ono3 - Pelican 1430 to S4 Portable VR Workstation

jØrd

S̳C̳S̳I̳ ̳f̳o̳r̳ ̳l̳i̳f̳e̳
sudocide.dev
SFFn Staff
Gold Supporter
LOSIAS
Jul 19, 2015
818
1,359
Army surplus stores occasionally stock external frame backpacks. Was a flood of them here going cheap just recently. Every hipster in town had one, they were just binning the external frame
 

Stevo_

Master of Cramming
Jul 2, 2015
449
304
I'd go with an internal frame pack, much more comfortable and keep the weight closer to your body where you want it. You can barely even find external frame packs for backpacking any more except for low end stuff maybe.
 

Soul_Est

SFF Guru
SFFn Staff
Feb 12, 2016
1,536
1,928
Just read it and it is a wonderful read. Thoroughly informative along with links to sources!
 
Last edited:

Rene Canlas

Efficiency Noob
Apr 11, 2016
5
0
Life's gotten in the way, but I've finally put up Pt 1 of a proper writeup, which covers various options for building the most portable VR rig possible: https://randomfoo.net/2016/03/28/the-most-portable-vr-workstation

After that will talk about DC and backpacking.

Gotta say, as a fellow S4 owner, you've certainly pushed the boundaries of what can be done with the S4!
Looking forward to updates on how your battery testing goes. I'm raring to try it myself for the Rift and Vive!
 

8BIT

Minimal Tinkerer
Apr 16, 2016
4
0
I posted some of my power/heat measurements in the main S4 thread, but I'll repost, along w/ more numbers w/ my move from the Zalman/Scythe to the Noctua cooler here.

Wattage is measured w/ a regular Kill-a-watt, so good mainly for ballparks. CPU monitoring/stress testing is w/ OCCT. GPU/CPU+GPU is with Furmark.

With the Zalman/Scythe (Scythe blowing air in, so the guard side towards the Zalman):

Using HDPLEX for power, GTX 970 unplugged
  • Idle (800MHz) - 40-45W, 42C
  • w/ Turbo (4.4GHz) - 173W, hits 100C (thermal throttle) in <25s
    • throttles to 4-4.2GHz - 150-160W, 100C
  • w/ Turbo disabled (4GHz) - 126W, 88-92C
Using SX600-G for power w/ GTX 970:
  • Idle (800MHz) - 95W, 42C
  • w/ Turbo disabled (4GHz) - 180W, 88-92C
  • Furmark (GPU 1088Mhz, 80-85C, 105% TDP, 85% fan, power limited) - 255W (CPU is about 65C, runs at 4GHz about 30% utilization)
  • Furmark + CPU Burner maxes out between 275-290W, both CPU and GPU around 80C.
Today I got a 6+6pin PCIe power cable for the HDPLEX. Here's what it looks like w/ the HDPLEX+Firebird PS and the Noctua cooler:
  • Idle (800MHz) - 100W, 42C
  • w/ Turbo disabled (4GHz) - 179W, 75C
  • Furmark (GPU 1088Mhz, 79C, 104% TDP, 77% fan, power limited) - 269W (CPU 60C, about 30% utilization)
  • Furmark + CPU Burner - 290-310W, both CPU and GPU around 77C.
Honestly I probably could have left the Zalman in - it's quite impressive considering the entire cooler is 85g and would handle a 35W/65W CPU fine, I imagine. The Noctua is much heavier (411g), although removing the Scythe and fan mount saves 100g, so it's not the worst tradeoff for a 15C cooler CPU under load.

Wait so the noctua cooler was 15C cooler than the zalman with the scythe fan?
 

lhl

SFF Lingo Aficionado
Original poster
Nov 16, 2015
121
143
Wait so the noctua cooler was 15C cooler than the zalman with the scythe fan?

Yep, this is on a 95W TDP 4790K mind you, and only when maxing out the CPU by itself. As you can see, when running Furmark+CPU Burner at full tilt, the difference is much lower.
 

CC Ricers

Shrink Ray Wielder
Bronze Supporter
Nov 1, 2015
2,234
2,557
Anyone able to try the Noctua cooler with the slim Scythe fan in place of the Noctua fan? I'm considering this if it means I can get similar temps with less noise.
 

Phuncz

Lord of the Boards
SFFn Staff
May 9, 2015
5,938
4,951
Awesome of you to report back in, I enjoyed reading it ! Do you have more updates or revisions planned for the build ?
 

lhl

SFF Lingo Aficionado
Original poster
Nov 16, 2015
121
143
I'll be traveling again soon so I won't be carrying around the backpack rig (among other things, I'm using a ~266 Wh LiPo battery, which is allowed on planes) so I probably won't be doing much more work on this, especially if the HP Omen X is any good. It's seriously compact:



Since I already travel w/ a portable screen and input, something like this would replace my S4 Mini and probably my laptop as well (the VR desktops have *almost* enough resolution to be usable w/o a screen).
 

Hahutzy

Airflow Optimizer
Sep 9, 2015
252
187
I just discovered this thread and it's awesome to see someone actualize the SFF backpack VR idea!

This was an idea that I had and wanted to take Hutzy XS to (which is why I added HDPLEX support), but I put it on hold because my case puts the GPU and CPU intake on opposite sides, and never had the time to really think about how to solve it.

It's interesting that you mentioned that you wanted to figure out how to make the setup fully wireless even between the HMD and the system (I only thought as far as making it wireless to unchain the human body from an outlet).

I do wonder how much latency a wireless solution would add to the USB 3.0 signal though. The latency might put it into the "technically possible but actually unusable" category... For now...
 

EdZ

Virtual Realist
May 11, 2015
1,578
2,107
Ideally, a later version of the Constellation cameras would do the basic thresholding and blob centroid finding on-board (a relatively simple operation) so all they need to pass back to the cost PC is a strong of coordinates, rather than full frames. This is how ethernet-connected mocap tracking cameras operate. It's possible that the tracking PC and rendering PC could be separated such that one very basic PC is connected to the cameras and handles the image processing, and hands off a string of coords over WiFi (or bluetooth, or dedicated RF) to the mobile host PC to perform sensor fusion and all other operations, but this is overcomplicated for a consumer system.

Wireless video to the HMD is a far more elegant solution in terms of carried weight and ease of setup, pending someone developing a wireless protocol that can handle ultra-high bandwidths with minimal LoS issues and minimal latency. That might take long enough for there to be a niche for media-converter-based extensions, swapping the existing 4-5m copper cables for much longer, lighter and more flexible fibre-optic cables.
 

lhl

SFF Lingo Aficionado
Original poster
Nov 16, 2015
121
143
Couple notes: added an addendum to my original writeup for a full parts-list if I were building a backpack system from scratch. Total cost is about $1300 and is available now (well, if you use a GTX 970 vs an RX 480). That may make the DIY route more relevant as it seems that MSI is targeting Holidays 2016 w/ a price of $2-5K, and Zotac and HP don't have any release dates/pricing announced yet.

I also stumbled upon the Mini-Box Y-PWR the other day, a cheap $20 controller board that should let you easily hot-swap batteries and included that on the parts list.

@Hahutzy - for a pure backpack system, the easiest thing to do would be to use something like the H5 SF cooler, which I also put in the final parts-list. It's probably not ideal to be sloshing around w/ liquid cooling on your back, but you get a lot of flexibility w/ that radiator and a PCIe ribbon cable. Conceivably, you can have all the PC components in a compact hydration pack form factor and have batteries (maybe a 5400 or 8000 mAh pack on each strap) on the front to help balance the weight (10lbs on your back isn't terrible in any case, but you could get really sleek w/ something like that With something like that Colorful motherboard, you could probably end up w/ something not too far from a Subpac form factor).

@EdZ I agree that offloading the Constellation code to an ASIC/microcontroller and making the tracking cameras wireless seems like a no-brainer, but Oculus seems to be having issues just putting one step in front of the other so I'm not going to get my hopes up too much for them getting it figured out before CV2.

As far as wireless protocols go, WiGig has everything covered w/ WDE/WSE already (bandwidth, latency, bidirectional interaction), but what's missing are good implementations. The challenges are difficult, to say the least (beam-forming for 60GHz, handling drop-outs and attenuation). Nitero have made some bold claims and promise to be demoing at E3, so we'll see, but if not, others aren't too far behind (Samsung, Intel, Qualcomm and Dell all have stuff, not to mention more specialized companies like Amimon, VidOvation, etc). Everything Nitero is doing already exists, but no one's brought it all together into a single solution yet.
 
  • Like
Reactions: Phuncz

EdZ

Virtual Realist
May 11, 2015
1,578
2,107
I agree that offloading the Constellation code to an ASIC/microcontroller and making the tracking cameras wireless seems like a no-brainer, but Oculus seems to be having issues just putting one step in front of the other so I'm not going to get my hopes up too much for them getting it figured out before CV2.
I think it;s less a problem of technical competence (something Ocyulus have in droves), but more that it's an added expense that isn't perceived as necessary yet. Going by the latest SteamVR hardware survey (which will be biased towards Vive users), about 70% of people do not have a space larger than 3mx2m (I can;t even clear a space for the minimum 2.5mx2m due to tiny London house sizes). For comparison, that's about the yellow rectangle. That is easily within reach of a USB3 cable even without a powered extender.
Wireless cameras offer many advantages in setup and configuration, so I suspect they will be moved to once Oculus can start shipping in megaunit quantities and amortise the ASIC costs of the chip required to do so. I'd be extremely surprised if the Rift, Vive and PSVR managed to ship over a million units combined in 2016, so the cost impact of custom components at such small manufacturing runs is very high per-unit, as evidenced in current pricing.