So here are a few items I need feedback on as a result of my test-fitting gear into the latest prototype:
1. GPUs/PCIe devices and the flex slot: Each flex adapter will support two PCIe devices (either 2 single-slot or 1 single-slot/1 dual-slot configuration). The problem is the supported device width. I wanted to use the dual, watercooled radeon frontier edition cards in my own build but their width with the EK 2x1 adapter is something like 140mm which causes interference with the PSU or the other flex slot. Basically, the largest PCIe width the S700 will support is just around 120mm - so blower-style coolers only if air cooled (and NO waterblocks as they would be too wide).
I think I need to stretch the S700 to support at least one flex slot with a width up to 140mm so here are my two questions:
a. Are we OK with increasing the height of the S700 by nearly an inch (and thus the volume by another liter or more) to support wider PCIe devices?
b. Should I increase the height even more (40mm or just under 2 inches) to support wider PCIe devices in both the flex slots? This would potentially increase the support from 120mm to 140mm radiators as well.
2. PSUs: I'm not liking my use of space near the bottom of the mainboard. Initially, I tried to support another DDC-style pump on this side but it has proven to be too narrow and would thus cause interference between the reservoir and the mainboard. The pump by itself could be installed, but the reservoir would have to be installed elsewhere. Based on the feedback from #1, I'm guessing the height will be stretched larger by at least 1", maybe 2. If that happens, I should have enough room to add mounts for 2x3.5" drives behind where the FlexATX PSU sits. If the height increases by 2", I'd easily be able to add a DDC combo mount that would clear the mainboard (maaaaaybe even if I only increase the height by 1" and shift the mainboard closer to the top of the case). I'm also going to add a giant passthruough hole at the bottom to help route the PCIe extenders without having to go behind the mainboard. Of course, if the top flex slot is used for PCIe devices, those cables will have to route behind the mainboard (and probably require around 500mm lengths). So my question is: What else can you think of to mount below the mainboard? (Assume the Flex ATX PSU might not be mounted)
3. The watercooling hardware I am using are primarily 90 degree elbows. I have both rotating and fixed hardware. The clearance is tight for the CPU elbows and if users want to use 16mm tubing, they will have to use the non-rotating elbows to clear the side panels. If 14mm is used, the rotating elbows *should* fit without issues. Most AIOs will fit assuming they have articulating hoses coming from the CPU block (I have a thermaltake unit I need to test fit to see if it clears but Enermax will not because the inlet/outlet are fixed and require like 70mm vertical clearance from the CPU IHS. The entire width of each chamber is only like 67mm so after the mainboard, socket, CPU, and standoffs are taken into account, there's only like 50mm of clearance from the IHS to the side panel. I could make each chamber wider to accommodate more types of hardware but I'd have to do so on both sides, thus increasing the volume. If I also increase the height due to #1, the volume of the S700 becomes at a minimum something like 24L which I think puts it in a gray area of inefficient use of space. Pros for increasing the width: 1. More hardware support for custom loops/AIOs. 2. Front panel could now support 3x140mm radiator. Cons for increasing the width: 1. Larger volume. 2. More prototype iterations (more $$$$$).
4. Anyone want a handle? It would probably cost me $25-30/each if I design it similar to the S401 handle.
I think I have abandoned the idea of a capacitive switch. I just don't have the time to develop a solution and make the custom harness for each chassis that would sell. I need something inexpensive and easy this go around so I think I'm going to stick with the centered 16mm pushbutton just like on the S401. I already have sufficient quantities in stock, their small, and they work. With the height increasing, I would have enough space I believe to mount a top-centered button on the front panel. Also, the front can be mounted upside down so users could choose to orient this button at top or bottom.
Let me know what you think. I've been in a design rut for the 3rd prototype. I'm putting extra pressure on myself to get it right because I can only afford to make 1 more. If it requires more major changes, I'm going to have to abort the S700. If the 3rd comes back and there are only minor changes required (like hole sizes) things are good. Thus, for the issues above I've started the 3rd design several times only to throw everything away because I'm not satisfied making huge changes to the design at this point. Given the number of S401 units I'm selling (low), I would expect to only sell about 10 of the S700. That already won't justify the $4k I've spent just getting to the second revision. By the 3rd prototype, I'll be in around $6k with no hope of making any of that money back. Further, I expect the cost to be at least $250 + shipping to purchase. I don't intend on getting the R&D money back (at 50 units it would take more than $100 profit to do so). Also, other cases (e.g. Cerberus) can provide similar specs, have a little less volume, and support more standard equipment (CPU air cooling, etc.). All these things combined, I don't expect a lot of movement on the S700. Maybe I get surprised, but I will not stretch my neck out like I did on the S401 with an initial order of 130. These will be manufactured as orders come in most likely (also raising the average selling price). Just a dose of reality and a bit of tempered expectations for myself and others as the design phase comes to a close.