Bt-20 accuracy

[primatarus]

OK, I've read the posts on accuracy, lack of WAAS, and the great chipset in the BT-20 (i.e. use of Kalman filtering), but have not yet seen any statistics on the accuracy of this device. I have passed a ? to the folks at Delorme in re accuracy (average, plus/minus, and probabilities), but am interested in user feedback (on a blue sky day, how accurate is this device vs. others on the market that operate under WAAS correction routines).

Thanks

[rcollins45]

I couldn't say how accurate the device keeps in inches, but I can say it's particular in what it's accurate with.

When connected to my laptop and using Delorme's 2009 software, it seems to be VERY accurate.

If, on the other hand, I try using it with my Blackberry (running Garmin's Mobile software), it plots my location as 7 blocks away from where I really am.

Using the puck from my TomTom Navigator for Palm package works with this same Blackberry device and is every bit as accurate as the laptop is with the BT-20.

[RegB]

Subjectively;
SA2009 in combination with BT-20 (on a USB cable) seems to be less sensitive than SA2007 was with LT-20.
I don't know where the filtering/sensitivity settings are, but I can put either the BT or LT above, below, left or right of my Garmin Zumo that is happily tracking 3D and the SA software stays in the acquiring mode for 1/2 a day.
The "sat info" tab shows maybe one or two red satellites, the rest in gray, no green ones.
I know I have a clear view of the southern sky, as I said the Garmin Zumo tracks just fine from that dashboard location and I can put the delorme puck over, under, left or right of it, but it still doesn't recognize enough sats to get a fix.

So, IS THERE software filtering of the data stream ?
e.g. check summing as a sanity check ?
and does it just discard messages that fail the checksum ?
I would LIKE to have ECC on the messages, but maybe that is expecting too much.