NP8150 Intel 6230 Ping Spikes (Not from polling)

So I seem to have an odd issue (have probably had it for a while and never did anything to notice it until now.)

Anyway I have what seems to be a 500ish ms ping spike about every 70 seconds.

This is not the old school issue of the wireless polling every 60 seconds from the vista days.

I don't have the issue on the wired side.

I have found forcing the driver into streaming mode resolves the problem but just trying to understand specifically what is doing it and wondering if any others have ran into it?

Other wireless devices on same AP seem not be impacted. I am using the 5.2Ghz side and connection is stable 300Mbps.

Some things I have done to try to resolve this is on nic property set preferred band as 5.2 and also set roaming aggressiveness to lowest.

I made sure only one wireless profile is on system and that connect to a more preferred network is NOT selected.

Did some google searches and best I could find was people with either the issue of it scanning for networks (polling) causing it every 60 seconds or that the NetworkThrottlingIndex option in registry was holding them back.

I disabled the NetworkThrottlingIndex anyway as no need for it in my case.

I went as far as downloading WLAN Optimizer which has "Disable background scan" marked off already also so that's how I am somewhat sure it is not that.

I am running latest Wifi driver 15.0.0.75 (Tested older driver 14.x and had same results).

Now my speed stays steady, stable and fast so that is why I never noticed it. I can download a 10GB file and watch the speed the entire time and it stays at my 3MBps.

I found issue by using voip on my PC and every 70 sec it studders and ping monitor playing online games (just started gaming on this bad boy, sad I know!) would spike.

Easy way to test if you have this issue is open two cmd windows.

ping -t your router IP and ping -t google.com (just to confirm both spike around same time)

After roughly 70-80 seconds it does it for me which is about 80 pings when I ctrl c to stop.


I work in IT for a living and not really seen this but tbh never really knew anyone that voip or gamed off wireless so may be why as that is what you notice it on which cased me to test with pings and find it.

Sorry for long post just wanted to try to save some questions.

 

Silly Questions:

Have you eliminated the possibility of any software causing the issue? Such as EA's Origin, Steam, or anything else that might upload user data?

Have you run a virus scan or checked for malware?

Have you logged into your router to see what traffic is on the network during the lag spikes?

Have you removed every other device on your network to ensure that the AP isn't experiencing any sort of overload?

Just making sure we're narrowing it down to your 6230.

Can you pull the NIC out of another laptop and try that?

 

It's indeed the 6230 (or at least the driver or some win7 thing).

Not AP as many other devices can ping nonstop fine.

Install is clean, I can see the problem just doing pings and they go away if I force the driver into streaming mode which should not need to be done to resolve this.

Just wondering if anyone else with an intel card can test via ping like I posted above and see outcome.

 

I'll run it on a couple machines when I get to work on Monday.

 

I just checked this against a few Intel Wireless cards I had around. I can see a spike on the 6230 of ~150-200ms every minute and it's also consistent on the 6300 in a completely different laptop model.

I'd imagine it's something built-in to the Intel hardware that's causing the issue as one of the machines is a linux box and is demonstrating the same thing.

 

Well that is good to hear and sucks at the same time.

Glad is not just me, sucks is linux also which tells me its a driver issue or hardware and I was in hopes was just some Windows 7 thing...

Ugh, wonder if I can get Sager to let me upgrade and just pay difference as this is useless to me for a gaming rig and even more as work now that I am using VOIP and VOIP is very latency dependent.

 

Are you noticing it much with VoIP? I've used the machines running SIP softphones and an Asterisk callserver and probably would have never noticed the spiking at all without this post. I've not seen any perceivable difference in call quality with the tests I've done with it. Latency is important, but a fraction of a second jump in lag shouldn't cause too much of an issue.

 

getting the same behavior in mine, and once it spikes to 80ms i get a time out on google
using intel 6230

however if i disable bluetooth i'm getting straight 2-4ms on my router even after 2 mins

 

I have the bluetooth disabled when it happens and as Anthony@MALIBAL stated a 6300 is doing it also and both also do in linux. I did test a 6300 also and confirmed it

My bet is it's a driver thing probably but no real way to get anyone at intel to acknowledge/listen or do anything about it myself.

I also have it happen on 2.4 & 5.2GHz bands.

Sometimes the spike is 60-100 and more often its 500-300ms spikes.

 

I might have some solutions up my sleeve. Here we go mate

You need to go to safemode, uninstall the driver then reboot and go to safemode with networking, after you are in safemode with networking you can go online on Intel's website.. now you haven't stated which OS you are using, but I'll guess it's Windows 7, therefore I'll support you with these links:

(64-bit) http://downloadcenter.intel.com/Det...64-bit)*&DownloadType= Software Applications

(32-bit) http://downloadcenter.intel.com/Det...32-bit)*&DownloadType= Software Applications

Now say the drivers ain't the issue (which as you said they are, but lets check it through).

Can you open your cmd (Windows Key + R, then type cmd and hit enter)and type ping 192.xxx.x.xx -t or (whatever your router IP is, you can do this by typing ipconfig in your cmd) then try to download (or upload for that matter) something and look at the pings and the latency in ms.. if you notice alot of latency above 50-300ms (as you yourself stated) and requests time out (quickly, or just do time out) then it's probably the drivers issue, and yeah step 1 is my only solution

EDIT: You don't necessarily have to use my driver links, you can also use your retailers/re-sellers recommended links (approved drivers)

 

I currently don't have a 6230 in any machines here, but I checked the N-130, and I couldn't duplicate this.

Correct me if I'm wrong, but the N-130 is using the same driver as the 6230 isn't it?

 

An 60-80ms ping jump you MIGHT not notice but in my case half the time its 200-600ms jumps and it does it for about 2-3 pings in a row and in windows each ping is 1 second so we are not talking a fraction of a second but a total of 2-3 seconds roughly.

If I am just doing VoIP alone I don't notice it as bad but who just does one thing on a machine? It just may be my random pings are a tad higher and thus more noticable. This is local latency ontop of ISP latency so you may be closer to the VoIP provider or say have a ping in the teens vs say mine being in the 40ms range for my ISP and as a result the issue shows its ugly head more because its local latency + provider.

Regardless this should not be normal to happen


As for jaug1337, driver version has played no role thus far in the 4 different ones I have tried. I am even on later WHQL ones than the Intel site which Intel plans to post in the next few weeks and those are also doing it.

14.1, 14.5, 15.0 etc all have the issue but I cant tell if just all the drivers are having the issue or something hardware causing it specific with these intel cards.

 

I don't think this can be driver related. Considering I've been able to duplicate the problem in Linux; that suggests that it's a hardware or firmware issue.

 

Might be that you guys are sharing the exact same router or same router brand.

I would suggest trying doing several tests at work, school etc. to get better results. This sounds very odd..

 

I am sure is not the router as other devices are fine and I also connected to two different routers (so three in total) and have the same problem.

That tells me it is the device.

I say it could be the driver even under linux because is possible that whatever part of the code is causing is present in linux but who knows, it could very well be hardware also.

Kinda a bummer but at same time I think most don't notice it because it really depends what you are doing on if you will notice it.

I don't see any actual speed decrease when it happens, it is just a latency issue which short of working with things that are dependent on that being good most won't notice but either way I see it as an issue/flaw.

 

I've been having the same issues recently. When I play league of legends, there's a lag spike every minute or to from 80ms to 300/400ms.

 

Have you tried running DPC Latency with and without the device?

 

Bumping this up, has anyone found a solution for this? Ping is 400ms+ with local servers, and 10ms with other machines in the same area.

 

get a bigfoot 1102
AnandTech - Bigfoot

 

While I would love to do that, I do currently have it installed in my NP8150, hence the revival of this thread.

I do believe it would be a bit more troublesome to remove the adapter as it is in a sensitive location. What did the OP mean by "force the driver to streaming"?

 

Being I am the OP figure would give some updates. This might be more of a Windows vs Driver issue though which I will explain further down.

I have worked with a few Intel Engineers and they have told me:

"I spoke with some of our driver developers today, and your prediction is accurate. Work on the scanning algorithms is happening while we speak. It’s a very difficult task. Different users have different requirements. We’ve got users in hospitals that mount their computers on carts and run down the halls, moving from patient room to patient room. These users want seamless fast roaming, and that requires accurate scanning. Then we have users that might sit in one location all day and be associated to the same AP. This requires minimal scanning. You could argue that perhaps configuration parameters be added to help define the environment, but one thing most users seem to want is minimal interaction in the configuration of WiFi. For the most part, they just want to turn on the computer, connect, and move on."


So straight from the horses mouth it is a known issue. This is from an email back in May and to this date is still an issue.

They did provide me the following which fixed the issue but has it's own downfall:

"There’s a parameter in the registry that will prevent the adapter from scanning while associated to an AP. It’s named ScanWhenAssociated. Default value is 1. Set it to 0 to prevent scanning. The negative outcome for this action is that roaming may be affected. When it comes time to roam, the WiFi adapter might not have an up to date scan list. To find the parameter in the registry you can do a search under CurrentControlSet."

After that change you will probably need to reboot.

Anyway so the options on Intel at the moment are the above or use the program I stated in OP "WLAN Optimizer" and selecting the stream option. (Disabling background scanning doesn't seem to do the trick, which is odd because based on the above regedit change that is background scanning...)

Anyway so to move onto the why I think it's also Windows and the Drivers together causing...

I recently got an Killer 1202 N card for this laptop.

Removed all old Intel drivers, installed card and installed latest drivers.
Guess what? SAME THING haha.
What I found on this driver however is the driver DOES have a setting to set background scanning to OFF. Doing so FIXED my problem and best of all if I click wireless it will force a scan so I can see networks, unlike the Intel registry change if you are connected to a network and have that regedit set, NO SCAN will happen, not even manual unless you disconnect from the AP.

People have found force installing XP drivers seems to fix this also (seems more likely XP just didn't support this exact method of however vendors are using).

On a final note, this is not router dependent, is all client side.

I beta test hardware for various router manufactures and have been on an 802.11ac router since before they became public (multiple vendor ones also) and them as well as my original 802.11n router all did the same thing.

So this all being said, it is basically a type of background scanning but has been modified somewhat. What I find odd is how the Intel vs Killer (Atheros) handle background scanning differently.

So my solution for now is I have left the Killer card in. It has BT4.0 so is newer vs Intel anyway. Only downside is the WLAN light doesn't work when using killer cards ha.

Anyway, good luck all, feel free to post any new sharing yourself also and again remember this is only in relation to a ping spike every 60-90 seconds, not a constant high ping... If you just have high ping it's something else.