anything close to whay you mean, @exxo.

The leg could be printed separately, which is generally a good idea if that means we can print the largest surface flat on the print bed. The suface only has to be flat on one side though, so the upper extent could still be part of the main surface.

Untitled.png (64.45 KiB) Viewed 5079 times

The ring in the lower hole protects the plastic against wear, otherwise it could break when over-tightening the screw or quickly wear off after taking out the PSU a few times.

I can't help but think this would be better as a perspex sheet with floppy like spacers added where it would mount to the case and then to the power supply.

Possibly, but keep in mind there a four erected parts on the connector end that would have to be connected to the sheet. I think it'll quickly become quite labour intensive.

[sandord]

By the way, I was thinking to change the main surface into a kind of beehive hexagon pattern surface with openings, to save filament and gain maximum structural strength. Perhaps it'll even aid cooling.

Bee_pattern.png (94.42 KiB) Viewed 5078 times

One thing I was wondering about was the shielding function of the original bracket. Perhaps we shouldn't make openings in the surface but paint it with shielding paint instead? I have no idea how significant the electromagnetic effect would be on the parts underneath the PSU...

[exxos]

By the way, I was thinking to change the main surface into a kind of beehive hexagon pattern surface with openings, to save filament and gain maximum structural strength. Perhaps it'll even aid cooling.

That's not a bad idea actually...

One thing I was wondering about was the shielding function of the original bracket. Perhaps we shouldn't make openings in the surface but paint it with shielding paint instead? I have no idea how significant the electromagnetic effect would be on the parts underneath the PSU...

EMI shielding only works in a enclosed metal area... And actually, the bottom of my new power supply PCBs, is a solid mass connected to GND anyway.. So if the metal plate did have any effect on screening, it is actually done on the PCB now anyway..

Actually was thinking we probably don't even need the base plate at all.. Pretty much just need the mounting brackets.. Back section where the switch and connect to go, could probably be a solid block, with some recesses to slot in some stand-off posts... So likely the 5 sections could all be easily printed.. Possibly the back section could even be done all as one piece... From section probably only needs a small adapter brackets.. I don't think there is any real need for the whole area..

The hex nut, I think should be on the bottom side not the top, once the PCB sits on top and the bolt tightened up, I can see pulling out the not for the PCB simply ripping out the plastic... I think it will be better to have the not on the bottom side so when tightened up, it compresses the plastic.. Of course if someone over tightens it may crack.. But as long as it is reasonably solid I don't think it would be a problem.

[sandord]

I think that we at least need a few bars (possibly with a large cross in the center) to give the thing enough rigidity. I don't think a few loose legs that get screwed to the PSU directly will provide enough rigidity.

The suggestion to put the nut on the bottom instead of the top of the post is excellent.

Okay, let me draw a few alternatives tonight, when I get home. I'll make a design with legs only (and a rear plate for the connector/switch) as well.

[exxos]

I think that we at least need a few bars (possibly with a large cross in the center) to give the thing enough rigidity. I don't think a few loose legs that get screwed to the PSU directly will provide enough rigidity.

The back section would be a solid block anyway, so that would be solid... front, yeah, maybe a linking section (left to right) could be good.. could print the front part in one go maybe.

[sandord]

Ran out of time today but I managed to add the rear panels to the original model (need to add 3 more holes though). Posted a picture just to make discussion easier.

Untitled.png (56.16 KiB) Viewed 5035 times

I hope I'll have time to do a simple 3D printing design tomorrow, which of course I will post here.

I'm planning to (based on your feedback):

- turn the rear section into a more solid block
- open up the main surface (replace by bars, change into grid etc.) or remove it altogether
- turn legs into solid blocks, possibly detachable for separate printing
- turn the four PCB support points into solid shapes with nuts mounted in the bottom
- make sure the area below the bracket isn't occupied more than in the original design for maximum compatibility
- make sure the overall size doesn't increase as the shielding doesn't allow it (appears to be the case after a quick glance into the ST)

I just realised that if we want to keep the bracket compatible with existing PSUs, we'll need to find a way to connect the ground wire to the shielding since our bracket will not be made of metal like the original.