Less signal strength with antenna?

h00ktern · EVDO Fledgling Joined: 25 Jan 2007 Posts: 11

Fair enough. Let me finish browsing real quick & I'll run outside & try as you suggest. I think your intuition & supposition deserves a test. I never thought of testing this way... I'm not sure how I'd manipulate the laptop on the roof while doing 70mph, or argue a warranty claim after doing so; I'll leave an explanation to the Highway Patrol up to everyone's imagination, lol

Will post a result in a few.

BTW, compared last night's test & results near the stadium today, where I was getting U720 signal (fluctuating much more than last night) in the realm of -55 to -70 dBi mostly hovering around -60 or so dBi. 3 or 4 subsequent speed tests showed high 200 to mid 300 uploads, & 800 to 1 meg downloads.

When I plugged-in the antenna only (no amp), the hovering signal seemed to fall off, if I recall, about +3 to + 4 dBi. Again this signal loss was fairly consistent & I thought totally contradictory to what I saw last night. However, I ran some speed tests in this lower signal, antenna config, and found the speeds, without exception, were all higher than with the U720's solo act *with* better hovering signal. The speed increases were not jaw dropping, but none-the-less, observable in the numbers. I'd throw out numbers in the 70 to 150 Kb/s range. I pondered if that might be due to less noise or a more consistent signal with the antenna, even though the hovering signal was 3-4 dBi weaker. I left it plugged-in for the rest of the day.

I'm beginning to wonder if there isn't a whole lot more to the performance of these modems than signal strength only. I'm beginning to sense that Sprint is somehow manipulating things, almost in a dynamic kind of way, reacting to changes in the communication environment between them & me. I could certainly accept that I am totally wrong, but I swear, it's like there's a ghost in the machine. I hate being ignorant...

Back with the results of your test.

Also noted another

n6gn · Posted: Sat Feb 03, 2007 5:54 pm Post subject:

h00ktern · EVDO Fledgling Joined: 25 Jan 2007 Posts: 11

Put everything back as it was last night. Spent about another hour trying to come to grips with the data spreads. Rough results for conceptual arguments:

U720 internal antenna, inside the truck on the center armrest - on average -76 to -78 hovers. Usual extremes: -74 to -81.

Same as above, but external antenna pluged in: hovers -73 to -75 with extremes -70 to -78.

U720 internal antenna, on top of the roof, near the antenna: for arguments sake, about the same as the inside results with the antenna plugged-in. Maybe more erratic signal swings.

Same as above with the antenna plug-in, pretty steady -80 - Hey!, is this modem getting warm! I did this just enough times to believe it's probably not a good idea to do it. Maybe it was just the fans on the pc. The absolute -80 & rock steady convined me.

Please don't quote my numbers as logged data points, or use them mathmatically. They are processed in my brain over the hour I do these "tests". I don't write anything down, and my method involves mousing over the signal meter in the Connection Manager, over, and over, and over and over... It's friggin tedious & I'm a one-man show. What I try to do is find a seemingly repeateable median, discounting the second by second swings. I seem to notice that the signal bars are not accurate representations of the numbers I see when I'm mousing over. There is a significant lag, but if the hovers are steady for awhile (more than five seconds) the bars seem to catch up; They definitely are not real time. For that matter, Conn Mngr cannot be argued as scientific measuring, lol

Oh, btw, the antenna cable is about 10' & the adapter cable (to the modem) about 1'. When using the amp, you use another, 6' extension, from amp to modem.

Summarily: If I'm reading correctly into your post, you are right: the cable loss plus antenna gain roughly equate to moving the modem onto the roof. However, in a moving car, that's not practical, and using the antenna offsets the loss of having the modem inside the vehicle. All bets are off when you place the amp into the equation; it appears to remove the cable loss, thus yielding the +- 6 dBi gain.

I didn't get into trying to tune the set-up for maximum performance. I'm a neophyte here, and certainly incapable of producing valid scientific data. Knowing the intrinsic inaccuracies in my reporting, I really only hope to find that in general, the equipment performs as expected. Once again, I have to give the nod of approval.

My eyes beginning to fade, so I'm going to put this away for now - re-read everything & see if I missed anything, and address it later.

Safe trips,
Jonesy'

h00ktern · EVDO Fledgling Joined: 25 Jan 2007 Posts: 11

I see I missed a few of your questions...

I think the coax is 174U.

The U720 comes with a Y-dongle, used to draw extra power from a second usb port & also, as you stated, get it awafrom the laptop. I have not been using it for the "truck" tests, but to my error in reporting, I *did* use it at the "stadium" test. That car is set-up as a mobil office. Away from the car, I just slip the modem directly into the UX's only usb port.

I'm still a bit stumped reference the "usb" cable you mention. Are you suggesting using one to extend the modem's reach to the exterior if the vehicle? Velcro'ing the actual modem outside, just to bypass using the antenna? Not as a practical use, right? Just for testing? I don't have a clue of the line loss involved in usb cable, nor, truth be told, 174U.

I certainly would consider, in a mostly permanent installation, shortening the coax length to that which is needed. I'm pretty sure the elimination of extra coax would not be frowned upon. Of course, everything has its pros & cons - adaptability is lost when you have less material to work with.

Thanks for your input; It's good you have my attention before the ADD sets in. I'm pretty sure once I have a grip on the effects of each set-up, and formulate a "I'll use this config for this scenario, and this one for that..." - I'll back away from hijacking this this Topic.

I think I feel a bit obligated to add something, in payment for receiving the access to this forum & it's many posters, such as yourself.

Thanks; let me know if you have other thoughts, requests, or suggestions.

onaquest · Posted: Wed Mar 21, 2007 10:34 am Post subject:

n6gn · Posted: Wed Mar 21, 2007 11:38 am Post subject: