Laptop troubles

Thought I'd seen it all, but this one floored me.

I regularly use a Realtek based USB adapter on my Preasario laptop (XP) to allow the use of high gain antennas.

I switch off the inbuilt IntelPro board and all is well.

Then recently when I switched the Intel back on I got the blue screen of death.

Didn't think much about it at the time, and today decided to check out the problem.

The PC workroom has been totally clutterd for quite a while so I did the job on the kitchen table.

Reloaded drivers and did all the usual stuff for half a day and just when it seemed OK, up comes the blue screen again, randomly, as soon as I turn on the Intel board.

I'm ready to concede defeat and assume that the Realtek and Intel WiFi managers are conflicting. Then the penny dropped.

The auto defrost fridge six feet from the table was going through it's cycle, as it regularly does.

I wonder........? Took the laptop into another room, and yep, you guessed it, no problem.

The damn fridge must be spewing out EMF all over the place and crashing the Intel board.

Thinking back on it, where did the problem first appear? Yep, on that same damn kitchen table.

Rob.

Reply to
me here
Loading thread data ...

Keep your fingers out of the 117VAC connections and you won't end up on the floor.

Does the Presario have a model number? Maybe some clue as to the wireless card maker and model?

Switch of how? Unplugged the card? Disabled in the BIOS setup? Disabled from the keyboard (Alt-F2)? Disabled in the Windoze control panel? Disabled using the Proset connection manager? Disabled using Windoze Wireless Zero Config? Yes, it makes a difference.

This is a clue. Check your Proset version:

I had some recent entertainment doing battle with an Intel 2200BG in an IBM Thinkpad T52. It didn't blue screen, but did just about everything else all wrong. I eventually discovered that the 2200BG card had literally the first version of the firmware, which didn't work with the latest Proset utilities.

Numbers are a good thing. Check your versions.

Presumably, you'll think more about it this time.

I know the problem. I eat lunch on my desk and play with computers on my kitchen table. The kitchen table is much bigger and somewhat less cluttered.

Part of the kitchen table is in the upper left of the photo.

Blue screens usually have an error number and module name that caused the fault. Did you catch those?

That's highly likely. Did you uninstall either or both drivers? It might also be something totally unrelated. I fought a similar problem that turned out to be a bogus USB driver (supplied by a dongle vendor). The wireless MiniPCI card shared an IRQ line with the USB device. The USB driver was hogging the IRQ line and not letting go, causing the wireless driver to complain bitterly. It was fixed with a new USB driver from the vendor.

Ummm.... I don't think so. However, moving the computah may have moved the screws, staples, and metal scrap you have floating around inside. Shake well and see what falls out.

Yeah, the beer is hard on Intel products.

Ummm.... I don't think so.

Reply to
Jeff Liebermann

Hi Jeff,

I wouldn't have believed that EMF could crash a PC but it looks that way to me.

Here's the details.

The laptop is a Compaq 2267AP.

Inbuilt IntellPro 2200BG - hardware 6.0.3 - driver 9.0.4.39

This is the current driver.

The USB adapter is Realtek RTL8187 - driver 5.1293.518.2007

The USB adapter and the Intel are both disabled via device manager.

I don't use Windows zero config.

The laptop has never been apart and there's nothing loose in it..

I've used this setup for over two years without a hitch. The problem can only be repeated in the kitchen near the fridge, and only when the fridge is going through the auto defrost cycle. It's a very large Kelvinator.

It only bluscreens at the point when the Intel is enabled/initialising.

Move out of the kitchen and it can't be replicated.

I was dumbfounded by this.

I've seen EMF in action at work where part of the IT area I worked in was directly above the main plant room, and some of the monitors were affected to the point that we got the OH&S guys in to check it out.

Of course nothing happened - as usual - lots of talk and a bit of chair shuffling. Maybe I'm irradiated :-)

That was coming through several feet of reinforced concrete from one of the main powerline conduits.

Cheers

Rob

By the way, your table is definitely more cluttered than mine.

By the way, your table is definitely more cluttered than mine.

Reply to
me here

Well, I find it unlikely the HARDWARE was effected to cause a blue screen crash, but it could well be likely the SOFTWARE can't deal with it and crashed.

berk

Reply to
berk

I think it's a software problem.

Maybe when the Intel manager kicks in it gets confused by some hash signal from the fridge it doesn't know how to handle.

It must be a pretty strong spurious signal because the fridge is NOT between the laptop and the access point, but it is close to the table and probably sending pretty omnidirectional rubbish out.

I rarely use the laptop in the kitchen and didn't pick up on this problem until recently.

Rob

Reply to
me here

Hi jeff,

I tried it with the Realtek software uninstalled. I know that these can cause problems, so that was the first thing I did.

The PC still blue screened in exactly the same fashion with the Realtek stuff removed and nothing in the USB port.

Your power adapter suggestion may WELL be a winner - I was running it off the same AC line at the time.

I recon that's a very good bet, because when I moved away from the kitchen I was running on the battery pack.

I'm not going to waste any more time on this, but I did find it peculiar and wondered if anyone else has had a similar happen.

I have seen wake on LAN triggered by spurious signals down ethernet, so I suppose anything's possible.

Lucky the beer was in the fridge when those shelves came down - could have been serious.

Rob

Reply to
me here

I know, but if the software can't process the signal, it can crash the program and take down the operating system as well.

Rob

Reply to
me here

I still don't believe it. I thought you were joking.

Any chance your laptop is sensitive to power glitches when plugged into the wall plug somewhere near the giant fridge?

I'll assume Windoze XP Home 32 bit. See any drivers worth updating?

Latest Proset version is 12.4.3.9. Latest driver version is 9.0.4.39 so that's correct.

Looks like you're current on the drivers. However, I'm not so sure about the firmware version. Somewhere in the Proset diagnostics or adapter details, it shows the firmware version. I don't have a suitable laptop with a 2200BG handy to check.

Incidentally, if you want to live dangerously, try using the lastest version 3.1 firmware intended for Linux:

RTL8187B or RTL8187L or RTL8187SE ??? I can't really tell if you have the current version as different chip vendors have different drivers. Here, it pays to experiment. Be sure to uninstall one before installing the next.

Have you ever considered just removing the unspecified maker and model USB device and trying the Intel 2200BG? Also try uninstalling the Realtek driver. That will effectively put things back to where you were before you started this exercise.

Ok. Just a minor note. Every, and I do mean every, laptop I've ever taken apart has loose hinge screws in both the base and the LCD display section. Sometimes, they fall out. I recently took apart a Dell Inspiron 1100, where I think there were 3 screws floating around loose in the LCD section.

Magnetic field induction? Dunno. If you borrow a power line EMF tester such as:

you might see something around the fridge. Sorry, but I'm still skeptical.

Look at the blue screen. Notice that it says "module" in there somewhere. What's the module or DLL that's crashing and at what memory location? That will tell which driver is complaining.

I'm still skeptical. However, I've been wrong enough times to remain open to such things.

I once installed a server on the wall opposite a big xray machine. The server would crash, hang, or go insane every time the xray machine was used. However, that was obvious as the room vibrated a bit when the xray machine was running.

That was the result of my fabulous carpentry skills. 4 shelves full of heavy junk peeled themselves off the wall. It took me about 2 weeks to clean up the mess.

Reply to
Jeff Liebermann

Well Jeff you were right.

I couldn't stand not knowing, so I tried to repeat the exercise in the kitchen with the laptop on battery pack only.

Absolutely no problem.

So it was NOT an EMF issue, it was dirty power supply off the AC line.

That's a relief.

I was getting a bit anxious as I have a new Ebay toy comming through the mail in a few days, which I'm just itching to try out.

formatting link

Not worth trying to build one at that price.

Thanks for your help.

Rob

Reply to
me here

On 28 Sep 2009 11:55:04 +1100, "me here" wrote in :

Software doesn't pick up signals. Hardware does.

Reply to
John Navas

Weird. Thanks for running the test. I don't think I could have slept tonight without knowing.

I've seen the power supply noise problem in the past. I only happened with both a dead LiIon battery and a flakey power supply. I had a customer with a Compaq v6000 laptop that did that. However, instead of blue screens, it would hang whenever a power line glitch arrived. I had both a spare battery and charger. If I replaced either one, the problem would go away.

Maybe. My guess(tm) is a dead battery and a dead charger as in the previous example.

I'm not totally convinced, but some random testing over the next few days or weeks should reduce the guesswork to certainty.

Ummm... all you need is the item number in the eBay URL:

I once built a 2.4 GHz yagi out of a 1/2" square length of oak and some aluminum roofing nails. The driven element was made from copper wire and semi-rigid coax. I'll see if I can find a photo.

Good price but a slightly odd design. The large distance between the driven element and the reflector seems wrong. If mounted outdoors, the plastic-rubber caps might fall off when hot. 16 elements and 65 cm (25.6") length seems about right for 16dBi gain. The mounting plate looks kinda flimly. There's no backing stiffener for the U-bolts, which might make the mounting plate bend.

The MFJ-1800 is also 16 elements, but only claims 15dBi gain. Close enough.

Here's another that's similar:

Note the lack of the large gap between the driven element and reflector. Only 13 elements and 60 cm length, but still claiming

16dBi gain. Probably not. I like the N connectors better than RP-SMA (unless you have a supply of RP-SMA extension cables). Also note that the specs list it as an "omnidirectional" antenna. Sigh. At least the u-bolt includes a backing stiffener.

Here's another, exactly like the one you bought, but with a u-bolt backing stiffener:

I still prefer panel/patch/planar antennas over a yagi. However, they're more expensive than these yagi antenna.

Reply to
Jeff Liebermann

Isn't 16 dBi about the maximum for 2.4 Mhz with this type of Yagi ?

.

Yes, I considered that one, but I just use the RP-SMA connectors and they work OK.

I'm not too worried about the mounting bracket as I intend to make one up to allow tripod mounting and vertical axis rotation.

I am a big fan of corner reflectors. I have also tried the salad bowl conversion, and built a biquad.

I never actually got any figures on the salad bowl (which could be a cousin of yours - by looking at it) but it did perform quite decently for what it was. But not a patch on a biquad.

From my tests with Netstumbler the corner reflector with 2 dBi rubber ducky is only marginally behind my homemade 150 x 150 mm biquad - supposedly 11 -12 dBi.

However the corner reflector works better at shielding out unwanted transmissions from behind it than the biquad.

My next door neighbour sprays quite a bit of signal at me and the corner reflector is much better at shielding out this stuff.

So I'm keen to see how this little yagi performs for $25 AUS.

The corner reflectors are so simple and look almost agricultural, so this comparatively awesome looking yagi must be much better, eh :-)

As for the laptop, it seems to be behaving itself, and as it's a bit old now I suspect there may be more to this than is evident - maybe power supply issues?

Cheers

Rob

Reply to
me here

Hi again Jeff,

Just one more thing on these cheap yagis, I notice that the one above is rated at 50 W maximum power, where as the one I'm getting is supposedly 100 W.

What do you recon and is that much of an issue?

Rob

Reply to
me here

Have a look and see if on the same phase you have another wall wart plug pack , it's common for two switchers in some circumstances to interfere

Reply to
atec 7 7

You won't be using anywhere near that much power, so no.

Reply to
alexd

On 28 Sep 2009 12:45:17 +1100, "me here" wrote in :

The software receives no "signal". The only thing that could happen is a hardware error (malfunction), which could be manifested in a variety of ways (e.g., Blue Screen Of Death ).

Reply to
John Navas

Nope. There's no upper limit on gain. However, realize that the antenna would need to be twice as long in order to obtain just 3dB more gain. That creates some rather absurdly long Yagi antennas. So, the common method is to place two antennas side by side as in:

However, note that EACH section has 24 elements on a 56cm (22") boom. That's many more elements than the eBay antennas. One might assume that this should have at least 3dB more gain than the various eBay antennas. Nope. That's because this antenna is for real, has been properly tested, characterised, and documented, while the eBay antennas are more like science fiction.

However, you are correct that 16dBi is about the limit for yagi antennas, but for a very different reason. As the gain of any antenna goes up, the useful bandwidth of the antenna goes down. Eventually, it becomes smaller than the 83.5MHz required to cover the entire

2.4GHz band without tuning. Different types of antennas become more narrow band at different rates. The Yagi is about the worst. For a less than 2:1 VSWR, a 16dBi yagi will have approx a 85MHz bandwidth. Meanwhile, a typical 24dBi dish and a 22dBi panel, will have approximately the same bandwidth. This problem, and the side lobes common to all Yagi antennas is why I prefer other flavors.

That's fine if your radio is within 1 meter of the antenna. In most cases, you can plug directly into the RP-SMA on a USB dongle or pigtail. However, if you want to mount the antenna on a roof, you'll need a coax cable extension, which should use heavier coax, such as LMR-400), which generally uses a Type N connector.

Ummm.... I think we have a problem here. Just how far away is the other end of this link? The problem is that with just the altitude provided by a tripod, the earth is well within the Fresnel Zone. See:

For example, for a 5ft high tripod, you can only go about 80 meters before the Fresnel Zone hits the ground. Any furthur, and you'll need some tripod extensions.

The salad bowl was a bad joke. You can see the pattern if you use a light at the focus and look at the reflection on the wall. Instead of a fixed point, as it should be with a real parabola, it's a cicular ring, that never comes to a point. Not exactly an ideal shape, but seems to work well enough. A fast test with a flat sheet of aluminum foil covered cardboard, yielded about the same gain. I did the same dish versus flat plate comparison with:

8.3dBi for the flat plate versus 7.7dBi for the dish.

It's really difficult to make accurate measurements without a pile of test equipment, an antenna range, or a proper RF test chamber. My best suggestion is to obtain a reference patch antenna, with about

8dBi of gain, and use it as a standard. The rubber ducky omni generates far too many reflections to be useful. I shoot across a wide valley to avoid Fresnel effects. I sometimes use Netstumbler, but with a step attenuator to measure the difference in gain from my reference antenna.
Reply to
Jeff Liebermann

I guess a better questions would be what determines the maximum power in an antenna. It's usually the point where the voltage produced breaks down some insulator or arcs over some gap. This varies with altitude. In other cases (i.e. cellular base antennas), it's the amount of heating that the antenna can tolerate over the range of ambient temperatures. With these Yagi antennas, the spec is probably a number pulled out of someone elses data sheet on the assumption that the antennas are similar. I doubt if they were ever tested for handling 1 watt, much less 50 or 100.

My guess is that the matching section will get hot enough to melt the shrink tube covering. Maybe arc over if it's a gamma match. The 1m coax cable probably will get rather warm.

Reply to
Jeff Liebermann

Well, in this case the 'signal' is code as converted from signal. The original poster has discovered that dirty A/C was causing the trouble but it is likely the trouble could have been with programming code that couldn't handle a great deal of dirty 'signal'.

berk

Reply to
berk

From the trenches: I once had a problem with noise from a refrigerator crashing a system that I was working because in the PCB, the traces from the AC/DC supply inlet was improperly routed with respect to the Ethernet (line-side) signals. If one of the SW guys hadn't had a small fridge in his office, it might shipped that way.

Steve

Reply to
Steve Fenwick

Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.