In case anyone is interested, I bought the device I linked to above. It works really well. I have downloaded Dashcommand with Mazda parameter and Torque Pro from the Android marketplace. I have only used it in the car for a couple hours, and taken a couple data logs. Here is a VERY quick review.
OBDLink MX: Awesome! I literally plugged it in and had it paired to my devices in less than 2 minutes. Very simple to pair. There is one tiny button on the device to push for pairing mode. Just pair up as normal with your android device. I am using a Galaxy Note 10.1 tablet, and will probably also use. I havent touched the device since I plugged it in, just fire up the engine and it wakes up and pairs with your device.
No Negatives yet.
Torque Pro ($5): Very nice, but lacks Mazda parameters. Has all the standard SAE PIDs and probably some more. Has logging capability that I havent used yet. Has DTC, live data, diagnostics, and a simplified data logger which graphs realtime data points on screen. The simplified DataLogger works really nice for simple plots that you can view real-time or while out driving. The array of gauge types and simplicity to customize your own dashboards is excellent. I much prefer the gauges in Torque Pro to that of DashCommand.
Negatives: No Mazda extended PIDs (that I can find). More about Mazda PIDs below in my review of Dashcommand.
DashCommand ($20 with Mazda PIDs). Not quite as user friendly or pretty as Torque Pro as far as custom gauge set ups. It does have some cool prepared Dashboards. It has all the functionality of Torque pro, plus Mazda Extended PIDs and maybe higher end datalogging capability (*havent used Torque Pro Datalogging much yet). It also has some interesting things going on with calculating shift points and horespower etc. Probably not too useful for the CX5. It asks you to enter things like drag coefficient, BSFC, VE, gear ratios etc. The datalogging for DashCommand is best analyzed at home on a laptop after exporting to a .csv. Honestly, I would use Torque Pro UNLESS you want the Mazda PIDs. Dashcommand seems geared more towards performance crowd who are trying to do custom tuning.
OK, so what can you see for Mazda PIDS? For me, the main one is Knock Retard, which is the learned timing retard based on knock events. This parameter is nice to look at if you are trying to decide if increased octane ratings may benefit in power and economy. It also has Cam timing positions, injector flow, and some Mazda specific PIDs that overlap the SAE ones (maybe more accurate?). There may be more, but I havent messed with them yet. I am going to take a look at the Cam timing positions today, since that is another important SkyActiv parameter.
Negatives: Clunky custom dash with limited gauges (e.g. no digital readouts etc). Datalogging is powerful, but somewhat of a PITA to get to a CSV format. You must download a trial version of the companies ScanXL to export. The pre-made Dashboards are very pretty, but annoying in that you cant change anything.
So, what have I learned so far?
I see very little knock retard, but I had just filled up with premium. Max retard I see is -1.5 degrees VERY infrequently. Considering that it is common to run 20 to 35 degrees advance, the 1.5 is negligible. I will log again with 87 later. Initial impression is that there will be no benefit to running premium.
The thing that has me wondering a little bit is that I see a long term fuel trim that is up in the 15% range at cruise. This may be due to cold temps or the fact that I had run about 3 tanks of 10% ethanol prior to filling with 91 straight gas. This would explain why I am seeing measured fuel economy down in the 25-26mpg range. I will reset the ECU and log again. I would eventually like to understand how much ethanol affects the mileage (we all know its a 5-10% theoretical reduction). I also believe that our winder blended gas here really sucks to. I normally get 29-30 mpg in the summer.
One other interesting item I have seen is that the MAP is indeed very high during cruise (26-28 inHg in the manifold vs 28.9 inHG barometric). I posted an article on this board which explains why. Essentially it is part of the quasi-Atkinson cycle operation. The intake valves actually control engine power and keep the throttle plate mostly open to reduce pumping losses.
So why am I doing this? Mostly just to learn. Its not like there is much in the way of tuning potential, nor am I really interested. I would like to have some data that confirms what the best fuel to use is, what the best cruise speed is etc. I also may use it as a test bed to get in to some of the systems engineering capabilities of OBD and CAN as that would be useful for my job.
