GPS Curiosity

[lolachampcar]

ADS-B out will require a TSO'd GPS source by 2020.  There are a good number of very inexpensive high performance chipsets out there with more on the way.  I was curious and read through the letter the FAA issued terminating its program working with LightSquared (http://apps.fcc.gov/ecfs/document/view?id=7021860337).  It provides some information on the requirements for TSO'd GPS receivers, specifically, their sensitivity and notch frequency rejection capibility.  The whole read, together with the upcoming mandate, made me wonder about what is involved in getting a Dynon type "GPS Puck" (die mounted chipset on the back side of an antenna substrate potted in an antenna mount enclosure) certified.

What is involved from a design and cost standpoint to get a GPS to meet TSO standards?  Are you aware of anyone out there working on doing this (like Dynon?)?  Given that the Dynon receiver is a WAAS unit, how does it compare to something like the certified Garmin units when it comes to accuracy?  If they are similar as I suspect they might be, is there simply a flag that is transmitted in the ADS-B out packet that identifies the certification status of the receiver separate from the signal integrity or accuracy of the reported position?

 

[dpbarnes]

Grand Rapids at Oshkosh was saying they are developing a certified gps. I was surprised because my understanding is that a company the size of a Grand Rapids or Dynon can't justify the large investment in time and money it takes to certify a gps. I haven't checked any further since all my eggs are in Dynon's basket now.

 

[lolachampcar]

Given that most of the TSO requirements seem to be performance based, I would think the chipset vendors could generate a reference design that meets the standards. It would be neat to take that to an open source or non-profit certified puck offering. Plugging that directly into a Dynon plane would do the trick

 

[dynonsupport]

The TSO for a GPS cannot be proven with just testing. You cannot certify anything with software in it via testing, you have to analyze and trace source code. It requires deep, intimate knowledge of the way the product works internally. Also, the way the TSO is written, it is likely not possible to certify a "puck" GPS, you can only do it with ones made of discrete components (which is how all current certified GPS units are manufactured). Of course, no modular GPS manufacturer is going to share source code or the inner workings of their units just to sell a couple thousand more units.

I believe that Grand Rapids is likely partnering with another company to co-brand a product. It does seem unlikely that they are doing all the development in house just so they can sell a few hundred units a year. Obviously only they can answer that for sure. If they can design or source one in the $1K range, then that will be a nice product for them. At their current price of $3,200 to $4,400, that seems like a very expensive way to get ADS-B out legal which you don't need for 8 years.

One thing to note is that having a certified GPS receiver does not suddenly make anything hooked to it legal for GPS approaches. In that case, the databases and whole map software need to be certified too. A simple certified GPS puck would only be helpful for ADS-B out.

P.S. The GPS TSO is 800 pages long!

 

[lolachampcar]

My thoughts were more along the lines of getting the chip set vendor to offer the certified reference design. I should think the nature of GPS chip sets would dictate that the silicone vendor would be involved with just about any certified product and that certified reference designs already exist. I could not fathom the task of digesting existing chip set code then trying to represent it to the FAA for certification. The code base would be obsolete before you could begin to understand it.

I understand your point about a Skyview front end on a certified GPS receiver not being a certified system. My primary curiosity was regarding ADS-B and how the need for a bunch of low cost certified GPS sources for ADS-B out was going to be addressed. I seem to remember a comment about dozens of certified sources coming on line over the next 8 years. It would be worth looking into if a chip set vendor or two had already done certification work on their part of an end certified product as perhaps the engine could be certified stand alone as a position source for the transponder.

Where does the degree of accuracy or level of certification appear in the -ES message? Is there a certified versus non-certified flag or just an accuracy factor for which the lowest is used for non-certified sources?

 

[lolachampcar]

Extended Squitter

This helps

Page 8
http://adsb.tc.faa.gov/WG3_Meetings/Meeting30/1090-WP30-21-Appendix_A%20Mods.pdf

 

[dynonsupport]

We are not aware of a "chip set" or "silicon" vendor that has done a certified design. All the designs we know of are done with discrete components. But we're far from experts on this. There is no "certified reference design" that we know of either. I imagine that if there was, there would be people selling affordable certified GPS modules.

We do expect that in the next few years, many vendors will come out with affordable, certified GPS receivers for the GA market. Remember that all GA planes in the USA will need these, so the market is a lot larger than just experimental and light sport planes, so there is a good market for someone to take advantage of.

The quality of the GPS is sent in an accuracy bitfield. If you know the accuracy, you transmit it, and the only way to "know" the accuracy is to be certified. In non-certified, you transmit your accuracy as "unknown."