RIP Webby

I've spent a lot of time talking about how Wi-Fi works and reviewing various products aimed at fixing Wi-Fi dead spots. But today, in our increasingly need-to-always-connect world, what do you do if you've got an LTE (cellular service) dead spot?

Searching Amazon for cellular signal boosters feels like navigating a minefield. Prices range anywhere from $180 to $1,000 and up, and reviews tend to be all over the place for every device offered. Serendipitously, the owner of www.repeaterstore.com reached out to me while I was dithering, and the site offered to let me test a few of the signal boosters they carry. I described two problem areas to him as fertile grounds for testing—my partial basement floor, and my parents' rural house—and we picked out three products to test.

Two of those products—Cel-Fi Go X ($899), and SureCall Fusion4Home ($499)—are products with outdoor Yagi antennas that should be roof or pole mounted and aimed directly at a nearby cell tower. From the outdoor Yagi, a leg of coax cabling needs to be routed indoors and fed to the signal booster, which then has an output port which connects to an indoor panel antenna via another coax run. Since the LTE signal is pretty poor everywhere in my parents' house, kits like these were their best bet.

My own situation was a little different. The LTE signal is good-to-great in the top floor of my house—it's just that pesky partial basement where things get dicey. So I might be able to get away without running any wires at all. The Cel-Fi Pro for AT&T ($695) is an all-indoor kit, with no antennas to hang and no cables to run. Just power up two 5GHz-linked electronic boxes, set them up via a smartphone app, and go.

So while it wouldn't be cheap and would also require a bit of setup, could cell signal boosters help in either situation?
Boosting weak rural LTE

About a year and a half ago, my parents bought a house only 20 miles away, so they could see more of us and their grandkids. But they didn't learn that no ISP was willing to deliver service to the house until after they'd already bought it. I ended up buying an LTE modem and putting an AT&T SIM card in it. The all-LTE Internet solution I came up with worked, and Mom insisted it was fine—but it seemed pretty sketchy to me. I really wanted to find some way to improve it.

My parents are both ham radio operators and my stepdad Scott likes to run extremely long distance 10-meter equipment, so he's well used to coax and antennas. Given his RF experience, I just dumped a giant cardboard box with two LTE booster kits, four antennas, and a boatload of coax on him, and he did all of his own installation. The first product he tried was the more expensive of the two, Cel-Fi Go X. Sina from www.repeaterstore.com felt pretty confident of both the indoor/outdoor kits he sent, but the Cel-Fi Go X offered a maximum 100 dB gain to the SureCall Fusion4Home's maximum of 72 dB. And having no experience with these things yet, I just wanted to see something work.

    Action shot of my parents' house Internet: a Netgear LTE modem feeding an Orbi Pro router. (Not pictured: two Orbi Pro satellites, elsewhere in the house.) Jim Salter
    Pictured here are two LTE booster kits, Cel-Fi Go X (left) and SureCall Fusion4Home (right). Each kit consists of an outdoor Yagi antenna, an indoor signal booster, and an indoor panel antenna, with everything connected together via coax wiring. Jim Salter
    This overhead shot of Cel-Fi Go X (left) and Fusion4Home (right) shows off the heat sink on Cel-Fi Go's underbody, and the backs of the indoor panel antennas, which hang from wall or window mountable clips. Jim Salter
    Getting an initial, unboosted baseline result for the Internet connection at my parents' house was a pain—tests frequently failed entirely during the upload portion, as shown here. Jim Salter

Cel-Fi Go X came with a small, simple bracket to mount its exterior yagi antenna to the roofline, and for our initial testing, Scott used it. Later, he re-mounted both it and the SureCall Fusion4Home's yagi to a home-made aerial pole, which allowed him simpler aiming and better elevation. There were three cell towers in range of the house, which he found using the free Network Signal Info app on an Android phone. We didn't see much difference between any of the three nearby towers, and aiming the yagi at any of the three worked well.

    The first kit Scott and I tested was the Cel-Fi Pro X. This is the exterior antenna for Cel-Fi Pro X, a highly directional yagi design. It's shown here mounted to the roof bracket that came with the kit, but Scott later remounted it on a homemade pole aerial. Jim Salter
    My stepdad Scott did all the antenna mounting, coax routing, and equipment setup for these reviews himself. Jim Salter
    When Scott hooked up the yagi for Cel-Fi Go X, he initially used the now-empty roof-mount bracket seen on the right, which came with the kit. When he hooked up the SureCall Fusion4Home later, he decided to mount both yagis on a home-made aerial pole. Jim Salter
    This is the base of Scott's home-made pole aerial. Home Depot and Lowe's sell these as weighted bases for patio table umbrellas. (Note: any such aerial pole should be properly grounded, in case it attracts lightning.) Jim Salter
    Both kits used an interior panel antenna to rebroadcast the donor signal from the exterior Yagi. The panel antenna radiates mostly from the front in a broad pattern.

After connecting Cel-Fi Go X to its antennas and registering it with the Cel-Fi Android app, the LTE modem's signal strength indicator shot from one bar to five. Any AT&T phone brought into the house also showed the coveted "full bars." But did anything actually work better? To find out, we did before-and-after speed tests using the excellent DSLReports Speedtest. Our method was to do a few speedtest runs on my Pixel 2XL phone, unplug the booster kit and do a few more, then reconnect the booster and try again. We got consistently and significantly improved results for each kit.

Since these are true Internet speed tests, we unfortunately had no way to get consistent results from day to day. AT&T shapes download traffic over their LTE networks very heavily, according to both how much data a given SIM card has consumed in a month and the current needs of other AT&T subscribers connected to the same tower. So it's not useful to draw conclusions based directly on which kit gets the highest numbers—particularly since the kits weren't tested on the same day. But we can draw meaningful conclusions from the before-and-after on each kit.

    The important thing to spot here is that MASSIVE improvement in upload bandwidth, for both kits. The tower can transmit to my phone just fine, but my phone had trouble transmitting to the tower without help from a booster. Jim Salter
    Getting an initial, unboosted baseline result for the Internet connection at my parents' house was a pain—tests frequently failed entirely during the upload portion, as shown here. Jim Salter

Browsing the Internet at my parents' house without the booster kit tends to be a frustrating, balky experience—you tap a link, and it takes forever before the new page begins to render. Once it begins, it usually finishes pretty quickly, but that initial pause will drive you nuts. Sometimes the long pause turns into a hard error and you have to hit reload. Both booster kits fixed that, making link and button taps responsive again. That's because when unboosted, my Pixel 2XL struggled to reach AT&T's cell tower, even though the tower had no trouble reaching it.

This is borne out in our speed tests. Even though the unboosted download speeds are great—better than I usually get upstairs in my own house, with plenty of signal—that weak upload means a lot of errors and retries when you're trying to request a page in the first place. In the real world, the greatly improved upload with both kits translated into responsive page loads and a better experience with the house's Roku TV streaming device.

It's more difficult to directly quantify the LTE boosters' effect on voice service. Without the boosters, calls would rarely drop—but the audio quality would frequently be poor for short stretches of a call. Calling from my house, I would sound fine to them, but they would frequently go through a few moments of sounding tinny or garbled on my end. This tracks well with the observations about upload and download we made on the data side, and these intermittent drops in audio quality seemed to stop occurring with either booster live.

The clearest indicator of the kits' success was my mom herself. She had always reassured me that the LTE modem I set them up with was fine—but after a few weeks of boosted signal in the house, she made it very clear she didn't want to do without one anymore.