http://blogs.msdn.com/windowsmobile/archive/2006/08/16/702833.aspx

The Windows Mobile Team has just published an article on the power consumption of Pocket PC devices, which is a continuation of their article on the MS Smartphone power consumption article. My avid readers that have also read my article on the Pocket PC power consumption issues will surely want to check it out too. It is a very interesting read. I'd like to add some of my measurement results to the statements in the article: that of the Bluetooth power consumption, which, in my opinion, isn't that bad. (The article doesn't say HOW bad it is - just that it should be disabled to enhance battery life.) That is, if you have a Pocket PC Phone Edition device that you often use as, say, a Bluetooth modem (just like me, who access the Net from home via GPRS/EDGE as a BT modem connected to both my other PDA's and my notebook for quite a lot of time) for your other Pocket PC's, notebooks or other kinds of computers and you are stick of always en- and, then, disabling the Bluetooth before/after every session won't significantly (with more than 20-30%) enhance the battery life, particularly if you keep your Pocket PC not that often in standby state (when its Bluetooth unit is still on and is listening to incoming requests, including Internet connection requests from clients). Of course, I've made some exact measurements to ground my statement. On a HTC Universal (a.k.a. i-mate JasJar, Qtek 9000, O2 XDA Exec, SPV M5000, MDA IV/Pro), the Bluetooth unit only consumes around 2 milliAmps. This means, assuming the Pocket PC itself (powering the RAM and the phone unit to listen to incoming calls) consumes at least 5 mA when suspended, always enabling Bluetooth will only result in a 40% availability degradation - at most. That is, only when when you don't touch the Universal (or, for that matter, any up-to-date HTC Phone Edition device - I don't think the built-in Bluetooth hardware of other, current HTC Phone Edition devices would be far inferior to that of the Universal and the Wizard, power consumption-wise) for days. In reality (assuming you do use the phone a lot and, therefore, you end up recharging it every (second) day), you will encounter far less availability time reduction with the Bluetooth unit always enabled (because other units - for example, the LCD backlight, the Wi-Fi etc. - consume far much power; compared to them, the additional power consumption of the BT unit is negligible). That is, it's not because of the always-enabled BT unit that you will end up having to recharge the PDA every day if you actively use it - then, its impact on the entire battery life may be as low as 1-2%. Note that it's, unfortunately, not possible to reliably measure the power consumption of the HTC Wizard (a.k.a. imate K-Jam, T-Mobile MDA Vario, Qtek 9100, MDA Vario, XDA mini S, SPV M3000, VPA Compact II, Dopod 838), but you can take my word: it has stellar battery life with the BT unit always enabled. I need to recharge it every 2-3 days only. (I don't use it much as a PDA, except for occasional mail checks and quick browsing sessions.) Here's an acbPowerMeter screenshot (please read my above-linked article on how the results should be read) showing all this: showing the Universal with the BT unit enabled for exactly two minutes (with the current used being between 125 and 127 mA's) and, then, with disabled BT (the current, then, was between 123 and 125 mA's). The difference (that is, the power consumption of the (idle, listening) BT unit) is around 2 mA's, which will be the same in suspended state too. Note that I've made the tests in a freshly hard reset device with the phone unit disabled so that nothing else had any effect on the battery consumption. All in all, if you need Bluetooth a lot and find it very tiring to always en- and disable the Bluetooth to conserve battery power, you may rejoice: always enabling it doesn't consume much power on any of new (WM5) Pocket PC (Phone Edition) device. At least not on any PPC (PE) device that I've tested or I've been using so far - there surely may be exceptions (but surely not manufactured by HTC - for example, I'm not sure about Asustek-manufactured devices like the Fujitsu-Siemens t830 because they may be using entirely different chipsets and hardware than PPC PE devices manufactured by HTC.) The Buetooth in the Universal, the Wizard, the Dell Axim x51v and the Pocket Loox 7xx (they have all been manufactured by HTC) definitely doesn't consume much power. UPDATE (21/08/2006): I've tested the power consumption of my HTC Wizard (which I, currently, don't actively use and, consequently, can long-lasting tests like these on it) with disabled radio, enabled Bluetooth (without being discoverable - I'll run some checks to see whether enabling it will result in a more pronounced battery consumption) and enabled infrared auto-receive (as I've pointed out in my above-linked article, it does NOT result in the device's consuming more power - as opposed to what most people think). It has turned out to be consuming about 6% battery charge a day - that is, the Bluetooth unit consumes really-really little power. I will continue my measurements; I'll keep you reported. UPDATE (08/26/2006): Read this article for further HTC Wizard results. As has turned out, the power consumption of the Bluetooth unit of the Wizard is almost negligible. UPDATE (08/30/2006): posted an article on the power consumption of the HTC Universal.