The BitFenix Pandora is a sleek and sexy aluminum-clad microATX tower with a refreshing design. It has the ambition and look of a high performance case but its dimensions prevent it from being a true powerhouse.

April 19, 2015 by Lawrence Lee

Product	BitFenix Pandora MicroATX Tower Case
Manufacturer	BitFenix
Price	US$100

The BitFenix Pandora is a breath of fresh air compared to most tower cases. It’s somewhat on the petite side, being a microATX chassis with a narrow 16.0 cm wide body that barely accommodates a full-sized ATX power supply. Like more and more cases these days, the Pandora has no 5.25 inch bays, tossing them aside in favor of a gorgeous sculpted curved front with brush aluminum side panels draped around it. There’s more aluminum rounding out the back as well and not a single screw is visible anywhere on the exterior. The panels simply pull off in a similar fashion to the NCASE M1 and SilverStone Fortress FT03.

The box and case.

The case is a fingerprint magnet, though, and the grain on the panels to be rather coarse. After initial cursory examination, I noticed a strange fine powder accumulating on the surface. I was the source of this material, as the aluminum had scraped against my hands leaving behind particulate and a metallic odor on my skin. Inside, it supports up to 3 x 2.5 inch and 2 x 3.5 inch drives, but it lacks a dedicated drive cage; the drive placements are scattered in various locations.

There are two 120 mm fans included, one at the top and one at the front, with a second fan placement sitting unoccupied behind the front bezel. The Pandora can accommodate video cards up to 35 cm in length, and the presence of an extra expansion slot suggests that a dual video card configuration wouldn’t be out of place. However, as the case is narrow, CPU coolers are limited to 13.4 cm in height, though you can use liquid cooling with a 12/24 cm radiator mounted at the front. It has the look and ambition of a high-end performance case but its dimensions may be the limiting factor.

The BitFenix Pandora, silver windowed version.

The Pandora is available in black and silver, with or without window, and costs around US$100, while a more basic version, the Pandora Core, sells for US$90. The only difference is the Pandora sports a small 2.4 inch LCD at the front used to display the BitFenix logo. The panel is connected via USB and BitFenix provides software to display any 240 x 320 image of your choosing. If you dislike shiny/novel things, the cheaper Core model has an embossed logo instead.

Accessories.

The Pandora ships with a few accessories including a short assembly guide, a few zip-ties, and two thick velcro straps. Thrown in with the screws are some rubber grommets for drive installation, a thumbnut for tightening standoffs, and an Allen key as expansion cards are secured with hex thumbscrews.

Relevant Specifications: BitFenix Pandora (from the product web page)
Materials	Aluminum, Steel, Plastic
Colors (Int/Ext)	Black/Black, Black/Silver
Dimensions (WxHxD)	160 x 420 x 465
Motherboard Sizes	Micro ATX, Mini-ITX
ICON™ Display	2.4″ TFT, 240 x 320
3.5″ Drive Bays	x 2
2.5″ Drive Bays	x 3
Cooling Top	120mm x 1 (included)
Cooling Front	120mm x 2 (1 included)
PCI Slots	x 5
I/O	USB 3.0 x 2, HD Audio
Power Supply	PS2 ATX, up to 180mm in length
Extras	Brushed Aluminum Side Panels, BitFenix ICON Programmable Display, Cable Management

EXTERIOR

The BitFenix Pandora is constructed of aluminum panels, a plastic top and front bezel, and a steel frame. It measures 16.0 x 42.0 x 46.5 cm or 6.3 x 16.5 x 18.3 inches (W x H x D) for a total volume of 31.2 Liters.

The top of the case consists of a massive dust filter and a curved panel with audio and USB 3.0 ports. The power and reset buttons are illuminated with white LEDs and are overly large, rocking noticeably when depressed at the edges. On our sample, the power button sticks out rather than being flush with the rest of the panel. The filter can be detached by pressing on two points along the left side.

The side panels hug the chassis but their hold loosens toward the front bezel, leaving a significant gap to serve hidden intake vents. It’s a clever design that give the case a unique look while serving a practical purpose.

The filter has a thick plastic hexagonal skeleton, a layer of steel mesh, and a light sheet of foam inside. It’s rather pointless to filter the top of the case, and this restrictive design is made worse by the fact that the top is the main exhaust. It’s also difficult to clean as several metal tabs have to be angled just so in order to take it apart.

The aluminum curves partially around the rear of the case as well. As the case narrow, there are no fan placements at the back, just a fairly open rectangular grill. There is enough room for two 60 mm fans but no mounting holes are provided. A surprisingly loose-fitting plastic compartment that falls off at the faintest of touches hides the thumbscrews for the expansion slots.

Underneath the chassis is a vent for the power supply fan at the rear is serviced by a magnetically attached filter. The front bezel at the other end has a large open intake point. The holes visible near the front set of feet are for mounting either a 2.5 or 3.5 inch drive on the case floor using rubber grommets and screws.

The side panels simply pull off, attached at four different points using standoffs that grip onto positions along the frame of the chassis. The fit doesn’t feel quite as secure as other cases I’ve encountered with this system. The panels are rather thick though, at a considerable 1.9 mm.

INTERIOR

As the top and front of the chassis are basically just rails for supporting fans and drives, so there isn’t much heft if you strip away all the aluminum. What little steel resides underneath is sufficiently sturdy with no obvious weak points or sharp edges.

Removing the side panels reveals the true nature of the front bezel. The curved portion of the panels hide huge rectangular intake vents which sandwich a glossy reflective center. The LCD is located about 1/4 of the way from the top but it’s difficult to pick up on camera when not turned on.

The bezel can be detached as well, revealing another dust filter held on with small magnets and two adjustable intake fan positions with one 120 mm fan pre-installed. The front display has a USB 2.0 header to power it and link up with the software for changing the logo image. There is no frame/guides to help hold the filter down, so it’s easy to accidentally knock out of place from the inside while messing with cabling.

The top filter covers just a single pre-populated 120 mm fan placement and a 3.5 inch drive position. The drive rests on four isolation pads, hooking onto two plastic tabs on the side while on the opposite side, two screws are used to secure it from underneath. This design unfortunately doesn’t allow for a second fan option.

The layout is similar to larger towers but there is no 5.25 inch drive support and only two spots for mounting 3.5 inch drives (at the very top and bottom of the case) are provided. The metal bracket at the bottom of the chassis resembles a drive cage but only allows a single 2.5 inch drive to be installed sideways. Its main function is to help direct/hide cables. The aluminum portion at the rear covers the expansion slot screws, explaining why they used hex thumbscrews and included a perpendicular Allen key.

The included fans produce a front to top airflow path that seems to focus on CPU rather than GPU cooling, which makes sense as the CPU heatsink height limit is just 134 mm. Given the size of the intake vents, a third fan seems like a good idea.

The power supply position is elevated/damped by four rubber pads (one fell off during unpacking). The metal bracket near the front is secured with two screws, one behind the motherboard tray and the other underneath the case.

The right side of the case has several points for tying down cables but there is very little space once the side panel is in place. There is a mere 5 mm available behind the motherboard tray. The depressed section toward the front of the case offers a third 2.5 inch drive placement, so it’s a bit roomier, with 14 mm of clearance provided.

ASSEMBLY

The assembly process is simple and straightforward and most of the issues are with regard to space and wiring. The chassis’ narrow body limits the choice of both CPU cooler and graphics card, and the restrictive space behind the motherboard tray you want to leave some of the thicker cables on the top side.

Our standard micro/full ATX test system fully assembled with some substitutions. The width of the case excludes our Asus GTX 980 Strix graphics card and Scythe Mugen Max CPU cooler. In their place is a narrower Zotac GTX 970 AMP! Extreme Core and the shorter Noctua NH-C12P. Also, the foam lining the top cover of the case has been discarded.

Our CPU heatsink stands 114 mm tall with about 20 mm to spare, confirming the specified height limit of 134 mm. Our 11.1 cm wide video card has about 2.3 cm of clearance above it. However, a GPU being installed has to clear the aluminum edge at the back of the case as well. If a card is overly wide near the bracket, it might not be possible to maneuver it in place, even if there’s enough physical space for it inside.

With so little room for cabling, only the smaller wires are tied up on this side. It should also be noted that the hole at the top/rear corner is too small to route either a 4-pin or 8-pin connector.

Most of the cables on this side are placed in the less restrictive front section.

The logo software is just a simple .EXE file that you can drag and drop 240×320 JPEG/PNG files onto. Head on, the display is bright and vibrant, but if you look closely, you can spot some dust under the panel.

While this is a nifty feature, the quality isn’t great due to poor vertical viewing angles. It only looks good within a 30° span near eye-level — beyond that, the image distorts noticeably. From below, the image fades, becoming washed out, while from above, it loses color.

The shape of the side panel window is odd. Instead of a simple rectangle, there is an extra bit to show off… the plain metal bracket?

TESTING

System Configuration:

• AMD A10-6800K APU – 4.1 GHz, 32nm, 100W, socket FM2
• Noctua NH-C12P CPU cooler
• Asus F2A85-M Pro
  motherboard – AMD A85X chipset, microATX
• Zotac GeForce GTX 970 AMP! Extreme Core
  graphics card – 145W
• Kingston HyperX LoVo memory, 2 x 4GB, DDR3-1600 in dual channel
• Seagate Desktop SSHD hybrid drive – 2TB, 7200 RPM, 8GB NAND Flash, SATA 6 Gbps
• Cooler Master
  Silent Pro M700 power supply – 700W, modular, ATX
• Microsoft
  Windows 7 operating system – Ultimate, 64-bit

Measurement and Analysis Tools

• Prime95
  processor stress software.
• FurMark
  stability test to stress the integrated GPU.
• Asus GPU Tweak to monitor GPU temperatures and adjust fan speeds.
• SpeedFan
  to monitor system temperatures and adjust system fan speeds.
• Extech 380803 AC power analyzer / data logger for measuring AC system
  power.
• PC-based spectrum analyzer:
  SpectraPlus with ACO Pacific mic and M-Audio digitalaudio interfaces.
• Anechoic chamber
  with ambient level of 11 dBA or lower

Testing Procedures

The system is placed in two states: idle, and load using Prime95 (large FFTs setting) and FurMark, an OpenGL
benchmarking and stability testing utility. This puts more demand on the CPU and GPU than any real life application. Throughout testing, system temperatures, noise levels, and power consumption are recorded. During the load test, the system and GPU fans speeds are adjusted to various levels in an attempt to find an optimal balance between cooling and noise while maintaining a GPU temperature of 80°C (assuming an ambient temperature of 22°C).

Baseline Noise

For our baseline noise tests, the system is left idle, the CPU fan is set to just 400 RPM, and the GPU fans are set to minimum speed (1060 RPM) by default. The system fans are connected to controllable fan headers and are set to a variety of speeds using SpeedFan. This gives us a good idea of what the stock fans sound like in actual use at different speeds with minimal interference from other sources.

Baseline Noise Level (Idle, CPU fan at 400 RPM, GPU fans at 1060 RPM)
Fan Speed Setting	Avg. Fan Speed	SPL @1m
0%	N/A	17~18 dBA
60%	640 RPM	18 dBA
80%	830 RPM	19 dBA
100%	990 RPM	22 dBA
Measuring mic positioned 1m at diagonal angle left/front of case. Ambient noise level: 10~11 dBA@1m.

The system fans have a nominal speed of just 1000 RPM, and as there are only two included, they produce very little noise compared to most tower cases. The noise they add to the system is not measurable below about 60% speed (~640 RPM) so there is no point running the fans any slower than this. At top speed, the system generates just 22 dBA@1m, so it’s likely that the internal components will drown out these fans in the load tests.

The quality of the noise emitted by the stock fans is pretty good. They produce an innocuous, soft, white-noise type sound with a distinct lack of tonality, even at full speed. The case also deals with hard drive noise fairly well as only faint vibrations can be felt on the side panels when the machine is powered up — the 3.5 inch drive mounting design does a better job of isolation than most cases. The bottom position is preferable to the top as it’s easier to hear the drive working through the top cover.

TEST RESULTS

 

System Measurements (80°C Target GPU Temp)
System State	Idle	Resident Evil 6 Demo (Peak)	Prime95 + FurMark
Avg. System Fan Speed	640 RPM (60%)	990 RPM (100%)	990 RPM (100%)
CPU Fan Speed	800 RPM	1000 RPM	1330 RPM (100%)
GPU Fan Speed*	1060 RPM	1370 RPM	1370 RPM
CPU Temp	30°C	66°C	69°C
MB Temp	36°C	52°C	59°C
SSHD Temp	35°C	39°C	39°C
GPU Temp	34°C	80°C	80°C
System Power (AC)	50W	301W	319W
SPL@1m	18~19 dBA	24~25 dBA	26~27 dBA
*set as low as possible to maintain target GPU temperature on load. Ambient temperature: 22°C.

Our usual hardware setup includes a larger CPU heatsink, the Scythe Mugen Max, with its fan spinning at a modest 800 RPM. So far this has been sufficient to adequately cool the processor, even on full synthetic load, in all of our recently reviewed ATX tower cases. The smaller top-down Noctua cooler used to accommodate the Pandora’s narrow physique isn’t quite up to the task when the system is fully taxed. Even with the system fans and CPU fan at full speed, the processor hits close to 70°C, at which point it begins to throttle. The system can handle the heat produced by the chip itself, but the additional thermal load from the GTX 970 is too much. However, keep in mind this is a far more demanding test than any combination of real applications you can throw at it.

The Resident Evil 6 Demo Benchmark is a more realistic gaming stress test, and at its peak load, the CPU fan can be slowed to 1000 RPM while maintaining a slightly lower and stable CPU temperature. It’s fairly balmy inside as the motherboard temperature sensor eclipses the 50°C mark and the SSHD, sitting in relative comfort on the case floor away from all the action, warms up to just under 40°C.

The noise level with these settings is 24~25 dBA@1m which is reasonably quiet for a gaming rig, and the acoustic profile is fairly smooth and benign.

AUDIO RECORDINGS

These recordings were made with a high resolution, lab quality, digital recording
system inside SPCR’s own 11 dBA ambient anechoic chamber, then converted to
LAME 128kbps encoded MP3s. We’ve listened long and hard to ensure there is no
audible degradation from the original WAV files to these MP3s. They represent
a quick snapshot of what we heard during the review.

Each recording starts with ambient noise, then 5~10 second segments of product
at various states. For the most realistic results,
set the volume so that the starting ambient level is just barely audible, then
don’t change the volume setting again while comparing all the sound files.

• SPCR ATX Test System in BitFenix Pandora – Baseline (idle, CPU at 400 RPM, GPU fans at 1060 RPM)
  — system fans off (17~18 dBA@1m)
  — system fans at 80% / ~830 RPM (19 dBA@1m)
  — system fans at 100% / ~990 RPM (22 dBA@1m)

• SPCR ATX Test System in BitFenix Pandora
  — idle, system fans at 60% / ~630 RPM, CPU fan at 800 RPM, GPU fans at 1060 RPM (18~19 dBA@1m)
  — Resident Evil 6 Demo Benchmark (peak), system fans at 100% / ~990 RPM, CPU fan at 1000 RPM, GPU fans at 1370 RPM (24~25 dBA@1m)
  — Prime95 + FurMark, system fans at 100% / ~990 RPM, CPU fan at 1330 RPM, GPU fans at 1370 RPM (26~27 dBA@1m)

FINAL THOUGHTS

The BitFenix Pandora is the best looking microATX tower to come through SPCR. Brush aluminum panels that pop off have been done before but the Pandora takes it one step further with its curved front edges. This gives it a snazzy profile and also serves to hide the substantially sized intake vents, helping minimize the noise directed toward the user from within. The black top filter and front bezel accent the silver nicely and the small display is a nifty addition that is also customizable. The latter is sure to draw the attention of enthusiasts, though the quality of the LCD is subpar due to inadequate viewing angles.

Inside, the 3.5 inch drive placements are most impressive. The Pandora can accommodate two such drives, one on the ceiling and one on the floor, and both locations are relatively effective at limiting drive vibration. The case fans have a nice sound, though they probably run too slow for most users. Unfortunately, cooling is not this case’s strong suit despite its fairly open airflow design. The width of the chassis restricts the CPU heatsink height to 134 mm and the single 120 mm exhaust fan has difficulty expelling hot air out of the case. This is perhaps one scenario where a AIO cooling unit may come in handy as it could direct heat generated by the CPU out through the front of the case and out of the heat path of the video card.

Cable management is probably the biggest annoyance. A few zip-ties and velcro straps don’t make up for the lack of space behind the motherboard tray. The foam lining the top filter is the most aggravating misstep; its presence goes against all common sense. Beyond that, all that’s left are some nitpicks like the weak feeling power/reset buttons, hex thumbscrews securing the expansion slots, and the coarse grain of the side panels.

The BitFenix Pandora is a beautiful small tower case with refreshing design. The smallish size hampers cooling but it’s so elegant that it may be justifiable for form to trump function. Users who use the most demanding components won’t like it, but it should satisfy the majority. I would like to see a wider, full-size ATX version of this case with a large rear exhaust fan and space for bigger components as there’s no reason to limit its inventive design to this particular form factor. For an aluminum clad chassis, its street price of US$100 seems reasonable but if you want to save $10, look at the Pandora Core, the same case minus the front display.

Our thanks to BitFenix
for the Pandora case sample.

The BitFenix Pandora is Recommended by SPCR

* * *

Articles of Related Interest
Be Quiet! Silent Base 800 ATX Tower
Cooler Master Silencio 652S Tower Case
SilverStone Kublai KL05 Budget ATX Case
NZXT H440 Mid Tower Case
NoFan CR-80EH & CS-60 Fanless Cooler & Case
Rosewill Legacy U3 Aluminum MicroATX Tower

* * *

Discuss
this article in the SPCR Forums.