I had been hoping to get eWEEK’s review of Palm’s new Treo Pro smart phone online this week, but testing took an unfortunate right turn, leading us to return the device and wait for a replacement before we could finish the tests. Although I am bummed about the delay, I had a lot fun troubleshooting the issues — even though I didn’t come to a satisfactory conclusion as to what the real problem was.
In a nutshell, the device was getting pretty terrible battery performance on eWEEK Labs’ talk time tests — almost half of what I expected given Palm’s published specifications. In my first talk time test, with the radio set to auto-select the best network, the battery only lasted about 2.5 hours, while the specs claimed 5 hours. According to the device indicators, the call was placed over the 3G network. Although the 2.5 hours would be an absolutely terrible result for a 2G phone, it still is poor compared with the 3G results we’ve seen recently from the iPhone 3G or the Nokia E71.
I didn’t think the problem was caused by background applications eating up CPU time because I had done a hard reset on the phone right before testing. Any background services running after that I felt were fair game, since they are automatically loaded on a freshly restored device. The restore also disabled the Wi-Fi and Bluetooth radios. And I didn’t think it was the battery, either, since I’d used the device over Wi-Fi plenty in the preceding days, without noticing anything out of the ordinary.
To ensure the battery wasn’t draining because the phone was thrashing between 2G and 3G networks, I forced the connection to 3G and tried again. And again, the battery died after 2.5 hours.
Confused and looking for answers, I enlisted Palm’s help for troubleshooting at this point. After I explained the tests and everything I had tried in an attempt to resolve the problems (and what I was seeing from Palm’s competitors in identical tests, to boot), the company told me that the AT&T network wants to give priority to the 3G network. When faced with a strong 2G signal and a somewhat weak 3G signal as reported by the device, the network nonetheless commands the handset to utilize 3G — unless the device is manually forced the other way or the 3G signal drops below some unstated threshold. Since Palm’s device gets really good reception (the company’s claim), it would almost certainly stick to the 3G network better than other devices would, and that extra love to the 3G network would mean worse overall battery performance.
Palm also told me some information that contradicts its claims about AT&T’s network a bit — telling me that not all SIMs are created equal, as some may default home to a 2G network over a 3G (or vice versa). And then I was asked to try the test using the SIM used in the Nokia test. (Note: eWEEK Labs does not have testing accounts with the various GSM-based operators. Instead, we test using the SIM sent to us with the phone.)
Something still didn’t seem right with these arguments, since ultimately the Treo Pro’s 3G numbers were still pretty poor. Figuring Nokia’s PR firm wouldn’t be wild about me testing another product on its dime, I tried something else to see if the battery was the culprit. I tried the talk time test again, this time forcing the radio to 2G, figuring that if the result was in the same ballpark, it had to be the battery.
Instead, I found there was something very wrong with the 2G radio. The phone would report an EDGE connection and show the received signal — four bars in the place I conduct the tests — but no traffic would actually pass to or from the device. I couldn’t place a call, nor could I receive one. No dial tone, no ringing — nothing. Just to make sure there wasn’t a problem with the local cell, I grabbed my gen 1 iPhone and made a call, then downloaded an application from the App Store. I tried placing a call with the Treo Pro again — nothing.
Not a cell problem, apparently.
I’m guessing there are a bunch of problems with our test unit, and perhaps what I’ve seen — the bad talk time performance and the lying 2G indicator — are just symptoms. The radio problem would not explain the bad battery performance on the test where I forced the connection to 3G. I could see a comatose 2G radio causing the problem when the radio is set to auto-select — there would be thrashing because the signal detected does not equal the actual traffic that gets through — but the 3G test was just as bad and there shouldn’t be any thrashing in this case.
I guess I will never know what the real problem is. Ultimately, there’s something wrong with the phone, and I need to get a review done. Palm is sending another one, and we’ll get the review online as soon as we can.