Author Topic: Altitude Accuracy  (Read 32830 times)

Admin

Re: Altitude Accuracy
« Reply #15 on: September 16, 2015, 06:05:30 pm »
Lee
Spoken with the boss (John Parker) - we are going to add the barometric sensor into all our boxes - when you are able to give us the pin outs etc.  We have ordered the magic bit (x10)!!!!!!!
Dave t

Hi Dave,

Its pretty simple actually, it uses the i2c interface, so only 4 wires to connect from the sensor,

GND
    GPIO pin 6, 9, 14, 20, 25 - take your pick !

3.3v
    I have used the 3.3v from the GPIO header pin 1 (for expediency), this is only rated at 50mA, so
    I would recommend using the 3.3v from the regulator if you are able.

SCL
    GPIO pin 5

SDA
    GPIO pin 3

I1
    unconnected

I2
    unconnected

Let me know if that is clear. I will update the build guide at some point.

Thx
Lee

trapdoor

Re: Altitude Accuracy
« Reply #16 on: September 16, 2015, 06:08:49 pm »
Have ordered two to add to my built boxes - this will fit quite happily to the Slice of Pod board. Again, happy to test the software when I get these installed (5 min jobbie )

The module says its temperature compensated - is there an issue with this getting baked on the instrument panel coaming?

Mark.

Admin

Re: Altitude Accuracy
« Reply #17 on: September 16, 2015, 06:19:48 pm »
Hi Dave,

Does this mean that if you have an ADSB equipped transponder already fed by an Altitude Encoder the barometric pressure sensor isn't needed - because it is only used if feeding data from pilotaware into a mode s transponder, or is the situation as below

In order for altitudes shown on Skydemon to be comparable with a received ADSB signal the pilotaware will require a barometric sensor (or the QNH updated manually in the web page)?

Complicated, isn't it  ???

This is nothing to do with the Altitude you are transmitting through your transponder, it is about comparing
relative heights between yourself and surrounding traffic.

ADS-B contains the Pressure Altitude, so altitude based upon an altimiter setting of 1013.25mb.
In order to correctly compare to ADS-B altitude against our own altitude, we need to either
1. Adjust the GNSS altitude using the given QNH
2. Use our own Pressure Altitude

The heights of traffic passed to SD, RunwayHD etc are relative +/- heights to our own position, not absolute heights.
So in order to calculate those relative height differences, in the case of ADS-B, we compare to our Pressure Altitude
in the case of PilotAware we use our GNSS Altitude.
The Navtool simply receives a message indicating the relative traffic height.

I hope that is clear

Thx
Lee

Admin

Re: Altitude Accuracy
« Reply #18 on: September 16, 2015, 06:28:56 pm »
Have ordered two to add to my built boxes - this will fit quite happily to the Slice of Pod board. Again, happy to test the software when I get these installed (5 min jobbie )
The module says its temperature compensated - is there an issue with this getting baked on the instrument panel coaming?
Mark.

I did some hand calculations, and I was not too concerned about the total impact it would have on the altitude figure.
What we are trying to do here is make the figures more accurate, they are never going to be within 30cm which is the
claim on the datasheet.

Unless you know something I don't ?

falcoguy

Re: Altitude Accuracy
« Reply #19 on: September 16, 2015, 08:12:16 pm »
hi Lee

perfectly clear, pin outs simple - but how to dismantle our IDE connector and insert extra cables I'm not too sure.  We have the digole module all soldered and glued so don't want to disturb soldered cable on the digole board.  Unfortunately we cut off unused cable on the ribbon! Damn!
I'm sure we'll sort it somehow
Will implement as soon as the devices come.  We have 7 working Pilotaware but I can, as yet, only confirm 2 with successful arf transmissions - progress is being made.

Fenland have a LAA and ultralight fly in this weekend (19th Sept) John Parker is suggesting we also use it as a PilotAware hotspot and meet up?  What do you think - if OK I'll put a topic on the forum

we will have 2 in planes by then not sure about the rest of them - we are still in initial ground testing

dave t

ianfallon

Re: Altitude Accuracy
« Reply #20 on: September 16, 2015, 09:15:05 pm »
Cool - ordered a couple of MPL3115A2 s. Looking forward to the s/w update.
« Last Edit: September 16, 2015, 09:31:37 pm by ianfallon »

SteveN

Re: Altitude Accuracy
« Reply #21 on: September 17, 2015, 12:09:06 am »
The majority of the GA fleet  (being old) will have a best altitude encoder resolution of 100ft (GRAY code 1bit = 100ft). More modern serial encoders MAY report to 25ft if set up to do so.  As there is no way of knowing which encoder is fitted to a target aircraft 100ft (minimum) has to be assumed.  Built in barometers in devices not connected to pitot will also suffer from cockpit air compressive effects.  So that's really 100ft at best AND if the encoder has been properly calibrated.  Some encoding altimeters can also be way out. Just walk around a hanger......

I don't think vanilla FLARM has a barometer but it is only looking at other FLARM for collision so they have a known calibration and resolution playing field. PowerFLARM definitely does have a pressure sensor.

So in our case we have a number of differing possible sources for reported FL to contend with.

For my 2p if it reports closer that 300ft above or below I'd consider it as a threat even with a barometer fitted inside our boxes.




« Last Edit: September 17, 2015, 01:23:32 am by SteveN »

Admin

Re: Altitude Accuracy
« Reply #22 on: September 17, 2015, 09:56:20 am »
Great Post,
I had no idea some Pressure Sensors were so inaccurate.
The Sensor I have been using seems very good, it claims 30cm of accuracy, which is way above what is required.

My understanding was that Mode S transponders had altitude reporting in steps of 25ft, I cannot find any
information regarding calibration inaccuracy.

Bear in mind that PilotAware is only using information from Mode S using ADS-B or other PilotAware units, so I
was under the assumption that the Mode S figures would be reasonably accurate.

There are situations where Aircraft fitted with both ADS-B and PilotAware would be transmitting 2 Altitude figures
1. The ADS-B Pressure Altitude
2. the P3I GNSS Altitude

At the receiving end, I use ADS-B in preference to PilotAware, in other words I receive 2 records of information with the
same ICAO code, lets say 0x40abcd
0x40abcd (ADSB) Pressure Altitude
0x40abcd (P3I) GNSS Altitude

If I wanted I could change the Altitude transmitted by P3I to be the Pressure Altitude, or even send GNSS & Pressure
altitude on alternating packets with an indicator to identify what is being reported.

I need to think on this a little further to decide which is the best piece of information to use

SteveN

Re: Altitude Accuracy
« Reply #23 on: September 17, 2015, 10:30:08 am »
There are plenty of Mode S Garmin transponder upgrades out there that retained a pre-installed GRAY encoder. The ubiquitous Ameri-King AK-350 is 100ft.

I note OGN's tracker protocol has a bit to declare whether altitude is Geo or pressure.  I also read OGN refuse to play the traffic avoidance game at all.

http://wiki.glidernet.org/ogn-tracking-protocol

I know it's a moot point with you Lee but have you thought about cooperation. For me today FLARM or OGN targets are a much greater risk to us than the current very sparse ADS-B stuff.  5 years time perhaps..

ianfallon

Re: Altitude Accuracy
« Reply #24 on: September 17, 2015, 11:47:02 am »
Been thinking about the connections for the MPL3115A2 with the Slice of POD.
I think this would be correct (?) ...
« Last Edit: September 17, 2015, 09:27:39 pm by ianfallon »

ianfallon

Re: Altitude Accuracy
« Reply #25 on: September 17, 2015, 11:51:18 am »
For ref: the other side of the SOP ...

Admin

Re: Altitude Accuracy
« Reply #26 on: September 17, 2015, 12:00:24 pm »
Been thinking about the connections for the MPL3115A2 with the Slice of POD.
I think this would be correct (?) ...

This is the version I used
https://www.sparkfun.com/products/11084

Your version looks slightly different, do you have a datasheet ?
It looks like it also has a 3.3v Regulator on board, and the 3.3v is labeled 3vo - is this an output ?
If so I would say it probably requires 5v input rather than 3.3v input - difficult to say without a
datasheet
Thx
Lee

(UPDATE)
Found the Adafruit description
Quote
Using the sensor is easy. For example, if you're using an Arduino, simply connect the VDD pin to the 5V voltage pin, GND to ground, SCL to I2C Clock (Analog 5 on an UNO) and SDA to I2C Data....
So this already has a regulator built in - I wonder what it can supply on its 3.3v output ?
« Last Edit: September 17, 2015, 12:07:28 pm by Admin »

ianfallon

Re: Altitude Accuracy
« Reply #27 on: September 17, 2015, 01:19:41 pm »
Ah yes - got that diagram slightly wrong - will remove it and correct it later.

Yea - I wonder whether it would drive the ARF - academic with the SOP board but curious.

ianfallon

Re: Altitude Accuracy
« Reply #28 on: September 17, 2015, 09:30:33 pm »
Corrected SOP diagram with 5v connections to the Adafruit MPL3115A2.

EDIT: Added it as described in my diagram and on the new software build it indicates it is working  :)
« Last Edit: September 20, 2015, 04:37:20 pm by ianfallon »

falcoguy

Re: Altitude Accuracy
« Reply #29 on: September 18, 2015, 12:28:05 pm »
Our MPL 3115A2 have arrived
Will install next week
Can I get on and wire up without a updated operating system?
dave T