The corsair 750w, which I bought new in August 2011(5 year warranty) is the non-modular version of this ;-

http://www.corsair.c...wer-supply.html

So I presime that the spec is similar i.e. This page reports that it can supply 62A on the main 12V rail - a fair bit of that must go through the 24 pin ATX connector which I understand the memory slots draw tbeir power from.

As said it has been running without problems with this Asus M4 mobo + Athlon Ii 630 (Rated at 95w) + 1 HDD + 2 x DVD/CD combo + 2 x PCI cards + 5 x USB devices and 3 x chassis fans since August power supply calculator calculates the worst case draw of tbis setup to be 450w - even if you discount the Corsair's rated delivery by the usual amount to reflect every day power delivery it still delivering 600w. Substituting the Phenom (125w) for the Athlon (95w) only adds 30w. Why that should bust the bank is beyond me !

Similarly, Overclock net states that the M4 board has 4 + 1 power phasing and is good for a TPD of 125w.

Nick

Your power supply can clearly deliver far more than your system could ever need at maximum power draw plus I find those calculators tend to over estimate the power required as they have to account for cheap power supplies which may produce high wattage but poor current, your PSU is definitely a good one. I'm running my desktop on a 375W power supply with a Core i7 940 (130W TDP), six memory modules, two optical drives, AMD 4850, SSD, HDD, PCI-E soundcard and a PCI-E wireless card with a range of USB peripherals and the machine has never missed a beat even though it spends a lot of its time at full pelt.

How cheap can you get a compatible motherboard?

I should have known better trying to uograde a pedestrian workaday mobo into a games machine. Of course, five years ago, the relativey undemanding games, in hardware reqjirement terms meant you could do a simple processor + memory + graphics card upgrade on a medium grade mobo and have had something lime Falclon 4 or IL2 running well at decent frame rates without crashes.

It seems that modern games, for whatever reason, have raised the hardware requirements bar considerably, to the extent that even with the moderately powerful Phenom and Sapphire graphics board that I selected (Which are rated on the Cliffs of Dover site @ 7 - 8 out of ten), an ordinary mobo won't in practice support them.

Perhaps, in future, with all this IT surveillance that has been promised by HMG (For our welfare, of course), users like me will be spared from making IT hardware and software decisions that they regret by the timely arrival of an advanced notification e-mail from. Cheltenam or the depths of Wiltshire - sic.

That' s enough of my moaning, I will now go off, boyed with increased brand loyality aising from the knowledge that Skoda have just declined to make any financial contribution to the cost of replacement of my ARB or console bushes (By relying on the 2 year limitation on the warranty on replacement parts (Nips at the ready for my new car order)) and attempt to unglue the heat sink from the Athlon.

Nick

The only thing you have done which I see could cause a problem is changing the processor over.

You may have damaged the socket at a wild guess, other than you have not done anything really.

Have you checked each cpu for bent pins, they can be straigtened if your gentle.

I had trouble with a new Asus board a while back, brand spanking new, worked for a week or two, then dead as Dillenger, mem light stuck on too

A bit strange tbh !!

But then computers are a bit strange by their nature sometimes

The pins on both processors are straight as a die.What I have noticed is that the board is very slightly bowed.Possibly the extra force necessary to secure the Phenom was been sufficient to crack or break one or more of the llines ideep in the board. Alternatively, the extra power draw of the Phenom could have fried the Voltage Regulators on the board, as suggested in the link. I posted above.Even more alternatively it could be malware embedded by the manufacturer or a third party (I been getting a huge number of phone calls recently from the sub-continentally based "Windows Technical Department" fraudsters) in the BIOS.As an outside chance it could be the PSU.

Irrespective of the cause, I think I'll get a new board more suited to the Phenom. The Asus Tuf Sabretooth with the 990 Chipset, with its military quality certified components, looks good and will take the 965 Phenom and existing memory and perhaps a power supply tester.

Nick

I would say the cracked board sounds more likely,I doubt a virus is your issue, we always get the calls from "John" too.

I don't think they are that advanced to infect your bios tbh.

I got a psu tester you could lend I can post to you if you like, as long as I get it back of course, alternativley iirc it can be done with a multimeter

Thanks for the offer, but I should have one as part of my standard toolkit - Amazon have them for £11 -. £30. Failing that, I could at a push use my Fluke multimeter - though not overly keen to do that because of the high "****-up" potential.

Nick

Failing that, I could at a push use my Fluke multimeter - though not overly keen to do that because of the high "****-up" potential.

I agree lol, never tried the multimeter for the same reason !!

I should have known better trying to uograde a pedestrian workaday mobo into a games machine. Of course, five years ago, the relativey undemanding games, in hardware reqjirement terms meant you could do a simple processor + memory + graphics card upgrade on a medium grade mobo and have had something lime Falclon 4 or IL2 running well at decent frame rates without crashes.

It seems that modern games, for whatever reason, have raised the hardware requirements bar considerably, to the extent that even with the moderately powerful Phenom and Sapphire graphics board that I selected (Which are rated on the Cliffs of Dover site @ 7 - 8 out of ten), an ordinary mobo won't in practice support them.

I don't think that's the case at all, budget boards may not have all the bells and whistles nor be able to handle decent overclocks but in your case there should have been no issues running the Phenom processor otherwise it wouldn't be listed as compatible, I think you've just been unlucky and something has broken as you never got anywhere near the stage where your components were at maximum power draw.

Yes. Point taken. The history in this case doesn" t support my assertion. I should have prefaced IMHO.

The problem I had in this case was whilst the processor fitted the socket, the extra heavy duty heat sink that was supplied with the processor in the AMD retail package had thin sheet copper wings projecting from the base of the heatsink (See photo link above) and on the version of the socket that I had on my M4 motherboard, these wings interferred sligthly with socket and made joining of the processor and heatsink a little problematic.

Nick.

I always worry when putting a new board together, the board screw layout does not support ATX boards properly and they do flex alarmingly when you are trying to insert even a couple of sticks of RAM, or the main power cable.

I try to do as much as possible with the board out of the case, and laid on an antistatic foam sheet (or A.S. bag on top of some carpet), before fitting it inside the case.

Pulled the board out of the case and totally disaasembled everything including separating the processor from the heat sink.

Then had to adjourn to the dealers to pick up the motor.

£500 lighter , I returned QTH and resumed playing around with the M4 mobo.

To my horror, I found that I had managed to bend a couple of pins on the Athlon CPU and x 6 pins on the Phenom ! = theres a first time for everything. !

Straightened the pins on the Athlon, thermal pasted it and managed to re-install it into the socket after a bit of tooing and froing. Phew !

Powered up the board on my "Test Bench" and then played around with the combination of the 4 DIMMS until I got it to boot ! Hurrah.

Unfortunately, it will only boot with 2 DIMMS one each in Slot A1 (Blue) and Slot A2 (Black). Its quite stable in this mode - I writing this post on it. However, that means I've only currently got 4 GB of memory (2 x 2GB) availalble.

Further, according to the M4 manual, this memory configuration is not a valid combination of memory and slots.

Permissable combinations are:-

1 DIMM in any of the 4 DIMM slots

2 x DIMM, paired in either the 2 blue slots or the 2 black slots,

or all four DIMM slots populated with a DIMM.

I've tried either a single or a pair of DIMMs in slors B1 (Blue) and B2(Black) and all four slots populated and it will not boot.

I checked the BIOS reports and all the timings, frequencies and voltages are in range and rock solid.

I wonder if this indicates that there is another processor pin bent or missing on the Athlon that i didn't spot ?

As for the Phenom. that can wait 'til tomorrow. that's a real mess.

Nick

Edited 3 April, 201214 yr by Clunkclick

Power supply tester arrived this morning. I got the following readings with both 12 Volt mobo.plugs connected to the machine:-

-12 volt line . . . . . . ..... .....11.70 volts

+12 volt line (V2 molex). . .12.00 volts

5 VSB line . .... . .. . . ......... 5.00 volts

PG . . . . . .. . . . . . . . . . .. .270 ms

+5 volt . ..... .... . . . . . . . . . .5.00 volts

+12volt line(V1 ATX) . . . .....12 volts

+3.3. Line . .. .. .. . . .... . . . .3.3 vllts

So, according to the Wiki page on ATX supplies, the above figures are within laccdptable limits.

So unless I can get DIMM slots B1 and B2 working then its going to be a new board.

Nick

Now I remember why I stopped building my own PC's.

It gets better.

I decided to buy a new set - up. The mobo and memory arrived yesterday. So I set it up with the existing Athlon II 630 4 core and it wouldn"t boot to the BIOS. Despite my efforts yesterday and today still can"t get it to boot. According to all the official sources, all the new bits I' ve bought are compatible with all the rest. Looks like its a mobo problem and its a top-quality mobo -Asus Sabretooth FX 990.

Deep research on the web has indicated an enduring problem with on-board SATA 3GB drivers provided by an Asus supplier J Micron and this conforms to the situation I find where the POST stalls at the point where the POST routine tests the drives. Apparently, you can turn-off this driver but it can only be done from inside the BIOS.

So, if I don' t get it fixed in the next couple of days that's another pile of **** that's going to be RMA'd back to Amazon.

In contrast, my previous board, the Asus M488TD-Evo USB, with integrated graphics, worked first time straight out the box.

Nick

Edited 6 April, 201214 yr by Clunkclick

Now I remember why I stopped building my own PC's.

When I was reading this topic it did remind me of an older AMD64 3000 build I had no end of trouble with, I'd moved to a hybrid socket 754/PCI-E board as I'd got a cracking deal on a PCI-E graphics cards when PCI-E was just out. However when I put it all together, the system wouldn't boot - everything powered up but I never saw anything on screen and didn't receive any beeps. I spent a long time trying to track down the fault without much success, part of the difficulty was that I didn't have any other PCI-E graphic cards to test (the ram, cpu and HDD/ODD drives were from an existing machine) nor could I borrow any from work plus the board had no onboard vga. I ended up buying a budget PCI-E graphics card for the machine but still I couldn't get anything out of it. So I forked out and paid for an RMA but the company said the motherboard passed all tests but when I asked which power supply and ram they were using they never gave me any more specific detail. I happened to see a good deal on a Thermaltake copper cooler which I went for (I'd been impressed with my flatmate's cooler of a similar design), installed it and PC booted straight away...turned out there was a problem with the existing cooler and the motherboard I was using, the heatsink touched one of the mosfets (or one of the other parts beside the socket) which clearly stopped it booting.I've also had trouble with Opterons and ram combinations, had a quad socket server that would not boot until there was much fiddling around with ram and it booted with a very strange combination that wasn't apparently supported either according to the manual (none of the supported ones worked either!). However as the machine was now working with all the memory it was left like that.John

Solved it yesterday, much along the lines you described, by using a 10 year old Savage S3 with VGA only output pulled out of a mini-server. It seems that the Sapphire i6870 initiallly only produces VGA output, even though its 2 x output ports are DVI only, duh ? Further, there was a secondary problem with the illyama LCD. Again, even tbough I was using the DVI socket on the screen, because the video card wasn't outputing a DVI signal it wouldn' t switch its input feed to take from the DVI socket - I wonder if that lack of that connection in turn meant that the video card did get a prompt to switch its output to DVI -a lovely chicken and egg conundrum.

I broke this loop using a DVI to VGA adapter and got video.

Once I got a screen, I was able to get into the BIOS, update it and load chipset drivers and, apart from having to re-install Windows 7 (Which wouldn't boot, or be repaired or updated), it was easy after that.

Unfortunately, the Phenom didn' t stand up to my first attempt to un- bend the pins and one broke off. That, of course, may not be critical, if its a reserve pin and I have seen videos on You Tube where the broken pin has been just been dropped in the socket in the correct position and the CPU has booted. So that's going to be a mini project in slower time.

In the event I purchased a six core Bulldozer to replace the 4 core Phenom and was surprised to see that it uses less power than the Phenom (95w as against 128w) and is ranked higher up the megaflop benchmark list.

The Bulldozer goes goes like stink - it was multithreading during the re-install of Windows 7, and considerably reduced the load time. A huge improvement on the Athlon 4 core.

Nick

Glad you got it sorted,,,, and interested in the bulldozer comment, I need to update my systems this year, and it is compatible with both my PC mobo's

I discovered yesterday why the Athlon II had had difficulties booting either the old M4 or the new Sabretooth board. There was just one severely bent pin at position A28. According to the pin diagram on AMD's datasheet for that processor family, pin A28 carries a Write enable signal for the DDR memory. So that might explain why the M4. board was booting with DIMM slots B1 and B2 inoperative and why a non-stadard DIMM confivuration was working in slots A1 and A2.

Even if booting in these circumstances was more by luck than design, I'm impressed at the robustness of these little processors.

Examination of the A 28 pin position in AM3 Socket on the M4 board showed that the pin had been so severely bent that it was touching its next door neighbour and had left a clear iimpression in the plastic of the socket surface. Obviously, it had been shorting-out, biig-time, a memory circuit.

Clearly, the bent pin couldn't be left in this state if I wanted to use the processor again. So, using eyebrow tweezers, plastic cards and miniature technical flat-bladed screwdrivers, I set to work. I managed to get the pin upright and straightened in one direction. Then I tried to complete the straightening in the other direction and . . . Disaster . . . it broke-off at the base.

So todays task is to find and cut a CPU pin length of bell wire (I reckon that will be pretty close to the conducting properties of the original gold plated copper and aluminium CPU pins), pop it into the A28 pin hole in the socket, mount the processor on-top and see if it flies.

Video to follow, if it works !

Postscript

Used a cut-down portion of a No 56 staple, instead of the bell-wire. And it seemed to do the trick. I discovered that in fact it was pin 27, which is a simple grounding pin (VSS). When it was bent over it was shorting out against Pin No 26 will has a signal on it to control DDR input/output.

Net effect is that the old M4 board is now up and running, although I still can't get the DIMMS B1 and B2 populated without it refusing to boot - hence only 4GB onboard. So looks like the abortive attenpt to fit the Phenom did do some permanent damage.

Nick

Nick

Edited 10 April, 201214 yr by Clunkclick

Have you seen there's a couple of Windows 7 hotfixes that are meant to improve performance by handling the Bulldozer core design better? They don't make for a big difference but still worth having.

Cheers.

I'll have a llook for the hotfixes.

Five days on, the Athlon "Repair" is still holdiing up - I was expecting the staple to have melted by now. !

Next job is to unbend the Phenom pins - hairddryer and disposable syringe at the ready.

Nick

Edited 14 April, 201214 yr by Clunkclick

Looking at the no power on thing, reminded me of a problem my son had with one PSU/MB. Bad contact with the soft power pins. Often he had to short out the soft start pins on the rear of the PSU plug.

OH,Nick - just seen your sig- you've missed out the Campbels & Tinkers

Edited 17 April, 201214 yr by VWD

Looking at the no power on thing, reminded me of a problem my son had with one PSU/MB. Bad contact with the soft power pins. Often he had to short out the soft start pins on the rear of the PSU plug.

OH,Nick - just seen your sig- you've missed out the Campbels & Tinkers

New categories to be added:- slow-off-the-mark car insurance brokers, bearded-wonder telecommunication providers, all sub-continetal branches of the incomprehensible "Windows Technical Department" , monolithically incompetent financial service providers and, of course, in-denial fault-tolerant car manufacturers (Guess who ?).

"Victor . . .its time for bed".

Nick

Edited 19 April, 201214 yr by Clunkclick