RonCraighead wrote:
My system can be had on the used corporate market for, believe it or not, $150... It's a 4 year old dell. I added a new Nvidia card to it, and it runs great. If I can build four or five of them for $250 each in hardware, all the better. I don't expect amazing graphics... Or even necessarily super accurate flight models.
+1 to above. Rather than using an old business PC, consider using a NEWer Haswell based PC. But here's the catch. Don't go with a high end CPU. Instead use a budget Haswell Mother Board that can be had for less then $70 and combine with a dual core Pentium CPU that can be had for less than $70 as well. This will give you a modern PC on a budget with performance that will exceed what a 4 year old Dell can provide.
It would also be smart to understand which OS the older Dell came with. If Windows 7 then some of the residual value of the PC will be in the OS. If Vista or older then this doesn't matter if the plan is to run Linux.
Another thing to consider is the very few purchased business PC's have large enough power supplies to run high end graphics cards. And many business PC's are difficult to upgrade. Some Dells have non-standard power supplies that make them almost impossible to upgrade. This is more likely a problem with the older Dells or the small form factor units.
On the subject of Video cards, my advice is to consider the video card carefully. Consider that a Urber high end video card needs a high end power supply. If building on a budget, making the wrong choice will have a cascade effect. For example the GTX750 series of Nvidia graphics card. This is a less than $120 card that in my system gives 45-70 FPS. It has very modest power requirements.
On the other hand, too little graphics horse power will ham string the system giving terrible frame rate. IMO once the rate drops below 20 FPS, it will go down hill quickly.
The key here is to build a matched system that stays within your budget. Avoid spending too much on any one component. Especially if not re-usable. That means no SSD. Don't over buy memory, cases, or power supplies in an attempt to make the system "future proof".
However in my opinion you will be much better off REPEATING the setup with newer PC components. Could you also explain why the goal is to have multiple sim setups. It seems to me that a single good set up would be better than multiple lesser quality setups.
So that you can compare my system specs:
Mother Board: ASUS Z87 Haswell
CPU: Haswell i5
8Gigs of RAM (again the cheapest available)
NO OVER Clocking. So no need for an upgraded CPU cooler.
450 Watt power supply (cheap!)
Graphis: ASUS GTX750 (I paid less than $150 more than a year ago. They are even cheaper now)
Saitek rudder pedals, Saitek yoke, dual Saitek throttle quad, Saitek radio statck.
Analog head phones with analog mic.
Over all X-Plane performance 45-70 FPS.
I have been able to get all of the gear to work together under an Unbuntu Linux 64 bit on PilotEdge. I have logged hours of stable flight.
The point about X-Plane's poor use of multiple monitors is a valid point. That is why if I were building this system I would get the largest and best monitor I could afford.
Now onto a completely different direction. If I were the owner/operator of a real flight school with real airplanes, I would at least look at the available touch screen systems on the market. Image the benefit of a touch screen system that almost exactly matched the planes the students were flying. The time to consider this is before too much money was spent to make changing directions overly difficult. IE, BEFORE it's time to create multiple systems (the multiplier effect!). Should touch be a possibility Linux is out.
http://www.flythissim.com/touchtrainer.php
http://www.flythissim.com/newsite/products/sim-avio/ (Simavo for more of a DIY'er solution).
To recap. Building a budget PC based on newer'ish hardware/CPU using budget Mother Board with a minimum investment in MB, CPU, PS, memory, video card, case would give you a good test bed to begin testing with. But expectation management is vital. Don't expect to look out the window and see the amazing graphics some people show case on Youtube videos. To me, if the cockpit dash is readable and the airport signage is readable and reasonable then much beyond that is just gravy.
The only change I might suggest is to attempt to use the older Dell only if you currently
OWN this device. Here's were Linux will give you a HUGE advantage. Linux is not married to the PC hardware like Windows is. There's a high likely hood that you could take the Linux OS HD out from the Dell and transplant into a newer PC and Linux will boot with zero issues. However I do recommend configuring both PC's with SATA AHCI enabled.
http://en.wikipedia.org/wiki/Advanced_H ... _Interface
To add to the comments about the Saitek Yoke and rudder pedals. I have zero experience in a real plane so I'm not the best to respond. HOWEVER, this is were the MISSION of the sim needs to be clearly defined. If the goal is to familiarize the student with instrument flying and not primary flight training, then any short comings in the primary flight controls is not as important.
IF the goal is also PRIMARY flight training, then the accuracy of the controls takes on far more importance. From the opening post, it seems to me that primary flight training isn't the goal for these systems.
