Testing iPhone4's antenna performance

Tests carried out by PA Consulting Group have indicated that Apple’s iPhone4 does indeed suffer from connectivity problems compared to other smartphones.

In the majority of the tests PA’s wireless technologists found the iPhone 4’s performance was in the same range as the other smartphones tested (Blackberry 9700, HTC HD2), but it was consistently at the lower end of that range. And, as found by other testers, when used in the “death grip”, the iPhone4’s performance was significantly worse than other smartphones.

Putting a rubber band around the iPhone, similar to but not the same as Apple’s plastic bumper, gave a significant improvement in performance.

Watch the video of PA testing the iPhone - see top right of page

Our tests have received external coverage, in such journals as: 

• Daily Telegraph (external link)

• Hexus (external link)

• The Register (external link)

• Inquirer (external link)

• Smart Grid (external link)

• Computer Business Review (external link)

See PA's iPhone media page for more coverage.

Test 1: Baseline measurements with no hands on mobile

We tested the maximum power transmitted on a 3G uplink to the Vodafone network as the path loss to the basestation increases before the voice call dropped. The mobile was supported on non-conductive foam. The power in dBm was measured in 100kHz bandwidth at approximately 1m from mobile. The average shown is the average of the linear powers, converted to dBm.

Figures above in dBm

A higher power level is better as the mobile will be able to keep communicating with the basestation further into a fringe area. The Blackberry’s radiated power exceeds the iPhone by 2.6dB, with the HD2 in between. This is marginal and the user may not notice much difference in practice. It is notable that all three phones had the same power in their worst orientation (-43dBm) and the difference was the variability between worst and best orientations.

Test 2: Variability when handling the mobile

We tested the maximum power transmitted on a 3G uplink to the Vodafone network as path loss to the basestation increases before the voice call dropped. The power in dBm was measured in 100kHz bandwidth at approximately 1m from mobile:

Figures above in dBm

In the case of the iPhone the grip avoided bridging the gap between the two antennas. When the “death grip” was used and the gap bridged, the signal level dropped to the point where it was not possible to conduct the test as the call could not be established to begin with. This effect was not seen when handling either of the other mobiles. Note that all these tests were conducted in a test chamber, calls with the iPhone could still be established outside the chamber even when using the “death grip”.

It can be seen that the iPhone suffers more signal attenuation overall than the other two when handled, though the Blackberry can also be made to suffer significant attenuation with a particular grip. The HD2 seemed relatively immune to handling and with a light grip performed as well as when not held at all.

Test 3: iPhone antenna fix

Apple have started to offer a plastic ‘bumper’ that fits around the edge of the iPhone 4. This was not yet available in the UK at the time of testing so a rubber band that fitted around the edge to perform the same function was trialled. The exact results of the Apple bumper may therefore differ. 

We tested the speed of download and upload from the internet via a 3G connection to the Vodafone network using Speedtest.net, in kbps:

Figures above in kbps

The upload speed is quite consistent and substantially less than all the download speeds. It is probably limited at some point in the network. The download speed varied according to the handling.

There was no appreciable difference between open and closed hand with the rubber band fitted. The speed approximately halved when gripped firmly without the band fitted.

Test 4: as Test 1 but on test network

This simulated the network in a screened room using test equipment with the mobile set to transmit at maximum power. The power measured was the average RMS uplink power at ~0.5m from mobile in dBm:

Figures above in dBm

This agrees broadly with the live network tests, but the Blackberry has a further lead over the other two, and the difference between the iPhone and the HD2 is negligible.

Test 5. As test 2 but on test network

We tested the mobile held in two perpendicular orientations.

Figures above in dBm

All the phones show a degradation compared to hands-free operation, though the results are now closer than on the live network.

Test 6. iPhone antenna fix on test network

This test was holding the iPhone in the “death grip":

And repeated with the rubber band fitted around the edge:

Figures above in dBm

This shows a consistent improvement for both hands and both orientations with the band fitted.

Conclusions

The iPhone antenna performance is comparable with the performance of the other mobiles when handled or hands-free, though at the lower end of the range.

All the mobiles suffer attenuation when handled. The so called “death grip” gives a substantial further drop in performance for the iPhone to the point where we could not quantify it using the same test method.

Putting a rubber band around the iPhone, similar to but not the same as Apple’s plastic bumper, gave a significant improvement in performance.

To find out more about PA's wireless technology services please contact us now.