Log 3LVIS: (3) (L)itre (V)ery small case with (I)nternal psu and (S)creen

[BaK]

The Swiss plugs (as much headache as they caused me when I moved there ages ago) are obviously of advantage here, just like UK plugs, because you know which wire is live or neutral.

I think the inlet will fit easily to the right of the I/O shield and the switch on top of it where the WiFi antenna holes are located. One problem solved, ready to tackle the next one...

I can confirm for the C6, it's a bit tight for the switch but that should do it!
How are you going to connect the cables to them? I don't recommend terminals, such as the following, because they will add to the internal depth.

Soldering will be better about that as you can solder the cables at a right angle, parallel to the back of the case. But that will leave dangerous contacts accessible to the touch, unlike the above terminals and their protective cap.
As right angles connector/cable are hard to protect properly with an heatshrink sleeve, liquid electrical tape could come in handy at that point.

I was more thinking of something like in your drawing with the red hinge, mounted on the front side. But with a slight change: the side of the new top would be a little bit lower so as to cover the hinge from side

That will let you have a flushed top and an hinge only visible on the front? Sounds good!

I think it should be possible to get a good result with flat hinges like these

Can't wait to see how you implement that! If it's still on your list that is...

 

[infoberg]

I can confirm for the C6, it's a bit tight for the switch but that should do it!
How are you going to connect the cables to them? I don't recommend terminals, such as the following, because they will add to the internal depth.

Soldering will be better about that as you can solder the cables at a right angle, parallel to the back of the case. But that will leave dangerous contacts accessible to the touch, unlike the above terminals and their protective cap.
As right angles connector/cable are hard to protect properly with an heatshrink sleeve, liquid electrical tape could come in handy at that point.

I was planning on soldering the wires to the inlet and connect them with insulted termimals to the switch and from the switch to the PSU. This would fit with the intended placement of the parts.

This is just an approximation of the placement. The ground connector of the inlet will be facing to the edge of the case and connect to the motherboard stand with a ring terminal.

There's basically two option for placing the switch. Either (1) on the left covering two ventilation slots and (barely) the 8mm hole, or (2) to the right between the 8mm hole and the 6mm hole on the right, covering the middle hole.

Not sure which one I will end up choosing. I will do the cutout for the inlet first and then decide if the stability of the case allows placing the switch above it (option 1).

That will let you have a flushed top and an hinge only visible on the front? Sounds good!
Can't wait to see how you implement that! If it's still on your list that is...

Yes, it's still on my list, even if I have no idea yet how to do it. But that's part of this journey

Next steps:​

1. case modification for inlet and switch
2. wait for a replacement case to arrive because I ruined the case in step 1
3. preparing all the necessary cables in the correct length (and waiting for connectors to arrive)
4. testing PSU and PDCB with the PSU tester
5. if all goes well: hooking it all up: PSU, LEDs, power switch, PDCB and modified USB3 ports
6. lots of thinking, googling and learning to tackle the monitor case lid...

 

[infoberg]

Time for a little update

Part 4: The first cut is the deepest...​

That's not only a beautiful song by Yusuf Islam (TAFKACS: The Artist Formerly Known As Cat Stevens), but also the repeating theme of my first case-cutting adventure. The first cutout I tackled was the one for the C6 inlet. At first I used a cutting disc, but they are a lot larger then the needed cutout, so the first cut I did went too deep. But luckily it wasn't too bad, it will be hidden by the inlets mounting frame. As the inlet consists of 3 overlapping circles, I then figured it might be done with a drill bit for the Dremel. But I came to deeply regret not following a friendly advice to get myself a vice for the metal works. Holding the case with one hand, the drill slipped when it touched the edge of the hole and the second "cut" went right into my hand. Lucky a second time, the hole I drilled into my hand missed any blood vessels, tendons or bones, so it will just be a nice little souvenir.

After giving it two or three days to heal and think about my clumsiness, I managed to get the job done with the help of a clamp to fixate the case while working on it. The result looks ok when the inlet is mounted. Next task was the cutout for the tiny switch. By the way, this switch which I took from the C14 inlet with integrated switch, is the smallest one I could find, even on AliExpress. Its outer dimensions are 21mmx15mm, cutout will be about 19mmx13mm, so not a big margin of error allowed, not more than 1mm to spare on each side before the switch is doomed to fall into the case.

It turned out that the rectangular cutout was a little bit easier and all went well, I managed to make only one hole in the case and none in myself. A little filing, sanding and spray painting the case made it look like new

After that I started getting all the cables ready with proper length, connectors and shrink tubing: LEDs, power button and the cables for the inlet and switch.

​

the future: integrated screen​

During recovery I had some time to think about the screen mounting options. I ordered an aluminium frame which would make the case higher when mounting it on top, but on the other hand it could be mounted with hinges. Another option might be just mounting the monitor on some aluminium bars running across the case, but then it could not be lifted up.
I also tried to find some thin hinges, but I wasn't too successful. Looking for lifting hinges as suggested by BaK, but they all were too thick and couldn't fit inside the case. After hours and hours of googling and searching I found only two viable options so far:

1) PVC hinges:
These are quite reasonably priced, about 24 Euros (or 28 USD) including delivery for a pack of 4. At a length of 38mm this would be enough to cover most of the front. The thickness is specified as 2mm, I guess that means 4mm when folded (as they need to be). So I would definitely have to hide those hinges behind some sort of cover running around the edges.

2) Kevlar hinges:
These are just 0.8mm thick, which might just be ok if used without a cover to hide them. They come in a length of 50mm or 100mm and are quite expensive. Ordering 2x 100mm to cover the complete length of the case will cost 110 Euros (about 130 USD) and even using only 2x 50mm hinges (e.g. on both edges of the case) will cost 84 Euros (about 100 USD), which would make the hinges the second most expensive component of this build, right after the CPU...

I ordered the PVC hinges to see how far they will get me, but I just can't get those beautiful kevlar hinges out of my head
To close the case on top, I will probably order an aluminium panel with a cutout for the screen, holes in the corners for mounting and ventilation holes. But before getting the necessary measurements, I will need to come to a decision on wheter to use the frame or not...

Next steps:​

1. case modification for inlet and switch: done
2. wait for a replacement case to arrive because I ruined the case in step 1 : surprisingly enough this wasn't necessary
3. preparing all the necessary cables in the correct length (and waiting for connectors to arrive): done, except for the the PDCB input cable
4. testing PSU and PDCB with the PSU tester
5. if all goes well, hooking it all up: PSU, LEDs, power switch, PDCB and modified USB3 ports
6. testing the computer for performance, thermals and airflow
7. lots of thinking, googling and learning to tackle the monitor case lid...: halfway done, I hope

 

[Valantar]

Looks great! IMO you did an excellent job on those cutouts. The spray paint probably helped with that impression, but that's also part of the job Looking forward to seeing this move forward, that monitor mounting is going to be interesting!

 

[BaK]

That's fun to see someone else going through the same challenges!

Except...

the second "cut" went right into my hand

Ouch! You've learnt your lesson the hard way!

I think I did all the C6 cutout with a round file removing metal around the stock hole of the case, much safer, I should have told you...

Anyway I think you did a really nice and clean job in the end, congrats!

wait for a replacement case to arrive because I ruined the case in step 1 : surprisingly enough this wasn't necessary

Too bad, I was expecting this could be an option to have some spare parts available! I indeed still have to hide the mess I did with the USB holes on the side panel...


Good luck with the screen/hinges, that's not an easy one!

 

[infoberg]

Part 5: A light at the end of the tunnel...​

... for the first modification, the internal PSU, that is. Some progress being made, soon it will all come together.

The last cable to be prepared was the PDCB input, which was straightforward. Hooking it all up outside the case and doing a first test to see if the PSU/PDCB combo works as designed. I decided to leave the inlet and switch inside the case - as removing the switch from the flimsy case seems impossible once it is in place.

Switch and inlet do their job, same goes for the PSU and PDCB, as the PSU tester shows stable voltages. Heat was of course no issue, as there wasn't any relevant current being drawn.

Test fitting the components inside the case also was successful, all parts play together nicely. LEDs were glued to the front panel and tested again, but at the moment I am a little bit stuck trying to get the new power button to operate correctly. I am still having issues finding the right way to add some padding, so the button will indeed press the tiny switch reliably.

Looks great! IMO you did an excellent job on those cutouts. The spray paint probably helped with that impression, but that's also part of the job Looking forward to seeing this move forward, that monitor mounting is going to be interesting!

Thank you for your comments, and you're absolutely right about the spray paint. Here is a little behind-the-scenes view of the raw cutouts, now you know why I tried to hide it from the public...

That's fun to see someone else going through the same challenges!
Except...
Ouch! You've learnt your lesson the hard way!
I think I did all the C6 cutout with a round file removing metal around the stock hole of the case, much safer, I should have told you...
Anyway I think you did a really nice and clean job in the end, congrats!

Thank you! I had indeed been wondering how you managed to cut the holes for the inlet so perfectly

Too bad, I was expecting this could be an option to have some spare parts available! I indeed still have to hide the mess I did with the USB holes on the side panel...

Good luck with the screen/hinges, that's not an easy one!

I am actually thinking about ordering another case, they are rather cheap now on AliExpress. I found a seller where the case costs less than 40 Euros, including VAT (which AliExpress collects for shipments to Austria since July 1st) and shipping.

As for the screen, I am not happy with those 2 hinges I was considering. Using them won't allow to prop up the screen at an angle, there would need to be another mechanism for that. And I think it's an absolute must to use the screen at an angle. While the horizontal viewing angle is great, the vertical is not. There's only a few degrees of tolerance, before the image looks all washed out. So more thinking and googling (or rather binging) needed in the days ahead on how to mount the screen and enabling it to be lifted up and fixed at any arbitrary angle.

Next steps:​

1. case modification for inlet and switch: done
2. wait for a replacement case to arrive because I ruined the case in step 1 : surprisingly enough this wasn't necessary
3. preparing all the necessary cables in the correct length (and waiting for connectors to arrive): done
4. testing PSU and PDCB with the PSU tester: done
5. if all goes well, hooking it all up: PSU, LEDs, PDCB, power switch and modified USB3 ports: done, except for power switch and USB3 ports, which will just be glued to the case
6. testing the computer for performance, thermals and airflow
7. lots of thinking, googling binging and learning to tackle the monitor case lid...: halfway done, I hope

 

[Valantar]

Wouldn't the best way to get a hinge that can hold the display up be to use a laptop hinge? A 180° hinge could be mounted to the inside of one side panel with a slot cut in the top/corner for it to emerge through, and would allow you to tilt the hinge to useful angles from 0-90° (assuming you have the display facing outwards when folded down) while making for a relatively flush mount. This is probably too beefy, but something along these lines could work, no?

 

[infoberg]

Wouldn't the best way to get a hinge that can hold the display up be to use a laptop hinge? A 180° hinge could be mounted to the inside of one side panel with a slot cut in the top/corner for it to emerge through, and would allow you to tilt the hinge to useful angles from 0-90° (assuming you have the display facing outwards when folded down) while making for a relatively flush mount. This is probably too beefy, but something along these lines could work, no?

Hmmm, true I haven't looked into using laptop hinges yet. I will check out if there are any viable solutions. Until now I had tried to find hinges that are as flat as possible or mounted to the side panels, but maybe there are other ways, just would need to hide the hinge with a front bezel...

 

[BaK]

but at the moment I am a little bit stuck trying to get the new power button to operate correctly. I am still having issues finding the right way to add some padding, so the button will indeed press the tiny switch reliably.

Quick comment, I don't remember if I have published this in my thread:

The added foam around the switch prevent the button to wobble and thus help having a centered pressure on the switch.
Hope that helps!

 

[infoberg]

Quick comment, I don't remember if I have published this in my thread:

The added foam around the switch prevent the button to wobble and thus help having a centered pressure on the switch.
Hope that helps!

Thank you for the hint, I think I saw it in your thread. I guess I need to make a trip to the cellar to look for some leftover materials then lol

 

[BaK]

Nice update!
Aside from...

Here is a little behind-the-scenes view of the raw cutouts, now you know why I tried to hide it from the public...

Let's say I didn't see anything!

I am actually thinking about ordering another case, they are rather cheap now on AliExpress.

This reminds me I got two cases in the first place, for the color mix, and only one case had a spring in between the stock switch and the power button to repel the later.
Do yours have one? If so you can maybe shorten it to make it compatible with the new setup, but the wires of the switch may interfere...

You are so close to testing the PSU on load, keep it up!

 

[richi-123]

good idea with the button, something interesting comes from here

 

[infoberg]

This reminds me I got two cases in the first place, for the color mix, and only one case had a spring in between the stock switch and the power button to repel the later.
Do yours have one? If so you can maybe shorten it to make it compatible with the new setup, but the wires of the switch may interfere...

Hmmm, mine didn't have a spring behind the button. Some sellers on AliExpress are advertising the case as "new version", maybe you got one of those. If I succeed with the build, I will order one or two more cases, I like the two-color version you created! If you're still in need of spare parts then, maybe we can arrange something...

 

[infoberg]

Part 6: It ain't over 'til the fat lady sings​

Getting the switch to work was quite easy, in the end it worked without any padding at all. Assembling and hooking up all the LED and switch cables, shielding the cables or shrink tubing them left me with this:

All fine and dandy - until I tried switching it on. Nothing really happened when pushing the switch, no booting up, no fans spinning or LEDs coming on, no output from the system speaker (which I always add until a system works). Only the EZ Debug LEDs (which are not documented in the MSI B450I manual, but rather only here) flashed very shortly, together with an undocumented LED on the PDCB. According to the MSI documentation the debug LEDs indicate an error when one of them stays lit, no word about the flashing. Still waiting to hear from MSI support about it...

I then tested the build using a standard ATX PSU and the PC booted correctly. I proceeded to check the input side of the PSU and the output side with the multimeter. As soon as the new power switch in the back of the case is switched on, the PSU provides quite stable 12.36V on the DC output side. I figure that this should be fine for the PDCB. Also the PSU tester indicated that all the voltages delivered by the PDCB are fine, with a power good time of 270ms.

At the moment I am suspecting that the setup and/or rise timings of the PSU (500ms and 50 ms respectively) are not good enough.

The setup time (or power good time) as measured by the PSU tester was 270ms, which is well within the required ATX specification of 500ms and far from the recommended value of less than 200ms.

But the rise time is definitely out of bounds at 50ms (per spec), I guess that's the reason the PDCB is not starting to draw power from the PSU.

Or maybe there's some other possible reason I am not aware of? Any ideas?

Next steps:​

1. back to the drawing board

 

[richi-123]

Doesn't the DC source always have to be on for pico-psu to work? maybe not 5vsb

 

[Valantar]

I was about to say, isn't the PSU on as long as power is connected, with the PDCB handling on/off duties etc? If so, rise time and setup time shouldn't matter. My guess - and this is purely a guess, mind you - is that the inrush current when turning on the system is high enough to entirely deplete the bulk capacitor in the PSU, with the PSU not reacting quickly enough to the changed load to avoid a severe voltage drop. If this is the case, it could theoretically be solved by adding a large capacitor anywhere between the PSU output and the PDCB input. You would have to find room for it, of course.

 

[BaK]

So close but not there yet, very frustrating!

the PSU tester indicated that all the voltages delivered by the PDCB are fine

The CPU 4 pin as well?

 

[infoberg]

Doesn't the DC source always have to be on for pico-psu to work? maybe not 5vsb

I was about to say, isn't the PSU on as long as power is connected, with the PDCB handling on/off duties etc? If so, rise time and setup time shouldn't matter. My guess - and this is purely a guess, mind you - is that the inrush current when turning on the system is high enough to entirely deplete the bulk capacitor in the PSU, with the PSU not reacting quickly enough to the changed load to avoid a severe voltage drop. If this is the case, it could theoretically be solved by adding a large capacitor anywhere between the PSU output and the PDCB input. You would have to find room for it, of course.

Yes, that was my understanding too. You connect the PSU to mains and it will always deliver 12V to the PDCB which handles "all the rest" i.e. the signaling, SBY power etc. That is if the additional switch I added between the C6 inlet and the PSU is on.

I did some more testing today, hooking up an external power brick to the PDCB, all fine. So the PDCB is definitely ok too. Tests with a horrible Sharkoon PSU to power the PC also showed no problems with the rest of the build. Using the PDCB and an external power brick showed nearly the same voltages as using the LED PSU, even with slower PG GOOD times than using the LED PSU. One thing I noticed and which seems to indicate that Valantar is right with his theory: when I connect the external power brick to mains, it immediately draws 20W to 50W, seemingly only to charge the capacitors, as there was no load connected yet.
Update: I noticed the same behaviour with my LED PSU, it also draws 10W - 30W for a few seconds.

But on the other hand, when the PC is turned on, it takes about 2-3 seconds for the load from the wall to rise from 0W to about 60W-70W. The multimeter seems to be polling too slowly, so it measures about 20W then 45W and then 65W-70W. Suppose this happens in 2 seconds, that's a rise from 0% to say 60% in what's well within the rise time specification of my LED PSU, if that spec is correct.

It's also not a problem with the new power button I made, using a screwdriver to short the relevant pins on the motherboard connector instead of using a power button didn't change the behaviour.

So close but not there yet, very frustrating!
The CPU 4 pin as well?

It is indeed frustrating, because all further testing did nothing to help.

The PSU tester requires connecting the 4- or 8-pin CPU cable and also shows the output, everything fine there. As mentioned above, the values were better than a "normal" ATX PSU or with the PDCB and an external power brick.

So at the moment I am really at my wit's end...

 

[BaK]

So at the moment I am really at my wit's end...

No clue either...

To sum up:
- There is no problem on the 'computer' hardware parts as you are able to start the computer with an ATX PSU
- You are able to start the computer when the PDCB is connected to an external brick, so the PDCB is also fine
-> then the only remaining part that can be faulty is the PSU

But
- The PDCB output values when attached to the PSU give good results on the tester
-> so the PSU seems to be working as expected, at least when there is no power draw or very few

Don't you have another setup that requires less power to test your PSU with? Like a motherboard with an embedded CPU drawing 20W from the wall or something.
Or having the PSU to take care of the motherboard only, and the ATX PSU for the CPU?
If that's not risky for your hardware that is, just thinking aloud here...

Well maybe it's time to summon our 12V chief magician @Thehack !

 

[infoberg]

No clue either...

To sum up:
- There is no problem on the 'computer' hardware parts as you are able to start the computer with an ATX PSU
- You are able to start the computer when the PDCB is connected to an external brick, so the PDCB is also fine
-> then the only remaining part that can be faulty is the PSU

But
- The PDCB output values when attached to the PSU give good results on the tester
-> so the PSU seems to be working as expected, at least when there is no power draw or very few

Don't you have another setup that requires less power to test your PSU with? Like a motherboard with an embedded CPU drawing 20W from the wall or something.
Or having the PSU to take care of the motherboard only, and the ATX PSU for the CPU?
If that's not risky for your hardware that is, just thinking aloud here...

Well maybe it's time to summon our 12V chief magician @Thehack !

Thanks for the perfect summary!

To be honest, I have been thinking about testing the setup with my previous build in the same case. It also has a PDCB (by Streacom) but an external power brick and the power draw is quite similar. After all this testing, I don't think my faulty PSU/PDCB setup is damaging the components in any way, so that's what I will try next. But I think I know what the outcome will be

Meanwhile the support of MSI proves not helpful at all, they don't say what the flashing of the LEDs could mean. they just told me that their board needs an ATX 2.31 compliant PSU and 132W is probably not enough - which is definitely wrong. The other build is perfectly fine with the Streacom 150 which delivers 150W in total, distributed over 12V, 5V and 3,3V - compared to the 200W PDCB I am using in this build which delivers 120W on the 12V rail. Only difference between the builds: the PSU...
So I guess the cuplrit is already clear, I am just trying to understand why it isn't working, so I can avoid making the same mistake again or try to find a solution for it.

Any additonal input/help from TheHack would of course be most welcome