Resolved Mem tester

What is the favorite RAM tester everybody uses? I still have a system freeze issue that started a month ago. In my experience, system freezes are caused by RAM not seated properly or dirty contacts (Thanks Neil !)

Two weeks ago, I pulled the RAM, cleaned the contacts with contact cleaner and reseated them... Pulled the video card, did the same thing. The Vid card is a Nvidia Geforce GTX 550 TI made by EVGA. The system ran weill for a couple of weeks and this morning the freezIng came back with a vengeance. I did a GPU stress test on the vid card with a utility supplied by EVGA. I did a SeaTools quick generic scan on the HD.. Pulled the data plug off the HD and reseated...Inspected the Motherboard and I don't see any blown capacitors, and I know what they look like (even ones that are starting)

I saw that Memtest is dated 2013.. Wondering if this the correct program to run?

Regards

 

With so many with very similar names, it is easy to get them mixed up. I recommend MemTest86. Unlike some of the others, it is under constant development to stay on top of the latest RAM types. It's latest version dated 13 Dec 2016, so it is just 2 months old. Allow the diagnostics to run for several passes or even overnight. You should have no reported errors – not even one.

Note also Windows own "Memory Diagnostic" program is pretty good too.

 

Yes, I used your utility for two hours and the RAM passed. This will be a thread on the issues on this machine over the next couple of weeks with the admin's approval.. After running memtest for a couple of hours, I escaped out and let the machine boot back into Windows. The temp readings were 122 degree F for the CPU (this is what Memtest did), (I watch and write down the baseline) So I know that Memtest stressed this machine for the past couple of hours and with the high temps, the machine did not freeze and Temps dropped dramatically after I terminated the Memetest program and went back into Windows (94 fahrenheit). I can rule out temperatures. With Memtest running for two hours, and the motherboard did not freeze, we can rule out RAM. I'm thinking, at this point., maybe the HD or Power Supply voltages (although PS voltages tend to cause a reboot or shutdown, in my opinion)

So I am now in Windows and the machine is running well, and I am on the Udemy website studying Python. Temps are good, and voltages are somewhat consistent.. I use the ASUS Pcprobe that came with the Mobo to monitor. Past experiences over the past month say when the mouse pointer starts spinning, its' gonna freeze in about a minute or two.

Regards

 

I read right past that before or I would have commented - I agree with Arie on this. Many many years ago, this was true, but with the way RAM mounts today, it no longer is IF inserted properly in the first place. And in reality, more damage is done by users reseating RAM unnecessarily.

If the RAM is inserted properly in the first place, it "clicks" into place and this is easy to see - especially since the release levers then align properly. And once the RAM is inserted properly, the mechanical connection between the socket contacts and stick contacts is so tight, it does not allow for dirt or grime to get in between the contacts. It is important to note that RAM contacts are gold, NOT because gold is a better conductor (it is actually 3rd behind silver and copper) but because gold does not corrode. So if the contacts were clean when the RAM was inserted, dirt will not be a problem.

Now if the RAM has been sitting exposed to the elements on a shelf for a long period of time, and if the user negligently handles the RAM with his/her greasy fingers on the electrical contacts (and was lucky ESD did not destroy a memory IC), then dirty contacts might be a problem. But even then, proper insertion should clean them off.

No software based memory tester is conclusive. No errors after just 2 hours is reassuring, but still not conclusive. To conclusively test your RAM, you need to use sophisticated and very expensive test equipment, like this $4,400 RAMCheck LX Memory Tester. Short of that, you either need to run with one stick for awhile, or swap in all new RAM.

Please note the computer industry uses Celsius even where Fahrenheit is normally used for weather and room temps. You will note that Intel, AMD, NVIDIA all use Celsius in their published specs. So when seeking help, please use Celsius to avoid confusion and to ensure everyone is on the same page. 122°F is 50°C and that is fine. So is 34.4°C (94°F).

And memory testers are not stress testers! So running Memtest86 for a couple hours (or days) does not rule out heat either. And the CPU is not the only heat sensitive device. An overheated chipset or GPU can cause system freezes, as can over heated other devices. So you might consider blasting a desk fan into the open case.

Some times. But it often means that application is frozen. Or that web page is not loading properly. It may have nothing to do with hardware.

I always want to ensure I am supplying good, clean stable power when troubleshooting potential hardware issues because everything inside relies on it. So I would swap in a known good supply and run with that for awhile to see what happens.

 

OK,. for being late, I've been dealing with a narly head cold this past. week. I'll say one thing about the RAM test, it must stress the system because the temps are high. I got the contact cleaning from Old Timer. He stated gold contacts might develop a film that prevents a good connection, and this helped an issue a long time ago. I agree that stressing the system is a better test, and it looks like running youtube videos in HD is the way--CPU and Mobo temps go way up on this system. The last time it started freezing, it did that mouse pointer spinning wheel thing and that was it. It froze.. I did wait five minutes and observed it and a couple times it did start spinning again for a few seconds and then stopped--but the system never resumed. I ran it all day today and no issues but I did not use the machine. I watched the voltages on the ASUS utilty Pcprobe, and the 5 volt line runs at 4.97 most of the time. The others are normal.

Probably stress it tomorrow and maybe uninstall reinstall Chrome and Firefox

 

Freezing can be hard to establish, pinpoint. Last year I had the perfect example also with a ASUS Motherboard. At the beginning it would only freeze on the odd occasion and mainly only when the computer was only idling not being used this slowly increased in frequency over several weeks I went through all the processes that you are going through, applications, windows HD cleaning changing ram modules, replacing PSU all to no avail, no obvious damage to the Motherboard, but in the end it was a faulty Motherboard 18 month old.

 

How high? What are the temps?

And note it does not stress the system any more than taking a brisk walk requires more energy than you sitting in your Lazy-boy and reading a book. Programs that "stress" or really "work" the hardware force the hardware to run hard at maximum capacity/full power (and beyond!). Memory testers simply plug values into each memory location, then read that location to see if the data remains intact and unchanged. That is hardly stressful "work".

Gold does not corrode. That is precisely why it is used instead of copper or silver even though copper and silver are both superior electrical conductors. Since gold does not corrode, it cannot "develop a film".

A film can be "deposited" on the contacts, however, if the RAM is left out and exposed to dust, cigarette smoke, kitchen grease, etc. Also, skin oils from mishandling and touching the contacts instead of holding the RAM by the board edges can leave a film and collect dust too. But it is important to understand that once the clean RAM is properly inserted into the RAM slots of the motherboard, those slot connectors are designed to prevent any dust from getting between the RAM contacts and the slot contacts maintaining a good and proper mechanical connection - essential for a good and proper electrical connection. So if the RAM was clean then properly inserted, it will remain clean virtually forever, and not need reseating.

In the good old days, the RAM slots were ZIF type (zero insertion force) connectors and grimy dust could, over time, work its way in. But they did away with those type RAM slots years ago for that very reason.

4.87VDC is comfortably within the ±5% tolerance allowed (4.75 to 5.25VDC). Note your probe (or any software based HW monitoring program) does not test for or measure ripple or other anomalies that affect computer stability. This is done by a qualified technician using an oscilloscope or power supply analyzer - sophisticated (and expensive) electronic test equipment requiring special training to operate, and a basic knowledge of electronics theory to understand the results. Therefore, conclusively testing a power supply is done in properly equipped electronics repair facilities. That leaves the only option for "normal" users swapping in a known good supply to see what happens, as I suggested above and would urge you to do before buying any replacement parts, especially a new motherboard.

 

Nothing new to report here. Ran some HD vids, and have left the machine for several hours.. No freeze.. I did uninstall both Chrome and Firefox, because using one or both of those browsers was when the freeze would happen.

Arie states that memtest does not stress the system, but in my view it does in the sense that the mobo and cpu temps go WAYYYYYYYY up--approximately 122 fahrenheit (after stopping memtest and booting back into windows, pcprobe reports those temps..,*but the temp was obviously higher as the system cooled down a few degrees after terminating memtest and booting back into windows.)

As I said earlier, pcprobe sez the 5 volt line is at 4.97. Once in a while it does go up to 5.02, but 90% of the time it stays at 4.97..

Anyway, let's see what happens

 

Once again, please use Celsius when citing temps. The entire computer industry, even in the US and other places where Fahrenheit is used for weather and environment measuring/reporting, everybody uses Celsius when monitoring and reporting CPU temps. AMD, Intel, and NVIDIA all use Celsius when citing temperature specs for their products. So do the RAM makers and drive makers too. You can too. It is not hard. You can easily set your monitoring program to C.

122°F tells us nothing. As I told you above, 50°C is fine. In fact, it is just warm.

So please, stop refusing to use Celsius and stop refusing to accept that memory testers don't stress the system. Systems are not running at idle when testing memory, but they are not being stressed either.

Idle, testing memory, and stressing the system is like sitting in your Lazy-boy, walking the dog, and running for life.

 

can't get it to freeze anymore.. Maybe next week!