Print Page
Testing TV Antennas With Nifty Gadgets

Doug Lung

This month, I’ll describe a nifty pocket-sized USB spectrum analyzer and show its pros and cons through some measurements on an ATSC signal, out after the mask filter and with some popular TV antennas.

A few months ago in RF Report, I described a USB spectrum analyzer the size of a small USB TV tuner stick. The device isn’t much larger than a USB flash drive and looks similar to a USB TV tuner stick except for its SMA connector. I decided to check it out.

The device has two separate bands—1 MHz to 850 MHz and 850 MHz to 5.35 GHz. Frequency spans are fixed and range from 1 MHz to 1,000 MHz. Fig. 1 shows the only data available besides what’s seen on frequency/amplitude display graticule.

Unfortunately, there is no option to measure channel power, which makes accurate DTV signal-level measurements difficult. This appears to be a software, not a hardware, limitation, so perhaps a future version will support it. Sweep is normally 2 seconds and burst mode allows it to capture short pulses to accurately display spread spectrum and frequency hopping signals such as Bluetooth and Wi-Fi.

Fig. 1: Triarchy measurement
(Click to Enlarge)
The input level range is –110 dBm to +30 dBm using the provided SMA attenuator. The device has a maximum usable display range of about 80 dB. Its noise floor is –115 dBm with a 5 MHz span at a –60 dB reference level at 1 GHz. Wider spans or different reference level settings will increase the noise floor and limit the usable dynamic range.

I hooked the TSA5G35 up to the output of a Channel 30 mask filter to see if the device might be useful in checking shoulder level. Fig. 2 shows the display from the TSA5G35 software. You can see the signal coupled into the Channel 30 antenna from a Channel 31 transmitter on the same tower. This was a clean transmitter—FCC mask shoulder levels were around 50 dB down—so only a hint of the shoulder is visible.

While it can provide a quick check of transmitter performance, this miniature spectrum analyzer is more useful for seeing what signals are present and their relative amplitude. The wide span and max hold make it ideal for tracking down interference. I used it to compare three popular antennas, the Winegard FreeVision FV-HD30 and Mohu’s Leaf Plus and Leaf Ultimate.

Fig. 2: 8VSB spectrum display
(Click to Enlarge)  
The FV-HD30 can be used indoors or outdoors and unlike many indoor antennas offers excellent performance on high-VHF channels. Even though it isn’t amplified, in Los Angeles it worked better than both my amplified Terk HDTVa log-periodic/rabbit ears and Winegard SS- 3000. I had been carrying the Mohu Leaf Plus in my laptop bag, but the Ultimate uses a USB powered bias-tee power injector that is not tethered to the antenna. It can be placed next to the tuner or laptop making it much easier to set up. It isn’t necessary to run a separate power cable to the antenna.

I tested these three antennas by placing them one by one against the same spot on an interior wall roughly facing Mount Wilson. The Los Angeles transmitter sites were obstructed by three walls and the Santa Monica Freeway. To evaluate antenna performance, I did plots with a 500 MHz span centered on 300 MHz (50– 550 MHz) and 600 MHz (450–850 MHz) plus 100 MHz span plots centered on 195 MHz (145–245 MHz) and 605 MHz (555– 655 MHz) for each antenna. You can view all of the plots at

While the antennas were all tested in the same location, you’ll see some variation in the response of individual channels due to reflections from vehicles on the adjacent freeway.

Fig. 3 shows the 500 MHz plot centered at 600 MHz from the Leaf Ultimate. At UHF, all three antennas provided mid-band signals approximately 30 dB above the analyzer noise floor. The Mohu Leaf Ultimate and the FV-HD30 provided slightly better high-UHF signals relative to the noise floor than the Mohu Leaf Plus.

Fig. 3: Mohu Ultimate UHF band spectrum
(Click to Enlarge)
The amplifier in the Ultimate has about 10 dB more gain and there does appear to be some filtering of the frequencies above the TV band, with 700 MHz LTE downlink signals about 11 dB below the highest UHF TV signals. The LTE downlink signals were only 6 dB below the highest UHF TV signals on the other two antennas.

Mid-band UHF output level from the antennas on Channel 36 was about –69 dBm for the unamplified FV-HD30, –67 dBm for the Leaf Plus and –50 dBm for the Leaf Ultimate. The specifications on the amplifier used in the Ultimate are: minimum 15 dB gain, <2 dB noise figure, and an output IP3 of 37.5 dBm. With this much amplification that’s important and given the performance of some tuners, an unamplified antenna may do better in strong signal environments.

I did not check low VHF performance. The FV-HD30 picked up Channel 9 at –68 dBm, with a spectrum analyzer noise floor of –100 dBm. The Leaf Plus had –52 dBm on Channel 9 with a –80 dBm analyzer noise floor. Channel 9 came in at –48 dBm, also with a –80 dBm noise floor, on the Leaf Ultimate. Channel 7 signal levels were about the same as Channel 9. I didn’t see any indication of FM band filtering on these devices, although it appeared the Leaf Plus had a lower FM signal level when compared to high-VHF channels.

As you can see in the plots, the software is poor. For example, the graticule marks don’t line up with the scale. I hope Triarchy or a third-party developer takes advantage of capability of the TSA5G35 hardware and updates the software. The software can send data over an IP connection for remote monitoring so it should be possible to grab it and post-process it to add measurements such as channel power. Another project to add to the list!

Questions or comments? Email me at

  Print Page