Production Lazer3D LZ7 - Quiet Gaming Cube PC Case

[K888D]

I have put together a survey for gathering the communities preferences on the design of the case, it would be highly appreciated if you could fill one in and let me know your thoughts:

http://******/forms/11Z6RdbyyaYZq5Xw2

In a week or 2 I will summarise the data on this thread for all to see. The data will be used to improve the design.

Thank you in advance!

 

[MarcParis]

just answered it..

Finally my only concern will led color in front panel. As I'm going to black/red theme, I guess it will be suitable to have a front RED led.

Solution on NFC S4 mini to have only Bulgin power button is also a very good solution. USB 3 cable is so cumbersome to rout inside a case, I see no issue to get rid of it on my LZ case.

 

[K888D]

Thank you for the feedback so far, keep it coming!

I think the LED in the IO PCB is a white light with only a tint of blue, the camera photo makes it look more blue than it is.

But, if you choose red corner pieces then the power button will be red, so in theory the light that shines through the button material will look red, I would need to check this out though.

Solution on NFC S4 mini to have only Bulgin power button is also a very good solution. USB 3 cable is so cumbersome to rout inside a case, I see no issue to get rid of it on my LZ case.

What is a bulgin switch? Is this the same as a vandall switch? I did plan to make an alternative IO panel available with a 16mm diameter hole for vandall switches that the customer can provide themselves.

This will be a very simple thing to do, so yes it's certainly possible. It should reduce the price by around £10 (if you supply your own 16mm vandals switch). And as you say, there would be less cable clutter to route.

 

[MarcParis]

Yes they are the same round metallic button, with ring led..
I must admit that I love their esthetics..

Related to your front panel that will match corner color, it's a really nice feature indeed!

 

[K888D]

Thank you for all the feedback so far everyone, I really appreciate it, there are some interesting results so far.

Keep it coming, the survey is HERE.

I will post a summary of the results at some point within the next week.

 

[iFreilicht]

I can't answer that poll, you're missing orange in the colour options

 

[K888D]

I can't answer that poll, you're missing orange in the colour options

I've updated the survey to include some Orange options

But.... Orange Acrylic is only available in 3mm for the prototypes and its also a slightly transparent, in fact most of the coloured panels apart from Black and White are slightly transparent. 5mm Orange should be available at the production stage.

For a prototype Orange could be used for all the 3mm sections of the case (left/right sides and rear panels), orange sides would go nicely with a White front, maybe? Here are some photos that might help describe what I am saying better:

Fluorescent Orange:

Dark Blue:

Dark Blue Tint:

Dark Grey Tint:

Light Grey:

 

[GuilleAcoustic]

Lovely !

 

[iFreilicht]

orange sides would go nicely with a White front, maybe? Here are some photos that might help describe what I am saying better:

Yeah I think that might work. That would also mean the HDD mount is orange, right? I still have a very hard time deciding, but thanks for adding it back in!

EDIT: I think what could look really nice is the same colour scheme your prototype has with the orange LEDs of the Phoenix and a cutout logo in the front with an orange HDD mount behind that. I'll just mark orange everywhere I could see it working, just so it get's a voice

 

[K888D]

That would also mean the HDD mount is orange, right?

Yes that's right, the HDD bracket is 3mm also.

 

[MarcParis]

Nice color panel.

 

[K888D]

This evening I am doing some temperature testing on the Samsung PM951 256Gb M.2 PCIe NVMe SSD (that's a mouthful!).

I was quite surprised at the temperatures these drives run at, they need good cooling to keep them running at full speed.

I am also shocked at how much they reduce in performance once they do throttle, I've discovered that the PM951's throttle temperature is 70C. At 69C the drive reads at 1,576MB/s and at 70C it reads at 315MB/s. So basically you have to keep your drive under 69C to maintain full speed.

I'm carrying out a variety of tests which include no system fan, 50% and 100% system fan speeds with and without the underside vents covered. Firstly I am testing the system in an idle state and stressing the SSD with CrystalDisk. If I get time I will also repeat the testing whilst the system is in a completely loaded state.

 

[GuilleAcoustic]

I am also shocked at how much they reduce in performance once they do throttle, I've discovered that the PM951's throttle temperature is 70C. At 69C the drive reads at 1,576MB/s and at 70C it reads at 315MB/s. So basically you have to keep your drive under 69C to maintain full speed.

That's why I plan to go with a PCIe NVMe board with a proper heatsink. This is the benefits of mATX over mITX.

 

[K888D]

That's why I plan to go with a PCIe NVMe board with a proper heatsink. This is the benefits of mATX over mITX.

Bare in mind that those temperatures have only been reached so far with the system fan turned off, with proper case cooling underside drives should be ok, although I haven't undertaken any testing with the whole system under load yet, will report back.

 

[MarcParis]

Thx for feedback.
Ssd throttling is real, HOWEVER, as these drives are really fast, it's happens rarely except in benchmark.. Don't bother too much on this point

 

[EdZ]

Maybe get a nice square of thin copper sheet (possibly even an adhesive-backed one) the size of an ITX motherboard, punch holes to match the standoff locations (so it sits down over them touching the case wall, not touching the motherboard backside) and use a thermal pad to couple the drive to the copper sheet to act as a heatspreader. The thermal mass should be sufficient to deal with heat spikes during normal use, and with a power consumption under 5W it should be enough to prevent thermal throttling under sustained loads by coupling to the case wall to remove heat more effectively (as well as being a larger surface area for any under-motherboard airflow).

 

[BirdofPrey]

I'd just as soon use those small heatsinks they make for cooing GPU VRMs and RAM with some aftermarket coolers than don't interface with them.

 

[Phuncz]

Mind you that fully loading these PCIe M.2 drives for even a minute is very hard to do outside of benchmarking. So benchmarking them for extended periods of time isn't indicative for real-world scenarios, just like red-lining a car engine in neutral is not indicative of real-world performance.

 

[iFreilicht]

I'd just as soon use those small heatsinks they make for cooing GPU VRMs and RAM with some aftermarket coolers than don't interface with them.

Those are quite tall, though. Depending on the standoffs in this case, there's only 6-10mm of space between the bottom of the board and the case, and the SSD itself is about 4-5mm off the mainboard. If the case had a metal bottom one could use thermal pads to just use the case as a heatsink, but EdZ' idea might work, too.

 

[K888D]

Those are quite tall, though. Depending on the standoffs in this case, there's only 6-10mm of space between the bottom of the board and the case, and the SSD itself is about 4-5mm off the mainboard. If the case had a metal bottom one could use thermal pads to just use the case as a heatsink, but EdZ' idea might work, too.

Thats right, there is only about 1mm - 2mm of space between the bottom of the case and the M.2 SSD, the case is also plastic so it can't act as a heat sink.

Maybe get a nice square of thin copper sheet (possibly even an adhesive-backed one) the size of an ITX motherboard, punch holes to match the standoff locations (so it sits down over them touching the case wall, not touching the motherboard backside) and use a thermal pad to couple the drive to the copper sheet to act as a heatspreader.

Yes this would probably work very well, perhaps a thin 1mm sheet. But this isn't something I would do as part of this case project. In fact with the system fan the SSD performed very well, I will explain in the full results shortly.

Mind you that fully loading these PCIe M.2 drives for even a minute is very hard to do outside of benchmarking. So benchmarking them for extended periods of time isn't indicative for real-world scenarios, just like red-lining a car engine in neutral is not indicative of real-world performance.

Yes I found this as well, the CrystalDisk benchmark test lasts for about 5 minutes if you max out all the settings, I found that the sequential read causes the drive to heat up the quickest so I focused on this test. The other random read and write tests didn't cause anywhere near as much heat build up. The heat build up seems directly linked to how fast the data is being read, and as the random read tests don't achieve the same speed they don't generate as much heat.

You would be surprised how close to its throttle temperature the M.2 drive sits when idle without proper cooling, in some instances it takes less than a minute of benchmarking to hit its throttle temperature. Active system cooling makes all the difference. Results to follow.