Asrock B350 ITX/AC (but the B450 should be just around the corner), silverstone's SST-ST45SF V3.0 450W and silverstone tundra slim cooler.
-
Save 15% on ALL SFF Network merch, until Dec 31st! Use code SFF2024 at checkout. Click here!
Production SENTRY 2.0: Evolution of console-sized gaming PC case
- Thread starter ZombiPL
- Start date
After many requests we made additional pack of photos showing configuration with 3.5'' drives and some without internal components. Link below:
We've *just* got first prototype and while there are things that worked out right away, like for example kensington lock, and for the most part new riser mounting system (we just need to adjust this slightly), there are still few things that we need to fix to make sure they won't cause issues in the production. So this first prototype is definitely not going to be the last before we order a test batch of 25 or more units.
So it'll be a bit more time and probably one or two more single prototypes before we know that we're back on track with the plan.
To summarise prototype 1 in context of the planned changes listed in the opening post of this thread:
NOT DONE | NEEDS MORE WORK/TESTING | MOSTLY DONE
1. Bigger ventilation area - we know now that it can be made at our subcontractor and the case body is still rigid. We will have to test this for comparison to 1.0/1.1 at some point, but the priority now is solving those issues mentioned above.
2. Less screws to screw the top cover - it seems to be working ok with the prototype
3. Less types of screws - changes in the riser mounting system and dropping support for 2.5" mounts in GPU compartment along with quad drive brackets lets us significantly reduce the amount of screw types needed.
4. Torx screws instead of hex screws (if we will be able to find a supplier for black ones) - we have sourced sample black torx screws but with "security pin" and those require specific type of bit/screwdriver which is not ideal and we haven't found yet standard black M3 torx screws without the pin.
5. Improved Riser Clamping System (RCS) - new mounting method works but need some minor dimension adjustments.
6. Redesigned top hdd/ssd holding bracket - new mounting bracket works but will be adjusted to tweak support for 120 mm AIOs.
8. Kensington lock - works as intended.
9. Wifi antenna holes with improved shape for one-tool assembly and rotation lock - work as intended.
10. Vertical stand rubber feet attached without glue - work as intended, but we need to see how prolonged use will affect the rubber feet.
11. Two additional holes in vertical stand for symmetry of installation - works as intended.
12. Two 3.5” HDD mounts in the GPU compartment instead of 2.5” mounts - work as intended, although mounting the 3.5" drives in front of PSU require plugging the sata cables first, with flat/straight connectors recommended, and it gets really messy with a lot of cables in the same area. Some planning is required to execute this neatly.
13. 120mm AIO Water Cooling mount in the GPU compartment - Fitting AIO works, but we need to adjust mounting hole sizes for both radiator mounting (the threads in AIO radiators are not aligned with enough precision to go for M3.5 mounts as they are designed for M5 fan screw mounts and do not need such precision although they use M3.5 screws) and fan mounting screw pass through (the big holes in the center wall above the fan). We are also trying to figure out a way to support 52 mm package, because the go-to AIO we assumed would fit within 50 mm is actually 52 mm (Corsair H75 has 27 mm rad instead of 25 mm as the website specs). Finally we need to figure out a proper way of mounting the pump for optimal bending angle on the hoses.
14. SFX-L not supported anymore due to various issues and for the sake of 120mm AIO - RIP SFX-L support
15. Chassis mass reduction where possible - a lot of metal was stripped from multiple internal pieces - final weighting will be made after we finalise the design.
16. Transport package redesign for reduced shipping cost and packaging time - new packaging is already designed, but we need to design new printout and there will be some adjustments depending on the way we want to package accessories.
17. Zip tie mount points for cable management - we have added multiple holes for zip-ties in the center wall (as visible in the photos) however with AIO or 3.5" hdd and short GPU those are not convenient to use and do not add much functionality. We will have to check if any of those are at least helpful to arranging the cables with full length GPU. We plan on adding one or two zip-tie mounts in the 2.5" drive bracket to handle AIO hose routing.
18. Motherboard IO backplate mount update - works as intended now with properly aligned screw holes. Minor tweak in the alignment for welding is required for the motherboard mount plate, but it the other end of the part near the PSU mount.
19. PSU mount clearance improvements for IEC C14 connector - new simplified PSU mount works as intended. There are no problems with corsair PSU as well (there were issues with those in 1.1)
20. Online-only manual - new manual will be made after the design is finalised.
21. No white colour version until our subcontractor is ready - still only black colour planned.
22. No quad 2.5” drive brackets - RIP quad 2.5" support
23. PCI slot brackets not included anymore - case can be assembled now for shipping without those thanks to changes in GPU bracket.
24. Economic postal package not available anymore, only DHL courier shipping - DHL shipping tested already
25. Relocating internal middle structural element to support slightly bigger ITX GPUs (Gigabyte/MSI) - Gigabyte 1080 Mini fits along 120 mm radiator. Cables on PEG connector are really squeezed because of the card oversize so low profile PEG connector is recommended, but not necessarily required.
26. Optimised air pocket above the GPU - new air pocket design will have to be extensively tested after we close design changes coming from this prototype.
1. Bigger ventilation area - we know now that it can be made at our subcontractor and the case body is still rigid. We will have to test this for comparison to 1.0/1.1 at some point, but the priority now is solving those issues mentioned above.
2. Less screws to screw the top cover - it seems to be working ok with the prototype
3. Less types of screws - changes in the riser mounting system and dropping support for 2.5" mounts in GPU compartment along with quad drive brackets lets us significantly reduce the amount of screw types needed.
4. Torx screws instead of hex screws (if we will be able to find a supplier for black ones) - we have sourced sample black torx screws but with "security pin" and those require specific type of bit/screwdriver which is not ideal and we haven't found yet standard black M3 torx screws without the pin.
5. Improved Riser Clamping System (RCS) - new mounting method works but need some minor dimension adjustments.
6. Redesigned top hdd/ssd holding bracket - new mounting bracket works but will be adjusted to tweak support for 120 mm AIOs.
8. Kensington lock - works as intended.
9. Wifi antenna holes with improved shape for one-tool assembly and rotation lock - work as intended.
10. Vertical stand rubber feet attached without glue - work as intended, but we need to see how prolonged use will affect the rubber feet.
11. Two additional holes in vertical stand for symmetry of installation - works as intended.
12. Two 3.5” HDD mounts in the GPU compartment instead of 2.5” mounts - work as intended, although mounting the 3.5" drives in front of PSU require plugging the sata cables first, with flat/straight connectors recommended, and it gets really messy with a lot of cables in the same area. Some planning is required to execute this neatly.
13. 120mm AIO Water Cooling mount in the GPU compartment - Fitting AIO works, but we need to adjust mounting hole sizes for both radiator mounting (the threads in AIO radiators are not aligned with enough precision to go for M3.5 mounts as they are designed for M5 fan screw mounts and do not need such precision although they use M3.5 screws) and fan mounting screw pass through (the big holes in the center wall above the fan). We are also trying to figure out a way to support 52 mm package, because the go-to AIO we assumed would fit within 50 mm is actually 52 mm (Corsair H75 has 27 mm rad instead of 25 mm as the website specs). Finally we need to figure out a proper way of mounting the pump for optimal bending angle on the hoses.
14. SFX-L not supported anymore due to various issues and for the sake of 120mm AIO - RIP SFX-L support
15. Chassis mass reduction where possible - a lot of metal was stripped from multiple internal pieces - final weighting will be made after we finalise the design.
16. Transport package redesign for reduced shipping cost and packaging time - new packaging is already designed, but we need to design new printout and there will be some adjustments depending on the way we want to package accessories.
17. Zip tie mount points for cable management - we have added multiple holes for zip-ties in the center wall (as visible in the photos) however with AIO or 3.5" hdd and short GPU those are not convenient to use and do not add much functionality. We will have to check if any of those are at least helpful to arranging the cables with full length GPU. We plan on adding one or two zip-tie mounts in the 2.5" drive bracket to handle AIO hose routing.
18. Motherboard IO backplate mount update - works as intended now with properly aligned screw holes. Minor tweak in the alignment for welding is required for the motherboard mount plate, but it the other end of the part near the PSU mount.
19. PSU mount clearance improvements for IEC C14 connector - new simplified PSU mount works as intended. There are no problems with corsair PSU as well (there were issues with those in 1.1)
20. Online-only manual - new manual will be made after the design is finalised.
21. No white colour version until our subcontractor is ready - still only black colour planned.
22. No quad 2.5” drive brackets - RIP quad 2.5" support
23. PCI slot brackets not included anymore - case can be assembled now for shipping without those thanks to changes in GPU bracket.
24. Economic postal package not available anymore, only DHL courier shipping - DHL shipping tested already
25. Relocating internal middle structural element to support slightly bigger ITX GPUs (Gigabyte/MSI) - Gigabyte 1080 Mini fits along 120 mm radiator. Cables on PEG connector are really squeezed because of the card oversize so low profile PEG connector is recommended, but not necessarily required.
26. Optimised air pocket above the GPU - new air pocket design will have to be extensively tested after we close design changes coming from this prototype.
We have been stuck on trying to figure out whether we should try to support 52 mm AIO package (27 mm radiator with 25 mm fan) and how to do it for a bit.
The problem with that is, we would have to move the center rail up above the radiator and fan while making sure the screws holding the riser and 2.5" hdd bracket will not go up as well, because that would limit 2.5" compatibility significantly.
We have spent significant amount of time this year simplifying the center rail to make it easier to manufacture and lighter (as visible in the photo galleries) and modifying it to support 52 mm AIO package would make it over complicated once again.
We also do not want one configuration that is not necessarily going to be a mainstream one to affect whole construction in such significant way. Most likely mainstream configurations will have a full length GPU instead of AIO + itx-sized one.
Because of that, we have decided to figure out first if that 50 mm AIO limit is actually a problem and if using slim fan with 27 mm radiator (which is most common among 120 mm AIOs) will be terrible or not.
We have done series of tests this week with Corsair Hydro H75 paired up with 20 mm thick DEEPCOOL GS120 SLIM adding up to a 47 mm package. We have tested three AM4 CPUs with different TDPs: R7 1700/R5 2600X/R7 2700X. Tests were done during two cloudy/overcast days with 25~27°C ambient temperature.
We have decided to compile captured footage with temps and clocks tracking instead of compiling data into charts because we feel like it will give you a better understanding how components behave while playing games rather than when looking at raw data.
Here's the video:
General outcome of those tests is that making support for 52 mm is NOT necessary.
Additional notes on such configuration with 120 mm AIO liquid cooling:
- Using standard profile memory sticks is STRONGLY RECOMMENDED. It is really tight with two RipJaws sticks, USB 3.0 cables, 24-pin connector and the tubes at the same spot.
- It gets really messy with cables between the PSU, radiator and power switch. It is not a configuration recommended to everyone.
- Modular PSUs have stiff ribbon cables out of the box and those make it really hard to handle around the power switch.
- 2700X with its 105W TDP works okay in games but when fully loaded with video or 3d rendering it gets hot really quick. I think it is not something that should be cooled off with a single slim fan.
The problem with that is, we would have to move the center rail up above the radiator and fan while making sure the screws holding the riser and 2.5" hdd bracket will not go up as well, because that would limit 2.5" compatibility significantly.
We have spent significant amount of time this year simplifying the center rail to make it easier to manufacture and lighter (as visible in the photo galleries) and modifying it to support 52 mm AIO package would make it over complicated once again.
We also do not want one configuration that is not necessarily going to be a mainstream one to affect whole construction in such significant way. Most likely mainstream configurations will have a full length GPU instead of AIO + itx-sized one.
Because of that, we have decided to figure out first if that 50 mm AIO limit is actually a problem and if using slim fan with 27 mm radiator (which is most common among 120 mm AIOs) will be terrible or not.
We have done series of tests this week with Corsair Hydro H75 paired up with 20 mm thick DEEPCOOL GS120 SLIM adding up to a 47 mm package. We have tested three AM4 CPUs with different TDPs: R7 1700/R5 2600X/R7 2700X. Tests were done during two cloudy/overcast days with 25~27°C ambient temperature.
We have decided to compile captured footage with temps and clocks tracking instead of compiling data into charts because we feel like it will give you a better understanding how components behave while playing games rather than when looking at raw data.
Here's the video:
General outcome of those tests is that making support for 52 mm is NOT necessary.
Additional notes on such configuration with 120 mm AIO liquid cooling:
- Using standard profile memory sticks is STRONGLY RECOMMENDED. It is really tight with two RipJaws sticks, USB 3.0 cables, 24-pin connector and the tubes at the same spot.
- It gets really messy with cables between the PSU, radiator and power switch. It is not a configuration recommended to everyone.
- Modular PSUs have stiff ribbon cables out of the box and those make it really hard to handle around the power switch.
- 2700X with its 105W TDP works okay in games but when fully loaded with video or 3d rendering it gets hot really quick. I think it is not something that should be cooled off with a single slim fan.
I brought this up somewhere else, but who even makes standard profile memory anymore? Everything above DDR4-2400 has huge 'heat sinks' on them. In the DDR3 days it seemed much easier to get standard height memory (and hey remember low-profile memory?)
Corsair Vengeance all the way up to 4600 MHz is standard height.
Corsair Vengeance LPX
Kingston HyperX FURY
Those are just slightly taller (1~3 mm) than standard profile. The issue with my RipJaws is they are really tall in the middle.
at todays ram prices?
no thank you, buddy of mine bricked an 8GB stick of 3000mhz DDR4 trying....
Not really, I'm not going to jump on the RTX bandwagon until I get to see how it performs in the games supporting it and that will happen in October afaik.
Unless of course I get a chance to test that by borrowing it from someone who jumped on the pre orders, but I don't think any of my friends or colleagues did that.
I'd say that RTX may start making sense in Q1 2019.
Dr. Zaber and CD Projekt Red...
Is it a Polish thing to make us wait for your wonderful products? =P
Sorry for the joke x), is there any news?
Is it a Polish thing to make us wait for your wonderful products? =P
Sorry for the joke x), is there any news?
I'm not sure. I'd say it's something like do it properly or don't do it at all kind of attitude, but I'm not sure if it's specific to our country.
There are still few details we need to figure out before we can update manufacturing drawings and order second prototype.
Jokes aside i'm actually very glad, i rather wait and have a great product than having it rushed and half baked.
Thanks for the info, mate!
hey saperpl, I'm looking at making some modifications to my sentry gen 1 and I was hoping you could provide some insight,
can I drill holes in the back of the GPU chamber? I want air to pass straight through my rad if possible,
for painting, will I need to sand it down first or is the powder coat already a solid base?
can I drill holes in the back of the GPU chamber? I want air to pass straight through my rad if possible,
for painting, will I need to sand it down first or is the powder coat already a solid base?
If you mean perforating the cover above the radiator then it's not technically a problem, there's no fluid anything like that inside, but it may get loud if you do that and also you may end up bleeding hot air inside (depending on the flow direction) making the whole air pocket useless.
Powder coat is a solid base... for another layer of powder coating, but if you do that you might have hard time fitting PSU and assembling the case because it will get thicker everywhere.
Other type of paint may not stick to it properly. You should use aircraft paint remover to strip the powder coating off the case. Note that it may be toxic so take safety precautions. You may also have your case stripped from paint in a shop that is stripping paint from car wheel rims.
Similar threads
