Project Lowbucks I : Update - 2G Eclipse Sound System

 After a few year of use, the original subwoofer enclosure designed for Project Eclipse began to show signs of wear and age. The original grey carpet covering shows areas of discoloration. The trim panels are also beginning to come off. 

Version 1 shown below is based on a 0.7 cubic inch sealed enclosure made from 3/4 inch marine grade birch. This is a very strong and very heavy wood to use and because of the relatively small size, did not require additional bracing.  The set features a rockford fosgate single voice coil P110S4 Subwoofer and a Pioneer GM3300T bridgeable power amplifier. This setup has provided more than  adequate bass without being a window breaker. On a scale of 1-10, I gave this a 6 for performance / output and 3 for aesthetics.

The subwoofer enclosure is big enough to produce great sound, yet small enough so as not to take over the limited trunk space of the eclipse.

November 2011 - Wanting a little bit more bass output from the single subwoofer and at the same time increase available trunk space, Version 2 of the enclosure was constructed. This is a total redesign and the old birch box was discarded in favor of 5/8" MDF. Again, since the box is relatively small, flex is minimized even on the longest dimensions of the box and did not need a lot in internal bracing. The front baffle board was made with 2 layers of the MDF glued and nailed together.

Even though the main box area looks smaller than the birch box, the decrease in material thickness and slight adjustments in angles and measurements resulted in a box with an internal volume of a little over 0.9 cubic inch.  Still originally planned as a sealed box, the possibility of doing a ported box came in. After a few more tweaks, a final box size of box 1.1 cubic inch was produced. A 3 inch pvc port was inserted on one side but provisions to plug this port were also made should the final sound produced was un-tunable. Technically, the P110S4 calls for a ported enclosure size of about 1.4 cubic inch - so 1.1 seemed a little bit too small. Fiberfill was added in the hopes of ever slightly increasing the perceived box size.

The completed box protruded less into the trunk area. A simple flat baffle board and black carpet material provided a more streamlined and minimalistic look.

As far as performance was concerned, the box was left ported since it seemed to provide a few more dB's over the sealed enclosure. It's louder inside the cabin but the sound itself felt like it was "less controlled". Speaker cone movement looked excessive but only when playing extensive bass notes at higher volumes.  Sealing the box did not seem to affect the quality as much but overall output noticeably decreased. 5 for performance and 5 for aesthetics.

Box version 2 never really saw road use. Since the eclipse was basically parked for the winter, testing and playback were only done on the driveway. Fast forward a few more months to the end of Spring 2012 and the beginning of Summer 2012 and a new build was underway.

Version 3 was now in the works. The tired Rockford Fosgate P110S4 is now being retired in favor of two JL Audio 10WX-4 subwoofers. The doubling of cone area should provide more bass without increasing the box volume extensively.

The original Pioneer amplifier, despite still providing sufficient power, will not be capable of running a two ohm load. So for this duty, a corresponding JL Audio JX250/1 was chosen.

[Both subwoofers and amplifier were purchased new from a Canadian authorized JL Audio dealer. Boxing day special had the subs at $38 each and the amp at $88 - at this price, one won't even be able to purchase a crappy 6 inch tube-type subwoofer]

MDF was again chosen. The original 1.1 cubic inch box was modified to about 1.3 cubic inches. A few angled cuts, some tweaks to the shape and some more fiberfill rounded up the installation. Each speaker was given its own screw-type terminal cup. Internal speaker wiring used was 14 gauge and 8 gauge was used for power wiring. The new box was covered in a more automotive grade black carpet which should prove more resistant to everyday wear.

Version 3 resulted in a slightly bigger box (but still does not take over the small trunk area). Output was noticeably greater than the single sub and quality was better than a single sub in the ported enclosure. The new carpet looks sleeker and more professional. 7 for performance and 6 for aesthetics. 

PC Builds

The average desktop PC has a functional lifespan of roughly two to five years. The length of the lifespan greatly depends upon the type of system purchased, advances in hardware components and changes in the software that we run. Over time, users will tend to notice that their systems just aren't as fast as they used to be, they don't have enough space to store their files or don't meet the requirements for the latest software. When this happens, users have the option of either upgrading or replacing their PCs.

To determine which path might be better for your computer system, it is best to look at a cost comparison of what you will get out of each of the two options. My rule of thumb is that upgrades should typically be done if the costs of the upgrades will be roughly half of the cost of getting a new system. This is just a guideline based on most upgrades giving you a functional lifespan of roughly half of what a complete replacement will get you.

The advantage that desktop PCs have is a greater amount of upgrades that can be made to them compared to a laptop computer. The problem is that with so many components that can be upgraded, the costs of upgrades can quickly outpace the cost of replacement.

My last major desktop was built around March 2003. Computer components were very expensive relative to the amount of performance you get. Back then, typical displays were CRT or glass tube monitors. My 19 inch Samsung Synchmaster 955DF cost almost $300. Fast forward a couple of years and TFT LCD monitors were beginning to gain ground. The price was still comparable to a CRT so a 17" LCD would still cost about $300 for the entry-level models. Today, you can purchase a sub-20" LCD for less than $100.

The same holds for other computer components. Hard drives grew from a few megabytes to a few gigabytes and now in terabytes. Back in the 90's I was one of the first to purchase a one gigabyte drive. Now, one Terabyte drives are fairly common with 3 and 4TB models available in single drive format on the mass market. There has also been a noticeable shift away from mechanical hard drives to solid state drives or SSD. While hard drives are the storage kings, the SSD's are definitely the performance leaders. Even the fastest single mechanical drive is easily overtaken by the slowest SSD in overall performance. 

The greatest developments and improvements are both in the CPU's (processors) and GPU's (graphics cards). Computers in the 90's had processors running under 200MHz. Now, CPU's are at least 3GHz and run multiple cores. Modern day CPU's can even be overclocked for daily use beyond 4.5GHz. Most users will not need to upgrade the desktop video card unless they are looking for extra performance or functionality with 3D applications such as gaming. PC games advances at a very rapid pace such that new graphics boards are released about every six months.

Today, the turnover rate for computer components are very fast. What is essentially the best, new shiny component today will be outdated in as little as a few months. There is really no way to be totally up to date with components. A video card purchased for more than $500 today will be half that price in a few months. 

PCs can be designed around a specialized task. A gaming PC would generally need an updated top of the line video card. That same high end video card will be useless for a graphics or CAD workstation. Multimedia PC used for video and sound editing need larger storage capabilities. These computers would still do well with a low to mid-range video card.

I tend to build and assemble computers based on specific tasks - a gaming computer, a multimedia server, a general workstation and a sort of jack-of-all-trades PC. Of course each of these modern rigs would still be capable of doing common everyday tasks so their functions will generally overlap with each other. The trick here is to spread out the builds between several months so you don't necessarily end up with multiple computers with outdated components.

After a long hiatus from custom-built desktop systems, I finally gave in and built a rig last September 2010. My intention here was to have a fairly capable gaming system that would be able to run the latest games at that time. I had to work around a budget of about $800 yet at the same time get the newest and most capable components at the time. What can I say? I was both cheap and greedy :-)

Five months after, I was thinking of either improving the hard drive on the gaming rig by adding either a Western Digital Black hard drive (or anything equivalent) to improve storage performance. The i5 was no slouch in terms of audio and video editing and encoding - but whenever the PC was assigned these tasks, I would naturally be unable to play any games. I then decided to build a dedicated multimedia PC. According to some reviews, longer threads and extended instructions work better on the new AMD Phenom II line of processors.  Sub-$200 6-core processors - seemed too good to be true. Intel's 6-core offerings start at $999 for bulk pricing.

Fortunately, AMD did not disappoint with the X6-1055T.

Three months later and after experiencing the capabilities of AMD's Six-core processors, I was ready to invest and create a desktop system that would be able to duplicate and surpass the capabilities of the budget gamer and multicore midget pc. This new system would be based on the tried and tested AMD Phenom II family but with an unlocked core so that additional tweaking and overclocking would be possible. The new target pc will be a mid to high level gaming computer with the capabilitiy to do high definition video and multimedia editing, extensive multitasking while operating continously 24/7. Why AMD and not intel? I do not have a real reason. Perhaps it was just the price point and the bragging rights of running a hexa-core system. (Intel 6-core systems range around $600-1000 at the time of writing). The AMD flagship 6-core, phenom X6-1100T was just a little over $250. Sure, it cannot trade punches blow-by-blow with the Intel processor but it's capabilities are just fine for what I need to accomplish. Besides I was thinking of getting a dedicated GPU based on ATI which is now owned by AMD, so I felt like "keeping it all in the family" as far as the major components are concerned.

After considering all my options and the functionalities I wanted to have, I decided to go a step down from the flagship processor. The AMD X6-1100T is clocked at 3.3GHz while the AMD X6-1090T is clocked at 3.2GHz. The 1090T was about $65 cheaper and just 100MHz slower - something that could easily be attained by very mild overclocking. All other specs are basically equal.

The other crucial component would be the video card or GPU. Again, I went the AMD way and obtained a factory overclocked HD6850. At the time of build, this card represented the best money value for performance. Sure, there are faster, bigger and more expensive cards, but this is intended to be the "sweet spot pc" - so no excessively expensive components ! The focus is on quality and reliability.

AMD also has a technology called "eyefinity" which allows up to 6 displays or monitors to run off a single video card. An eyefinity setup can actually span and spread the video image to all monitors connected - more like a smaller scale video wall. This is a unique feature of AMD graphics products that cannot be found on any other consumer graphics solution at the time of the build. This capability allowed me to use 2x24 inch Samsung LCD screens and a 37 inch Toshiba LCD panel. This setup is excellent for multimedia content but not ideal for gaming applications. However, it does serve the purpose I need and generally turn off eyefinity when gaming.

Fast forward another 6 months and I was again looking at another build. This time I had two goals: Build a reliable budget box, that would still be able to do HD video playback and at the same time use AMD's new FX Bulldozer processor line.

There were great expectations with the new processor among the AMD crowd. Unfortunately, it has turned out to be more of a disappointment. Performance was less in comparison to the Phenom II line and in real-world applications, the FX quad cores were more on-par with the Intel dual core i3 lineup.  Still, four cores is four cores.... or is it? As it turned out, the new micro-architecture is a totally new way of looking at multi-core computing. AMD has re-introduced the "Clustered Integer Core" micro-architecture, an architecture developed by DEC in 1996 with the RISC microprocessor Alpha 21264. This technology is called clustered integer core technology and referred to as "modules" by AMD. In terms of hardware complexity and functionality, this "module" is midway between a true dual-core processor and his integer's power (each thread have a fully independent core) and a single core processor that has the SMT ability, which can create a dual threads processor but with the power of one (each thread shares the resources of the core with the other thread). Basically, a dual-thread Bulldozer processor has one single core (or module), a four-thread processor has two cores and the eight-thread processor has four cores.

In theory it was good. In efficiency, in proved to be excellent. In stability and overclockability, the entry level, 95W TDP fully unlocked FX-4100 runnig at 3.6GHz reached up to 4.6GHz  on regular air cooling without the need for an aftermarket cooler. The only "upgrade" was to use another OEM cooler from a 125W TDP processor instead of the one that came in the box. Not to mention the fact that the processor was actually undervolted during the overclock. 

However, despite the higher clock rate, (compared to the phenom II), it still underperformed. Another setback was that Windows 7 does not know "what to do" with this new processor microarchitecture. As a result, Microsoft released several patches to optimize Windows 7 systems. The result was about a 2% increase in performance.

(Windows 8 systems are said to make better use of the processor and is optimized for this new architecture). Performance gains? - we will find out when the official product in no longer RTM and available off the shelf.

As of July 2012, Budget multitasker has been running 24/7 without any issues. It has proven to be a very capable desktop system and for it's sub-$500 price, still offers very good value for the money.  Quality components with the possibility for future upgrades. Unlike pre-made branded systems, this one uses regular off the shelf parts as opposed to proprietary components that will make changes or repair possible and less expensive.

A new build is being planned this time probably based on AMD's other line of processors with a built-in graphics processor unit on a single chip. This build will be a media center pc that will only be used for HD playback, internet content streaming and file (movie and music) server functions. Instead of going with a quad-core setup, this rig will probably use either a dual core or triple core chip, have a built-in 6xxx series video with the capability to do AMD crossfire with a standalone additional video card. Crossfire technology basically combines two compatible video processors to almost double the performance. At the time of writing, these 4 processors are being considered:

A4-3300 dual core @ 2.5GHz with onboard HD6410

A4-3400 dual core @ 2.7GHz with onboard HD6410

A6-3500 triple core @ 2.1GHz with onboard HD6530

A6-3650 quad core @ 2.6GHz with onboard HD6530

When combined with an inexpensive HD6450 (approximately $20 after rebates), the A4's will have the GPU performance of what is equivalent to an HD6510 while the A6's will be similar to an HD6550.

The only problem at this point is that these processors (subcode: Lynx-Llano)  are based on the FM1 motherboard socket. There is a new line of Fusion processors expected to come later this year or early next year (subcode: Virgo-Trinity-Piledriver) that are supposed to be slightly more efficient, and sporting 7xxx series onboard graphics. However, these processors are expected to utilize an FM2 motherboard socket which is incompatible with the previous generation.