Laptop runs solid, tried to run MemTest86 8.1 and have Test 6 errors! False positive?

[PIRATA!]

Hi there!
I have a problem with my laptop ram.
Computer runs solid, and for fun I tried to launch MemTest86 last version UEFI in the Free edition, that is 8.1, and had errors only on Test 6: 20 errors at first pass, 150 at second.
Memory location always differed from previous pass, and have tried both all CPUs, and after each CPU separately.
If using only one CPU at a time, I still have errors, but not so many: 2 errors at first and second pass, and 12 or 20 at third pass.
Tried also the old ram mounted on the laptop for almost 1 year in the past, and the errors are confirmed: always Test 6.
Quite similar behavior.
Laptop is updated to latest bios available, and there is no external drive connected. Only usb mouse and usb pen-drive (the MemTest86 one).
Could it be a false positive? Computer runs solid!
Thank you all.

[David (PassMark)]

Can you post the test report. At least the bit showing the address and the expected and read values ( to determine the number of bits in error)

What's the make a model of the laptop?

[PIRATA!]

Can you post the test report. At least the bit showing the address and the expected and read values ( to determine the number of bits in error)

What's the make a model of the laptop?

Ok, sorry for the delay but I had to remake the tests as I formatted the usb pen drive where I had the logs saved.

Here are 3 logs:

1. first is the one made with the same ram as the one mentioned above, but only had 60 error now, and I stopped after 2 passes.
2. second is done with my old ram and I had no errors for 2 passes, so I stopped.
3. third is interesting as I have mounted a different stick but equal to the ram as the first log (the one that gave me so much errors at beginning), and now I had no errors on pass 1, only two errors on pass 2, and no errors on pass 3 and 4 (I went all through reaching test 6 in pass 4 just to see if it was going well, and then quit without going all through the finish of pass 4).

The laptop is a Lenovo Yoga 500-14IBD.

I'd like to know if its a ram problem, or a compatibility problem, and if I can live with that 2 errors in the second ram stick I've mounted now (the one from third test).

Thank you.

[David (PassMark)]

Looks like a real RAM error to me.
Maybe the error is only triggered on certain timings, or depends on external conditions (e.g. temperature)

False positives tend to have errors that are more than 1 or 2 bits in error and tend to be in a tightly defined address range (especially in the low address ranges where is there is more devices mapped into the memory addresses).

[PIRATA!]

Looks like a real RAM error to me.
Maybe the error is only triggered on certain timings, or depends on external conditions (e.g. temperature)

False positives tend to have errors that are more than 1 or 2 bits in error and tend to be in a tightly defined address range (especially in the low address ranges where is there is more devices mapped into the memory addresses).

But why do I have an error only on the second pass, and not the third or the forth when the laptop was much more hotter than the second pass??

Regarding the laptop, it is certified to supporto only max 8Gb or ram, and I'm having errors with both of the 16Gb of rams I tried, one like hell, and the second only a couples of errors on pass n.4.
Could that might be the cause?
I have seen that these kind of modules built like this should work on this laptop, but I have not had any feedback from anyone about going through a MemTest86 testing.
I've tested also the second stick of 16Gb (the one with only 2 errors) with other software, and only MemTest86 gives errors back.

I'll try to make the test again with a much more ventilated environment so to see if it is temperature.

In your humble opinion....could I live with that, or will I have for sure problems of any sort? What could consequence could I might have in the future? What is the most common senario I could experience?

Thank you.

[David (PassMark)]

1st pass is a quick pass. Subsequent passes are longer. Some tests also use random data that is different each pass. Also not all faults are consistent. Sometimes a bad memory cell might only fail intermittently.

Maybe the laptop can't keep the voltage level high enough to drive more than 8GB. I don't know. It is a question for the manufacturer.

The consequences of bad RAM are effectively random and unpredictable. You might never notice a few errors, or it might crash every 5min.

[PIRATA!]

1st pass is a quick pass. Subsequent passes are longer. Some tests also use random data that is different each pass. Also not all faults are consistent. Sometimes a bad memory cell might only fail intermittently.

Maybe the laptop can't keep the voltage level high enough to drive more than 8GB. I don't know. It is a question for the manufacturer.

The consequences of bad RAM are effectively random and unpredictable. You might never notice a few errors, or it might crash every 5min.

Well, I could try other rams, order a new pair (these were sold in pairs...is it ok to use them alone?) and confront them, and them send back the faulty, or the less functional.

The computer works fine, and I have tested in gaming as well: not a problem till now.

I'm testing the second 16Gb (the one giving back less errors) in MemTest86 only on Test 6 now, so to reduce time for the testing, and see if the errors are the same. For the rams, as I've seen that on all other tests they are ok, is it ok testing only the test that fails, or running it with the other tests is important for I don't know the stability, the stress, the voltage, the temperature?

Thinking about it, as I'm running test 6 only from lets say 10 minutes that the pc is on since a complete night turned off, the temps should be low, but regarding the voltages, what would you say? Could it have better voltages?

Regarding the voltages, in MemTest86, when I run the SPD to see the specs of the ram, I see that the voltages indicated are 1,5 - 1,35 V. In CPU-Z from Windows I see that only 1,35. Could it me that these rams are set by SPD to need more than 1.35V and that really my motherboard can't give more? Have you ever seen an SPD like this from MemTest86?

I'm running a test now. When finish I'll upload the results and the screenshot of the SPD from MemTest86.

[PIRATA!]

Ok, test ended. Test 6 only 4 passes: all good! What does it mean then??

In attach the test result and also the screenshot of the SPD.

Thank you.


EDIT: Hah!! I just noticed that I've done the Test 6 on battery...could it matter?

[PIRATA!]

1.

Well, there is been some evolution here!!

I have done tre full complete test and left it to go for the full 4 passes...and this time I was all success!!

The only thing that I have changed is the plug in the wall where it was attached my laptop!!

It all started when I realized that I had a successful pass-through in test 6 for 4 passes on my battery...and then I remembered that my laptop, as it is made of metal chassis, suffers from bad ground plugging.
I don't know if it is only mine, or if it is a general thing with metal laptops...but I have realized this as I was feeling a buzz under my fingerprints while working at my desk at home.

What I mean is that, if I attach the wall power cord to a wall-plug that does not have the ground, while touching the metal around the keyboard of le laptop with the fingertips, a very slightly buzz can be felt underneath them. If I plug the paltop to a grounded wall-plug, everything is ok!
I even use my laptop to check grounded wall-plugs now!!

So I changed wall-plug into one where I don't have such behavior, and which does to ground....and now I have finished the test successfully!!

I don't know if it is that, but for sure the power going around the laptop it is not a good thing.
Maybe my voltage to my ram is at its limits like you suggest, as with the guaranteed 8Gb, MemTest86 goes through even with this electric buzz going around the laptop....but if using only grounded wall-plugs should be my only thing so to have stability....well it is even not normal to have not grounded wall-plugs...so yeah!!

Said so.....could it really be this un-grounded thing here..or is it only casual that now I have a complete 4 passes successful?

In attach the results.

Thank you.

[David (PassMark)]

I thought the Lenovo Yoga had an external power brick. Meaning only low voltage DC gets feed to the laptop. Not mains power.
Also, at least for our Lenovo Yoga, the power brick only had a two pin mains plug. So grounding was never possible. But maybe it is different for different models.

[PIRATA!]

I thought the Lenovo Yoga had an external power brick. Meaning only low voltage DC gets feed to the laptop. Not mains power.
Also, at least for our Lenovo Yoga, the power brick only had a two pin mains plug. So grounding was never possible. But maybe it is different for different models.

I have an external power supply...the standard stock one, never changed it. It has a 3 pin plug from the side of the wall plug cable, and the power plug that goes from it to the laptop it has a 3 pin as well. I have the rectangular plug with the metal externally and internally, plus the little pin at the center.
I've checked on Google and physically my plug has 3 contacts, so I imagine I have some sort of ground connected to the laptop.

Maybe it's just a power supply failure, as I have such thing only if the power supply is connected to the wall power, or maybe is some internal circuit regulator in the laptop.
I imagine that all these laptops are done for all markets, and that IBM al least provides different power supplies regarding the country it is sold.

I'm from Italy, and we have the ground. But for example TVs are not sold here with a ground power plug, but they don't give back any buzzing...maybe because they are made with plastic chassis, or maybe because they have an internal regulator for preventing such buzzing...I don't know, just thinking....

The fact is that I have not made a second try of MemTest86 as now I need for my job, but will do it again in the weekend.


But regarding my situation...what are your personal thoughts and concerns for now? Have you ever experiences such thing? Should I stay cool with my ram or what?

Thank you.

[David (PassMark)]

The electrical insulation & grounding of laptops is somewhat outside of our scope of expertise.

[PIRATA!]

The electrical insulation & grounding of laptops is somewhat outside of our scope of expertise.

Well, I understand of course Lets keep out the electrical issues.

I've done the full test with 4 passes and it went all good again. I imagine I'm ok.

Does it matter to make tests at first boot after a night resting, instead of make it after a day of continuous working??

Thanks!!

[David (PassMark)]

One can never be 100% sure, but if there was 4 passes without error, there is a reasonable chance it is good (for example there is a small amount of RAM that can't be tested due to memory mapped IO).

Shouldn't matter when the test is done unless the machine is susceptible to some outside influence (Temperature, EMI, etc..)

[PIRATA!]

One can never be 100% sure, but if there was 4 passes without error, there is a reasonable chance it is good (for example there is a small amount of RAM that can't be tested due to memory mapped IO).

Shouldn't matter when the test is done unless the machine is susceptible to some outside influence (Temperature, EMI, etc..)

Thank you.
and for being sure, I'l try to do it after a full day of working onto the laptop.

About this ram stick, I'll keep it then!

Thank you for everything...and I'll post the results here after a full day of working...when I'll do it

Thanks!