K7N2G-ILSR refuses to POST


Lord Darkness

For those few, if any, of you that are still bothering clicking this thread :-)-]]]) here's a little update.

I was rather afraid that the store I went to today was gonna say that they couldn't help since I bought this stuff thru an online retailer (although somewhat connect to the store) but as they themselves got to be rather interested in this strange case, they let me go into their testlab filled with different PSUs and memorybanks etc etc. This was a dream come true after 3 weeks of pulling my hair.

First, I wanted to prove you all wrong :-)lol_anim:) so went for the Antec480 PSU, and just as I suspected, it was a dead boot, then one that post:ed and then 4 more dead ones. OK, PSU out of the question.

Then, new memory modules into the board (Kingston) but no luck. OK, so the motherboard is dead? That's really strange considering this is the second board I try out, and the fans are spinning (on both boards). So I before leaving I put a Corsair 256MB module into the board and voila! It booted, posted and restarted without troubles 5 times. God damn that felt nice!

So.... I guess I answered my second question myself. Yes, although my 2*512 TwinMOS are listen in the compability list, as well as my freinds Kingston (and the Kingston at the store) they do not work on this board. The Kingstons works very well on 2 other Nforce2 board (my friends and one at the store) but not on the K7N2G-ILSR, and the TwinMOS worked fine on one of the Nforce board as well as a VIA based Abit board but not on the board at the store and not on my MSI board.

Nforce is in other words very picky about RAM, and the K7N2G is even more picky that your avarage Nforce2 board (something to do with the integrated graphics accessing/reserving parts of the memory?).

Now, I could of course exchange my TwinMOS into Corsair memory that I know work on this MSI board and pay the $65 dollar difference, but then again I could just change the board.



New member
Jul 7, 2002
but he was right about the ram though was he not

we are all just users here

and most of us only have the ram we use and dont have access to others

there is a great big sticky thread were we have tried to sort out what does and does not work

as to the minimum psu you will not get them to commit to that as hardware varies so much

myself i would not recommend a psu has less than 28a 35a and 16a with the nforce 2 and more is better

Lord Darkness

Yeah and if you'd read that great big sticky thread you'd seen that my memory is listed as recommended while in fact it is unusable on a K7N2G. I'm not saying a good PSU is not worth your money, in fact I know very well that good PSU is important. But a new board should at least post on a generic 350W supply, not to mention the Enermax, although 2 years old it's still good enough to power up an empty system.



Can you post or take a picture of your CPU? I had a similar problem before, and I was told that it's a CPU issue. Seems voltage incorrect or something like than. It was also a Barton...


I am presently having the same problem with my board that I installed today, I bought a 2600 xp Barton and used the memory from my old MSI k7n420 pro. Nothing, dead. HELP!!!


I find it funny that the numbers that are listed for that memory in the compatibility list were different than what you bought and you say you're an experienced builder and all but that means the modules that you bought came from a different batch of modules and probly a different date code...that means that the chips you bought WERE NOT in fact the same as what were used in the test by nVidia...

Basic to me who's been building some 20+ years now and I've seen all sorts of scenarios concerning PSU issues, memory issues, etc...your condescending attitude could have been avoided had you read MY WHOLE STICKY POST instead of stopping at the top to click just the nVidia list link that I put there...but then again I doubt you'd have believed it anyway...

Good luck in your endeavors...

Bonz :cool:


I understand the reluctance of Lord Darkness to spend >$100 on a new PSU that may or may not solve the problem. (I am currently 'sharing his pain' with my own problem) Would a good alternative be to use use a DVM to measure the voltages at the ATX power connector? (Assuming the expertise is available)

If the voltages are close to rated, then shouldn't the PSU be OK? I don't know if during POST the current draw would be greater than during a failed post.



DVM's aren't fast enough (for quick response you are better off with analog meters), and you really need to see the voltage at the motherboards regulators.