Asus M4A88TD-M EVO socket AM3 motherboard currently running with AMD Athlon II, on-board Radeon HD 4250, 8 GB of Corsair XMS 3 memory, Seagate 500 GB Sata Hard Drive and powered by Corsair 750w enthusiast power supply (Bronze).

I want to upgrade the processor and graphics so that I can run UBIsoft's Cliff's of Dover.

According to the CoD and Asus websites the following hardware which I have purchased is compatible the existing motherboard and rates with CoD @ 7-8 out of 10:-

AMD Phenom II processor 965 (Black edition, retail) 3.4Ghz, HDZ965FBK4DGM, C3 stepping, 125w.

Sapphire HD 6870 graphics card.

So, I started, after cleaning out the cabinet, by removing the existing Athlon and replacing it with the Phenom and the supplied heat sink/fan, making sure to align the triangle mark on the processor with the corresponding mark on the m/Board. No problems, to that point.

Powered up, disk lights lit, but no video or POST routine. Repeated, but still no joy.

Removed CMOS ram battery, shorted out the terminals, waited and then re-booted. Same again, disk lights came on, but no video or POST.

Suspected a hardware conflict, so removed memory SIMMS and disconnected hard disk, DVD drive and cycled the CMOS battery again. Re-booted, still the same.

Checked the install of the processor, looked and felt sound as a pound. Hit the mem-reset button on the m/board. Still no difference.

According to all the sources of advice I've referenced on the web, this processor/motherboard combo should be fine.

But somethings blocking video output and the POST routine - that"s fundamental.

Just wondering whether I've got a bit of Amazon "Fire damaged" Stock ?

Any ideas welcome ?

Nick

Do you mean a 'Phenom IIX4 965' - that is listed as being supported by the Asus website for your motherboard (http://uk.asus.com/M...M_EVOUSB3/#CPUS)

Have you switched your computer back to the original CPU to check it still works with that? If so, then I would do the following:

- disconnect all HDD/DVD/etc drives

- remove all the RAM

- insert new CPU, HS & fan

- try powering it up - see if you get to the BIOS

Try turning the the machine on a key onthe keyboard held down - see if that generates a keyboard / key stuck error message (and beep).

If still fails, assume CPU faulty.....

Edited 30 March, 201214 yr by mbames

Have you updated the BIOS? If not, sounds like dud/unsupported CPU to be honest.

Yds that's the one.According ro Wiki ghere are two types -stepping C2 @ 140w and stepping C3, 125 w.

I'll re-install and see what happens.

Nick

Re-installed old processor.

Still no video or POST or beeps.

Tried everything - Holding down del key on power-on;Holding down the memory re-Sync key on the motherboard. Powering up with only one memory DIMM. Clearing the CMOS; Removing the CMOS battery.

I suspect that the Phenom, being a gamer"s processor @ 125 w power(Compared to the Athlons 95w)has fried either the on-board GPU/Chipset/Cmos.

Also, it looked as if the CMOS clear jumper had been parked in the wrong position from the outset, so that the CMOS may have already been trashed before I started the processor swap - god knows how that happened. If that was the case then changing the processor would have blitzed any remaining values in the CMOS and after that it would be unable to take new ones.

Looks like. I may be in for all new full set processor, memory, motherboard. Lucky, I didn' t install the Graphics card.

Just in case its a correctable CMOS thing, I'll try pulling the battery and powering it down overnight.

Nick

Any beeps without the CPU? It sounds like something might have fried your motherboard, or your PSU is now dead (not not supply the correct amount of power on the requested lines).

I don't think the clear CMOS jumper being in the wrong place is likely to the root of the problem. The BIOS should still work even if the CMOS was corrupted/not set.

Might be worth popping to a local computer shop and see if they can test the various bits for you before you shell out on a replacing everything.... might cost you a few quid.....

No beeps CPU out, but the system powered-on ! Is that supposed to happen ? Unswitched power to components attahed to the Motherboard ? I thought that removing the CPU would kill everything on the motherboard, including power distribution.

I hope its not the power supply that's only a year old.If it was that it would be without prior indication.

Nick.

Edited 30 March, 201214 yr by Clunkclick

The motherboard should be able to detect a faulty/non-inserted CPU and make suitable beeps that indicate so.

Maybe your new CPU draw a bit too much power and broke the PSU? Still you said it was a 750w unit which seems a pretty fair size to me - the last machine I built had a 400w unit, apart from the dual xeon which is under my desk at home (and consumes power like it is going out of fashion)

Did you check that you had the latest BIOS ? It may be the mobo detected the new cpu, set itself up and became unstable/locked up due to a firmware or BIOS coding fault.

Have a read of the mobo manual to find how to use CrashFree Bios 3, then try installing the latest BIOS via a usb drive.

Thanks. I was aware of that.

According to the references as long as the BIOS number exceeds 1012 then it should handle the Phenom.

Whilst my BIOS hadn't been flashed to the most up-to-date version 1017, the board was a late purchase in the middle of 2010(Manual dated 2010) and therefore, I suspect had a post 1012 BIOS.

At the moment I've lifted the board out of the case, completely removed all attachments and have given it a thorough blow-thfrough/vacuum and inspection - nothing to be seen. I just wanted to check that nothing was shorting it out on the reverse side. Later today, I will re-assemble and re-install - I've often found in the past that just allowing an extended time with power-off is sufficient to clear any "Stuck" data bits. I often wonder whether these things might get affected coming through the post if they happen to go through an X- Ray scanner

A quick check on the Asus power supply calculator showed that the Phenom II 965 would actually consume less overall power than the Athlon II 630 (500w/550w):-

http://support.asus....owersupply.aspx

I can't beleive that.

I stand to be corrected, but I still think that the Phenom may have fried the board.

Taking AMD' s power consumption off of their processor web-site:-

Athlon 630

95 watts, volts 0.9 - 1.4. So calculated max amps would be 106 (Rounded).

Phenom II

125 watts, volts 0.825 - 1.4. Calculated max amps 152 (Rounded).

I.e. 50% higher. Frying tonight !

Yet you are faced with the fact that the Phenom fits the socket (AM3, 941 pins) and Asus declare it as compatible.

Anyway, technically the board is still under warranty . . .. .

We shall see what happens later today

Nick

,

Edited 31 March, 201214 yr by Clunkclick

The only time I have had a PC that gave no beeps on boot up was when the motherboard was dead. I usually fit a motherboad, CPU and fan together to make sure they are compatible.

Well I re-installed the board again and, you guessed it, same as before, no beeps, video.

Also, although, on power-up it appeared to be reading the crash-free BIOS disc in the DVD drive, but there was still no video.

I suspect that the Phenom frizted at least the on- board Radeon graphics circuitry and possibly other stuff. And that would stall the POST software early on in its routine -hence no beeps. If other motherboard circuitry is OK then I may well still be able to run the thing using a plug-in graphics card. Taking that line of tbought a bit further, I'm just wondering whether the POST software when it detected the Phenom switched graphics requests from the on-board graphics to the PCI graphics slot., which is vacant. But this theory, wouldn't explain why the thing fails to boot after a pull-out and disconnect and subsequent re-installation of the ofiginal processor.

I have a gash and rather ancient. S3 savage PCI graphics card which I used in the past to test boards, so assuming the original PCI edge connector goes into the new PCI slot then I may be able to run it. It that card goes phut, it won' t cost me anything and will fairly conclusively prove that the mboard is screwed.

Nick

Have you tried the old cpu again ??

I doubt the processor has fried the motherboard simply because it is a higher TPP, you've mentioned the processor is meant to be compatible, it sounds like you have a decent power supply and the TDP limit more specifically refers to the thermal dissipation of the processor. It's a vague number as it doesn't refer to the absolute maximum amount of heat the processor can dissipate instead it's meant to be the upper thermal limit when the processor is running at full pelt for sustained periods and is particularly important for laptops which have minimal space and compact cooling systems. Obviously something that can dissipate more heat will need more power but there doesn't sound anything wrong with your setup and I wouldn't think the processor would be able to immediately draw full power hence I tend to find in practice if there's something drawing too much power the system will start but the machne will randomly and suddenly restart without warning.

Bios-wise I wouldn't be surprised if your bios was out of date as the motherboard could have been sitting for some time and manufacturers generally don't seem to ensure devices they're producing ship with the latest firmware/bios. Unfortunately I doubt the graphics card on its own is just fried as it is integrated into the motherboard chipset which means if the board has been shorted, other critical components will also be damaged.

You've taken all the steps I would recommend, stripping down to basics, checking the board and the pins for any short and resetting the bios to defaults. Frustratingly few boards seem to have any sort of diagnostic system for pre-POST issues, it's very handy on OEM systems where they have even something as simple as a quad set of LEDs which show the exact point where the board has stopped in its start up sequence.

John

Update.

After doing a second re-install of the Athlon, this time doing it properly (Putting the processor in first followed by the heat sink - those thermal pads which come pre-fitted on the AMD retail product heatsink are a bugger to get off - hairdryer + penknife), I got it to beep with no DIMMs installed.

One long, two short, repeating.Web guidance says that tbis is memory. So after a bit of playing around, it would start to boot and then the memory LED on the board would come on constant. The advice in the Asus board manual, is that this indicates incompatible memory is being used. Well, this is obviously a false indication as these memory sticks (4 x 2GB Corsair XMS (CMX 1600. 1,65v, matched pairs) have been used from day one with this machine without any previous problems and are listed in the Qualified Vendors List in the Asus manual as fully compatible.

The advice in the Asus manual about getting a constantly illuminated memory LED is to depress the "Mem OK” button (Next to the LED) at power on and then the machine goes on to attempt to derive a workable set of latency values. Whilst it is searching the memory LED will flash and the rate of flash quickens the nearer it gets to an optimum solution. If it finds a set it stores them in CMOS and then restarts the machine and the LED is extinguished. If it doesn' t find a solution, then the machine should automatically restart and continue the search - LED continues to flash.

Well, after numerous cycles of this process, I got two types of outcome;-

1. LED kept on Flashing,no matter how many shut downs it did.

2. LED would extinguish after one or two cycles but then the machine would shut down during the power up that followed and wouldn't stay on no matter how many times afterwards I pressed the power on button.

The process didn't work properly because I was having to start the macine manually after each machine induced close- down.The two types of outcome appeared to occur randomly and independent of which memory slots were populated or how many memory sticks were used.

Cycling the CMOS memory by shorting it out (Only after turning off the power, pulling the power lead and removing the CMOS battery) wiped these "Optimised" memory latency values Nd the process started from step one.

In my view the Radeon graphics embedded in the motherboard are buggered because the POST procedure seems to reach the next step on from Video checks i.e. Memory checks before it stalls, but there"s sttill no video.

And the memory ciircuitry on the main board is fritzed as is the CMOS.

So all of the kit I' ve installed so far may be affected i.e motherboard, Athlon processor, Phenom processor and 8 GB of Corsair memory. All in the bin (Sounds like Basil Fawlty ?).

I can' t be sure of the. Corsair 750w power supply. That may have to go unless I can conclusively verify the output -- and that will require a £30 tester.

I seem to recall that I had similar symptoms with a previous Asus board - An A8. But in that case I knew it was faulty, as I remember seeing the puff if smoke come off it after I dropped a screwdriver on it whilst it was powered - silly boy.

Nick

From what I recall of that chipset, the Radeon graphics card is not a discrete part as it's integrated into the main motherboard chip which also contains the PCI-Express controller and the links to southbridge (only the memory is separate due to the on-die memory controller).

I'd be willing to bet that if the board is borked, all the other bits will be fine, I'll be honest, it does sound as though something is dead.

Whenever friends have had issues with onboard graphics there has been a way to disable as if installing a separate dedicated card.

Is this possible on your board? This should give you some beeps or at least allow the memory check to complete.

If it is any consolation, I had a 2 month old Gigabyte board that scrambled the pci timings for the wireless card and shortly afterwards "broke" the onboard gigabit ethernet port, I swapped the board TWICE and experienced EXACTLY the same two fault on both of the new boards within a few weeks.

According to the people I bought from, it is a known problem with the Dual Bios systems from Gigabyte.

On the other hand, I have never had a problem with an Asus board.

True, the focus of the problem appears to lie at the moment with the memory, but it could lie anywhere upstream i.e. CPU or PSU. The fact that the error outcome conditions encountered during power up are consistently of two types may suggest that it isn' t generated by the PSU, which I suspect would be more random.

That said there are plenty of reports on the web of PSU being instrumental in failures to boot for reasons ranging from broken supply wires on new units, to dodgy CMOS batteries, to insufficient component loading on high-power PSUs.

I'm looking at the moment at the CPU socket, which may have been slightly distorted by the installation of the Phenom - as you can see from the photos in the link below the heat sink supplied by AMD with the retail package (Left) is a far bulkier affair than the Athlon:-

http://www.flickr.com/photos/54474784@N05/7034785247/in/photostream

I did have a bit of a job securing the metal strap of the Phenom to the lugs on the socket surround of the mobo and when I came to reinstall the Athlon following the boot failure using the Phenom, I found that with the socket arm raised and the Athlon inserted, tbe Athlon CPU pitched very slightly between diagonally opposite corners, though this effect disappeared when the socket arm was. closed.

If this yields no joy, then I think iI'll start topdown with the PSU -that will mean buying a tester.

Nick

Have seen this problem many times, here is a test.

ensure the power is not plugged in

remove all RAM

power up

You should get POST error beeps

Next remove power again

add 2 of the RAM sticks

Power up again

If it now comes on this is the PSU not able to provide enough power

It is very misleading the PSU power rating system, the rating is the total power the PSU can provide ie 750W

what it does not tell you is how this is distributed, this is on a label on the side of the PSU and can vary from PSU brand to brand.

If yours does not have enough in one area the symptoms you are getting show up.

Hope this helps

John

Edited 1 April, 201214 yr by jjc

As regards testing its getting a bit more inconsistent since yesterday. Yesterday, with all ram removed, I was getting one long + 2 short beeps (Endlessly repeating) on power up. Today, when I repeated it, nothing.

But I'll give it another go with your m ethod tomorrow.

Just to clarify that's, power off and remove all ram. Power on and if beeps, then power off and install all ram (In this case 4 x 2GB DIMMS). Power on and then off. Remove one DIMM, power on. If it boots after this then suspect PSU.

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

Edited 1 April, 201214 yr by Clunkclick

sorry did not see you had 4 memory modules so after the initial power up with no memory and you get the beeps

power up again with 2 modules fitted

Have amended original post so it reads better :-)

No definitive idea why this happens on system that worked OK before upgrade but have found exactly the same as you are seeing on two other occasions. Sometimes it can even be just upgrading the memory modules, though this system had been a bit temperamental to start up for a while.

John

IRe-testing will have to wait ti tomorrow, as I pulled the CPU for a second time today to see whether I would get any beeps with the CPU out on power up.Answer no. Now I got to unglue the CPU from the heat sink - hairdryer to the ready.

Busy and expensive day tomorrow.Getting the console bushes and ARB (Second time) on the Fabiia done. So if the Desktop PSU has gone as well hat's me kissing goodbye to £800.

Hopefully, the nice Mr Holland from Skoda Customer Services will respond to my Thursday call and advise me how much of the Fab's repair cost they are going to meet.

Nick

I just wonder whether this afticle sheds any light on what's happening:-

http://www.overclock.net/a/about-vrms-mosfets-motherboard-safety-with-high-tdp-processors

Nick