HELP!....my emachine wouldnt turn on, so I thought I had crashed the harddrive, so I replaced it.....still didnt work, then I read about the notorious faulty power supply, so I replaced it.....still doesnt work.....when you push the on/off button, the green light goes on for just a second, the the light goes off....the fan runs...the computer does the hum noise (like any computer)...but nothing goes....nothing on the screen....????....HELP.....is there a quick fix to this??....all this happened just 1 week after the 1 year warranty expired...(go figure)...

JLamb, After reading your post, I could not help but respond. I used to work for a company that provided hardware technical support for these systems. It was my job to identify, research, and document hardware support issues for the technical support call center, the field technicians (ha!) dispatched for on site service (for those customers who had such warranties), and the technicians who worked in the repair depot. I have spent two years supporting these machines and investigating their problems, using and testing various models supplied to us by ECS in a lab we had set up for that purpose. I have also supported a variety of other systems, including home computers, servers, workstation, laptops, and thin clients from IBM, Dell, HP, Toshiba, Compaq, and others, so I have a reasonably well-balanced perspective when it comes to assessing reliability and value. Get rid of your e-Machine. I can tell you without the greatest certainty that ECS makes the worst motherboards on the face of the earth (even worse than those awful PC Chips and Tseng boards!) and that their systems are likely the worst of any produced by a major vendor. In terms of hardware quality, reliability, and expandability (I'm completely ignoring performance), these systems define the bottom of the barrel. For a start, the power supplies are very cheap and unreliable. They were a frequently replaced part at my former job. Good power supplies should have a circuit that checks to make sure the power is stable and at the correct voltages before sending the Power_Good signal that allows the system to boot (During those first few milliseconds when you turn the system on, the power can fluctuate wildly.) In two years, I did not see one ECS-built system that had a power supply that had this circuit. They all faked it by tying one of the output rails to the Power_Good signal and delaying it for a few milliseconds. This delay was sometimes not long enough, causing all sorts of problems, including corrupted CMOS memory. The power supplies were invariably way too small for the systems. They were usually barely able to power the system in its original factory configuration. When users added devices, there often wasn't enough power to ensure reliable operation. Some systems even shut down when users plugged a network cable into them! The tiny draw from the ethernet cable was enough bring the whole system down instantly! (This issue was prevalent among PowerSpec computers.) When they did work, the power supplies put out all sorts of hash, noise, and interference. On an oscilloscope, it was a mess. This went unnoticed by users, but a sure sign the power supplies were made about as cheaply as possible. The real indication that they were cheap was that they tended to fail... a lot. The motherboards, as I said were the worst on earth. Incredibly, the DIMM sockets on many of them were too deep! (Again, this was primarily a PowerSpec issue, specifically with the 3xxx models.) On certain models, it was almost impossible to install memory modules because of this. You could seat a DIMM as far into the misshapen socket as it would go, but it would never quite make good contact with the contacts at the bottom of the slot because it was too deep. DIMMs would lose connection with the socket contacts doing shipping and even when users moved systems from one room to the next. Upgrading memory was a Zen-like excercise; it was often necessary to remove and re-insert a DIMM nearly a dozen times to get it seated well enough to get past the POST. The act of inserting a second DIMM usually unseated the first, multiplying the problem. It was often possible (after much work) to get two DIMMs properly seated simultaneously, but I never heard any reports of anyone getting three to work in these systems. Motherboard failures were routine in these systems. In fact, I'm surprised your power supply went before your motherboard took out. More than half of all failures in these systems were motherboard failures. Some of these systems had hacked-up BIOSes. The idea was that if you removed access to features available through the motherboard's CMOS setup utility, users wouldn't be able to mess around and screw up their motherboard BIOS CMOS settings. Well, this *generated* tremendous call volumes. It was impossible to add some peripherals or change several settings that users really needed to change. In addition, many of the CMOS batteries provided with these systems were old stock, having been stored in a warehouse for god only knows how long. Many systems exhibited failing CMOS batteries failing within months. Replacing the batteries with service parts solved the problem, but only for a short time. The replacement batteries came from the same place as the originals, and were already near the end of their useful life. I have personally seen one system that had FOUR service calls to replace the battery! The exasperated customer eventually went to Wal-Mart, purchased a replacement battery, installed it himself, and solved the problem. There were also many electrical design issues with ECS motherboards. It's impossible to say how many customer problems were caused by this and not by a flaky power supply, cheap or improperly memory, or some other problem, but there were definite grounding issues. Currents would begin flowing on the chassis ground. These were low and didn't necessarily cause problems, but it became common practice to insert plastic washers and electrical tape in strategic places where the motherboards came into contact with screws and other metal objects. Many of these systems contained liberal amounts of HOT GLUE. That's right--hot glue. Most of them had gobs of hot glue around the floppy drive, hard drive, and CD-ROM drive cable connectors, and on just about every other connector in the system. I was told that this was to prevent people from disconnecting the internal cables. It looked to me like it was done to hold the system together! You'd be hard pressed to find a cheaper way to assemble a system. Most of the ECS-built systems that had slots for the CPU. However, they came with CPUs that plugged into a SOCKET. To connect the Socket 370 CPUs to the Slot 1 motherboards, they used a socket to slot adapter (frequently referred to as a "Slocket"). The CPU plugged into the adapter and the adapter plugged into the slot. While not the most desirable arrangement, this is not necessarily a bad thing. There are many good slockets available and many people use them without a problem. However, ECS used poorly made, cheap slockets that frequently came unseated. What's far worse--and as I see it, UNTHINKABLE, is that many of these systems had CPU slots with GOLD contacts. That's good. The slockets that plugged into them had TIN contacts! These systems would work for a bit, but as current passed through the junction of soft gold pressed against hard tin, a lovely chemical reaction would occur, building up an oxide layer between the connection, causing the connection to go "high resistance". This cause infrequent lockups and truly bizarre behaviour that was impossible to properly diagnose. The problems would become increasingly frequent and severe until the system was rendered useless. Simply removing and re-inserting the CPU solved the problem (temporarily), because of the self-cleaning nature of the contacts. As far as any real solution for customers with these systems, there was none. They needed a slocket or processor with gold contacts; all we could send were slockets with tin contacts. The memory installed in these systems was often marginal. No one noticed this as long as they ran DOS or Windows 95, 98, or Me. If you tried to run Windows NT, Windows 2000, OS/2 Warp, Solaris, Linux, or some other "real" 32-bit operating system, they would crash. In fact, these systems usually did not support Windows NT, 2000, or any operating system other than what was pre-installed. (This was less true of the e-Machines, but more true of the PowerSpecs.) Systems that were sold with Windows NT and Windows 2000 were actually specially built for NT and 2000 (Although they weren't advertised as such. People were given the impression that they were exactly the same as all the other models with the exception that they came pre-loaded with NT or 2000. The truth was that the "regular" models' hardware was too marginal for NT or 2000 to work!) I left just as the first Windows 2000 machines were being rolled out, so I'm sure this situation has changed. The hard drives in some models were known defective rejects from Western Digital. (I normally don't name manufacturers or vendors unless they've really done something truly reprehensible.) Years ago, Western Digital had a huge lot of hard drives with defective logic boards. Drives with this defect worked perfectly except for one small problem: When the read/write heads moved backward over the platters, it read and wrote garbage to and from the drive. (It worked perfectly as long as the heads were moving forward, however.) This problem was very serious because it eventually destroyed data on the hard disk, eating away at it until it screwed up the FAT or some other logical disk structure or corrupted the Windows registry, a dll, or some other key system file and rendered the system unbootable. Since the data corruption only happened when the head was moving backward, the problem usually went undetected for several weeks or even several months, depending on the amount of disk activity (Customers who frequently defragmented their drives, for example, tended to experience problems within a couple of weeks.). ALL of these systems eventually became unbootable. Every single last one of them--no exceptions. The phone lines at the call center and return lines at stores became clogged with people because of this issue. There was a 100% failure rate in the systems with these hard drives. Authorized representatives from vendors (Microcenter in particular) repeatedly lied about the problem, giving us several different but equally unlikely explanations, all of which were quickly disproved by testing in our lab. The problem was easily seen using diagnostic software to test the hard drive, performing a reverse seek test. Everything passed when writing and reading with the head moving forward; with the head moving backward, it couldn't read or write a single byte without garbling it. There were hundreds and hundreds (possibly a few thousand?) of systems with these hard drives. It would have been impossible for vendors and ECS not to have known what was going on. I find it impossible to believe that Western Digital didn't know these drives were defective. I was never able to find any evidence that any Western Digital drives packaged for retail had this problem. Presumably, Western Digital was careful enough not to put these drives on store shelves--they sold them to ECS instead. If this is the case, I'm sure ECS got a great deal on these drive. Unsurprisingly, replacing the OEM versions of these drives with the retail version solved the problem. (If you bought your system after 1999, it probably doesn't have one of these drives. They were all IDE Western Digital Caviar drives with a 3.1 or 3.2 capacity (I've forgotten the exact model.)) Even the floppy drives in these systems had problems! Failure rates for these systems were unbelievable. They were many, many times higher than those of IBM, Dell, HP, Compaq, Toshiba, or any other type of system we supported. (I'm not recommending these particular brands, these are just the systems I had available for reference.) Every day, there were angry, frustrated customers who'd had service call after service call, parts replaced again and again, and eventually had whole systems replaced two or three times calling, sometimes represented by their lawyers, threatening class-action lawsuits, civil suits, etc., etc. I'm amazed that ECS stayed in business. I have been working with home computer systems since 1982, I've done tech support for years, and I've seen a *LOT* of crappy systems but I have never seen anything from a major vendor as poorly designed, cheaply made, or as unreliable as the majority of systems made by ECS or that were built using ECS parts. The PowerSpec systems and e-Machine systems are similar and share a common parent. Many top-end PowerSpecs have the same parts as some e-Machines. The e-Machines *are* a step above the PowerSpec line, but it's not a very large step. When you consider how bad the PowerSpecs are, well... Some people have reported little or trouble with their PowerSpec, V2 Premier, or e-Machine. I'm amazed and am unable to account for this. It makes me wonder if they're even using their systems at all. I would not accept one of these systems if it were given to me for free if I couldn't resell it. I'd have to be PAID to take one! Honestly, the best recommendation I can make is to get rid of your e-Machine and get something that has nothing to do with Elitegroup Computer Systems! I could tell you stories about them you wouldn't believe... I'm sorry this post is so long, but instead of going on and on about how bad these systems are, I wanted to provide at least some real information about these systems, not just my own personal opinions (which are, of course, also included at no charge). Again, the e-Machines are probably the best systems available from ECS, but the company's products have a hideous track record for hardware failure and unreliability.

By the way, although I didn't mention the name of my former employer (I still have friends who work there), I'd just like to mention that it wasn't Stream International, Warrentec, or AON, companies who also, at one time or another, has had contracts to support these systems. I didn't for Microcenter/MEI (MicroElectronics Internation) either, although I close dealings with them on a near daily basis.

Andrew, Thanks for the insider information on E-machines... I won't ever consider recommending them again. Unfortunately, my church uses one to run power point presentations for Sunday morning worship. I will be keeping a huge eye on it. Hot glue is a common practice these days, and it actually works very well. It helps to eliminate some common shipping problems such as (as you mentioned) cables becoming unplugged, memory sticks knocked loose or displaced video cards (compaq actually has a special bar incorporated into the case that holds the video card in). If they hot glued anything else then I would have to wonder. Thanks again for the info. -Michael

I knew that e-Machines were pretty nasty but, DAYUM!

Well, I didn't mean to frighten e-Machine as such, but looking back over my previous post... well... it's all quite true and I wasn't exaggerating. (And I'm amazed all over again!) Supporting them was quite an... interesting experience. At first there's confusion, then shock, then horror, then when you learn to laugh at it, you laugh your ass off 'cause it seems like some sort of sketch comedy or sitcom. I do know that hot glue isn't that uncommon, but in the case of the ECS systems, many of the connectors had surprisingly little grip. (Once, when I was unplugging the data cable from a PowerSpec's hard drive as part of a demonstration during a training class for the call center support technicians, the plastic connector on the end of the cable just disintegrated! The small class of just under a dozen techs erupted into laughter.) It was a running joke that without the hot glue, the systems would just fall apart. Oh--and they hot glued just about EVERYTHING. As far as I remember, they used hot glue on every single cable in the case except for the power connectors, including the IDE/ATAPI data cables, sound card audio cable, all the cables for things like the power LED, soft power switch, speaker, etc. The problems I cited were with systems built by Elitegroup Computer Systems or by someone else using ECS components in general. Again, the e-Machines were clearly the top of the ECS line, but the top for ECS is hardly a dizzying height. The PowerSpec systems were the lowest. I would definitely keep an eye on the e-Machine. If you can find out who made the motherboard, you might rest a little easier. There were some e-Machines that were NOT equipped with ECS motherboards. In fact, even a few PowerSpec models sported boards from other manufacturers (the only one I can remember is Soyo.). If your e-Machine has a decent motherboard, well, that's a large part of the problems with ECS machines. Failure rates for ECS boards were unbelievably high. The other critical component I'd worry about is the power supply. Failure rates for power supplies in these machines were also very high. One thing you might check regarding the power supply is that the fan is an actual ball bearing fan, not one of the sleeve bearing variety. (Many of the el cheapo power supplies used in these machines were sleeve bearing fans! I was amazed that anyone would do something like this, but it happens all the time.) If the PSU fan is the sleeve bearing type, I would replace the fan with a ball bearing type. Ball bearing type fans cost a little more than the sleeve bearing type, but they are far more reliable and last longer. (According to Scott Mueller, author of Upgrading and Repairing PCs, ball bearing fans generally have a service life TEN TIMES longer than their sleeve bearing counterparts.) The oil in the sleeve will eventually dry up and the fan will seize. Some very good power supplies may be able to manage without proper cooling, but these... probably not. If you replaced an e-Machine's motherboard and power supply and made sure it had decent memory, you'd probably have a decent machine. It's a pity the parts manufacturers and vendors tend to skimp on happen to be the most critical to the entire system, but I can see why they do it. Numbers sell. If a vendor can advertise a CPU with a high MHz rating or a CD-ROM drive with a high speed rating (never mind that you could only get the advertised speeds on the outer tracks where there's probably no data to read and even then you can't get those speeds because of disc vibration), the better they sell. Nobody's systems are perfect and everyone--including Dell, IBM, Compaq, Gateway, HP, et al--has had issues. There are no saints in the computer industry, only businesses, and the bottom line for these businesses is the profitability they need to survive. You can argue that providing quality products and services is a good strategy for long-term corporate success and I'd tend to agree with you, but many of these companies either don't see it that way or simply have no interest in the long term. Even for the companies that seem to place greater importance on quality, there's occasional problems. Dell tends to make good systems, but you can still buy a bad Dell. It can happen. You can get a bad *anything*. Nobody has zero percent failure rates and everyone cuts corners somewhere. With ECS, however, the failure rates and poor quality are faaaaaaaaaaaar beyond anything you'd expect. I think they (along with PC Chips and Tseng) are defining the limits of unreliability and shoddiness for major vendors.

Just a sample of ECS satire. This was written at about the time Microcenter/MEI made a deal with Microsoft to give away "free" PowerSpec computers with a three-year subscription to MSN. (Ironically, these "free" PowerSpec computers couldn't function for the duration of the contract--they all had the Western Digital hard drives with the fatal reverse head seek problem!). Hell to Offer Free PCs, Eternal Suffering Hell Inc., the troubled Internet start-up (Nasdaq:HELL), will soon begin distributing free computers in hopes of boosting membership. Applicants must buy a three-year subscription to Hell's DevilNet ISP, a wholly owned subsidiary of AOL, to receive a free PC. "Recent free-PC ventures have basically tanked, but we thought, 'What the hell? Let's give it a shot,'" said Hell CEO Satan. "We see free PC distribution as a great way to set right Hell's bad name." Microsoft has also made a pact with Satan--each Cerberus 400 PC that Hell gives away will be loaded with $1,500 worth of Microsoft software. "We're really excited about entering this untapped market," said a Microsoft spokesperson. "The deal will undoubtedly put Microsoft products on the computers of every damned PC user. That's a population we've never been able to reach before." The first 100,000 people to take advantage of the offer will also receive a deluxe cherry coffin with a satin interior. A subscription to Hell costs just one soul and an eternity of damnation.