Memtest86 not seeing 4gb of ram in DFI SLI DR board

[agent_x007]

@N2 Gaming Just to be clear : Under normal circumstances you never tick maximum memory box under msconfig.
Yes, overclocking RAM may make it unstable, and limiting it then can help with stability (not really, you just tell OS to not use some of unstable part of memory), but that's not what you want in this case.

CPU-z probably detects memory by SPD info, regardless of how much of actual capacity can be used by system.
That's why you get 4GB under Memory tab in CPU-z and only 2GB in System info.

Again, it's not OS issue as BIOS itself only sees ~2GB, so it doesn't matter what OS you put in (or how many times reinstall Windows/Linux).
OS is ultimately limited by how much memory is seen by BIOS, and Your BIOS only detects ~2GB of RAM.

Not english language, but at least numbers/settings should be recognisable :

 

[N2 Gaming]

Yes I installed it with only 2gb or ram because that is all I had at the time of OS install. Os was installed a couple of months back during the HWbot Team Competition.

It's not the bios version, it's your OS. Was this a fresh install?

I can't wrap my head around this idea. what agent is saying makes way more sense to me. How can the os make the bios see or address more ram on an older non UEFI bios board?

 

[Mr.Scott]

Yes I installed it with only 2gb or ram because that is all I had at the time of OS install. Os was installed a couple of months back during the HWbot Team Competition.


I can't wrap my head around this idea. what agent is saying makes way more sense to me. How can the os make the bios see or address more ram on an older non UEFI bios board?

Just try it.
If I'm wrong I'll eat crow.

 

[N2 Gaming]

Ok I just got a couple of new 512mb ssd drives yesterday so I'll try a new win 7 install on one of those when i get home.

EDIT: Does anybody know where to get some fresh crow meat??? no luck with fresh install :/

While looking for a web page that lists Ram by manufacture and timings to indicate memory chips like TCCD or BCCD or BH5/CH5 etc I stumbled upon this tool that I have yet to try from passmark called RAMMON. https://www.passmark.com/products/rammon/

Anyone every try it?

 

[Bones]

I've never heard of it, much less tried it but I guess it can't hurt to give it a go.

 

[delshay]

I've never heard of it, much less tried it but I guess it can't hurt to give it a go.

It will show the same info as CPUZ but in more detail, but It will not fix the issue. I still think it's a CPU controller or motherboard fault. Remember OP has a de-lidded CPU. I have broken a few CPU on my first three tries, but I have now mastered my de-lidding skills.

 

[Mr.Scott]

EDIT: Does anybody know where to get some fresh crow meat??? no luck with fresh install :/

Crap.

 

[N2 Gaming]

It will show the same info as CPUZ but in more detail, but It will not fix the issue. I still think it's a CPU controller or motherboard fault. Remember OP has a de-lidded CPU. I have broken a few CPU on my first three tries, but I have now mastered my de-lidding skills.

My curiosity with it is if it can tell the type of Ram chips are on the module or stick of Ram like BH5 etc.

Crap.

I don’t expect anything here but if you ever get the urge

It was suggested to me over at OCN that if and when I flash the bios to use AWDFLASH.EXE instead of the Award Winflash windows based tool . It was said that Winflash has been known to cause problems and including boards being bricked from a Winflash. With that said I went surfing and and discovered a lot of AWDflash versions. I’ll have a look to see if I have a copy already that came with the board but if not is there a known version I should be using to flash the bios.

This page has a list of many different awdflash.exe versions

Award Flasher (awdflash, winflash) utility- Download all available Award flasher versions

Download all Award BIOS Flasher (awdflash, winflash) utility versions from Wim's BIOS

www.wimsbios.com

The user manual does not specify from what I can see what version of awdflash.exe should be used. Page 116 talks about using awdflash.exe

EDIT:

Being one to always overthink everything it occurred to me to ask if the problem really does exist because of the cpu then is it feasible to think that it might not be wise to flash the bios with this cpu???

 

[Mr.Scott]

For DFI Ultra D, this one.

 

[delshay]

My curiosity with it is if it can tell the type of Ram chips are on the module or stick of Ram like BH5 etc.


I don’t expect anything here but if you ever get the urge

It was suggested to me over at OCN that if and when I flash the bios to use AWDFLASH.EXE instead of the Award Winflash windows based tool . It was said that Winflash has been known to cause problems and including boards being bricked from a Winflash. With that said I went surfing and and discovered a lot of AWDflash versions. I’ll have a look to see if I have a copy already that came with the board but if not is there a known version I should be using to flash the bios.

This page has a list of many different awdflash.exe versions

Award Flasher (awdflash, winflash) utility- Download all available Award flasher versions

Download all Award BIOS Flasher (awdflash, winflash) utility versions from Wim's BIOS

www.wimsbios.com

The user manual does not specify from what I can see what version of awdflash.exe should be used. Page 116 talks about using awdflash.exe

EDIT:

Being one to always overthink everything it occurred to me to ask if the problem really does exist because of the cpu then is it feasible to think that it might not be wise to flash the bios with this cpu???

OCZ used Infineon but most are fitted with cheaper chips can't remember the name at this time. They act almost the same as Infineon at low frequency but don't overclock as good as the Infineon chips.

EDIT: I think the cheaper chips were AEON, which I call rejects.

 

[N2 Gaming]

For DFI Ultra D, this one.

That will work with the SLI-DR? What version is that?

 

[Mr.Scott]

SLI-DR uses the same bio's as the Ultra D.
Flash utility's come packaged with the bio's. No idea what version.

FYI, I have all of the bios's for those boards, stock or modded.
Any particular one you're looking for?

You need to visit my site sometime for all this old stuff.
Link is in sig.

 

[Bones]

It will show the same info as CPUZ but in more detail, but It will not fix the issue. I still think it's a CPU controller or motherboard fault. Remember OP has a de-lidded CPU. I have broken a few CPU on my first three tries, but I have now mastered my de-lidding skills.

This could be it, if the delid job was botched then there's nothing you can do short of finding a damaged surface component to replace will take care of it.
If you have another chip to try it with I'd do that and see if that changes anything.

SLI-DR uses the same bio's as the Ultra D.
Flash utility's come packaged with the bio's. No idea what version.

FYI, I have all of the bios's for those boards, stock or modded.
Any particular one you're looking for?

You need to visit my site sometime for all this old stuff.
Link is in sig.

Yes - Do pop in anytime you'd like, we'll be there (Somewhere).

 

[N2 Gaming]

Registered over there just waiting for the Aye Oh Kay from the Mod! In the mean time I was reading this thread and something OZZ said made me think about an issue the board could potentially have from when I was using an Opty 180 with bent pins I straitened out with a clicker pencil and a credit card trick. OZZ said he was able to remove the cpu socket thing and look at the pins and see that some of the pins were opened up more than normal and potentially was not making contact. Oh boy just what I'd need if this was the case :/ Hopefully the 180 did not damage the socket and it's not the socket and it is the bios or cpu. Will also try the long clear cmos trick mentioned in that thread.

[quote author="@ozz"] cant be sure as i really cant remember but i had a problem on a board similar to this a long while ago now and i think you can take the socket cover off and it exposes the pins, 2 pins side by side like in a y shape and the cpu and pins slide into them to make contact when you lock the cpu in, i found 1 of the sides was opened up more than the others, so i got a small pointy tweezers i think and closed the pin entry gap up on all the pins and it fixed my problem, damned if i can remember what was wrong with it back then or what board it was, something you could check anyway, may help , may not

 

[delshay]

This could be it, if the delid job was botched then there's nothing you can do short of finding a damaged surface component to replace will take care of it.
If you have another chip to try it with I'd do that and see if that changes anything.

It's not a surface component they all capacitors. If any of the capacitors are shorted computer will not boot. It is stressing of the PCB itself. If you look carefully, you can expose copper trace(s). This does not mean this is the fault as I think that trace is just a screen ground plane & you can get away with scratches like that. I suggest OP should try another CPU. If this does not clear the fault even with a BIOS update, then the motherboard is fault, ie memory socket fault possible broken crack trace, but my money is on the CPU. ...Remember we are talking memory mapping fault, that's on the CPU.

I say try the de-lidded CPU on another motherboard or get another cheap CPU around 5 USD & test, but the very first thing to try next is any compatible BIOS update.

 

[N2 Gaming]

This is my only socket 939 board so no way to test cpu in another board. I did just purchase an X2 4400 and it has not shipped yet. Hoping to have in by next week end if the seller ever ships the dang thing out.

I recently saw a brand new A8N32 SLI Deluxe but seller was asking $150 so I made a low ball offer and it sold to some one else in the mean time :/ . I have a couple of AM2, AM2+AM3, AM3+ boards like M2N32-SLI-Deluxe, Foxconn Destryoer, NF980-G65, GA-990FXA-UD3, CHV-FZ Heck I even have an older socket A EPOX board practically brand new with an old Duron 900 or 800 in it. Any way crossing fingers the bios is the issue.

I'm going to take a close look again at all the caps to make sure none are bulging outward on the tops.

 

[FreedomEclipse]

I bought three of these motherboards off EBAY some time ago & every single one had a fault. I still have them here. I could repair them as 3 into one I should be able to do the repair, but I can't be bothered.
I have more important projects/hardware modding to do & complete.

Im guessing those boards all need re-caping.

 

[delshay]

Im guessing those boards all need re-caping.

Possible, but that's not the fault. I did suggest OP should try 3 sticks of ram to see what appears in windows. It seems to me something is not fully active when four DIMMS are inserted. I would like to know what it shows when 3 DIMMS are inserted in any order/slot.

 

[Bones]

It's not a surface component they all capacitors. If any of the capacitors are shorted computer will not boot. It is stressing of the PCB itself. If you look carefully, you can expose copper trace(s). This does not mean this is the fault as I think that trace is just a screen ground plane & you can get away with scratches like that. I suggest OP should try another CPU. If this does not clear the fault even with a BIOS update, then the motherboard is fault, ie memory socket fault possible broken crack trace, but my money is on the CPU. ...Remember we are talking memory mapping fault, that's on the CPU.

I say try the de-lidded CPU on another motherboard or get another cheap CPU around 5 USD & test, but the very first thing to try next is any compatible BIOS update.

Yes and no.
With my first attempts at delidding, I took the top off of a chip and nicked one of the components on the surface of the chip's PCB that's underneath the lid - Didn't kill the chip but it lost it's L2 cache.
That's what I'm taking about - If one of those components got nicked during the delid it could have an effect similar to that.

 

[delshay]

Yes and no.
With my first attempts at delidding, I took the top off of a chip and nicked one of the components on the surface of the chip's PCB that's underneath the lid - Didn't kill the chip but it lost it's L2 cache.
That's what I'm taking about - If one of those components got nicked during the delid it could have an effect similar to that.

There all capacitors. For example, I operated my phenom with four missing capacitors for a while before replacing them. Did not see any difference in stability when four was missing.

If you damage any of the capacitors when de-lidding, the best thing to do is remove them completely. Then find a cheap or even a non working CPU & de-soldered the caps from that & replace the missing caps on your CPU. This is what I do if I crack the capacitor package. Another thing to note if you crack the capacitor package you must check for a short before powering up the CPU as you can blow-up FET if a short is present.


OP Photo: below I have highlighted the ceramic capacitors that's linked to memory. Most or all of them should be memory.

 

[N2 Gaming]

Possible, but that's not the fault. I did suggest OP should try 3 sticks of ram to see what appears in windows. It seems to me something is not fully active when four DIMMS are inserted. I would like to know what it shows when 3 DIMMS are inserted in any order/slot.

Tested using just 3 sticks of ram with one stick missing in every configuration possible which results in a no post issue.

 

[Deleted member 185158]

I guess just stick with 2 sticks.
1:1 ratio 2-2-2-5 as fast as they'll go.

 

[delshay]

This video indicates what happens when I use only 3 sticks of ram in any of the slots. I tried it with all of the slots. Fair warning don't troll the youtube channel unless you really really want to be bored. Don't say I didn't warn ya

You didn't need to post a YOUTUBE video on this, just post the result here. Probably a good idea to delete that video unless you want to keep it.

 

[N2 Gaming]

I'm going to take a close look again at all the caps to make sure none are bulging outward on the tops.

All the caps look good from the top with just a visual inspection using my 2.0 reading glasses.

You didn't need to post a YOUTUBE video on this, just post the result here. Probably a good idea to delete that video unless you want to keep it.

And it's gone.

I guess just stick with 2 sticks.
1:1 ratio 2-2-2-5 as fast as they'll go.

That's not a solution I want to live with :/ Default timings are 2-3-2-5 by the way. I have never tried tighter timings.

 

[Bones]

There all capacitors. For example, I operated my phenom with four missing capacitors for a while before replacing them. Did not see any difference in stability when four was missing.

If you damage any of the capacitors when de-lidding, the best thing to do is remove them completely. Then find a cheap or even a non working CPU & de-soldered the caps from that & replace the missing caps on your CPU. This is what I do if I crack the capacitor package. Another thing to note if you crack the capacitor package you must check for a short before powering up the CPU as you can blow-up FET if a short is present.


OP Photo: below I have highlighted the ceramic capacitors that's linked to memory. Most or all of them should be memory.View attachment 271051

Yep - What you circled is what got nicked in mine.

I already know to get these from another chip and how to do all that, I just haven't had opportunity to do it yet.
The one that got nicked in mine affected the cache as said because before that it was all working correctly but afterwards, the L2 cache wasn't working. Checked it and saw the affected component, knew what to do from there - I just haven't done the deed yet.