Menu Sign In Contact FAQ
Banner
Welcome to our forums

ADL / Golze satellite weather system

A protocol is just a piece of software that can be implemented …

Sir_Percy wrote:

A protocol is just a piece of software that can be implemented …

Maybe call Sebastian and ask him?

EGTK Oxford

One more happy ADL user On the very first flight with the system we could already make very good use of it: There was a small area of heavy precipitation right on our way to our IAF.

We could see in flight that nothing bad hides behind and a little diversion to the south did the trick.

EDFM (Mannheim), Germany

Very useful for me yesterday too. With the wind info, managed to turn a 25kts headwind to a 15kts crosswind by descending a bit. While it was towards the end of a 7h trip (EGHH → LEBG → LPSO), the feeling of doing a successful optimisation is worth many times the minutes gained!

Also getting Metars for fields without ATIS, allowing for planning well in advance of approach (and nice routing optimisation)

Terbang,

Were you VMC when you went to the south?

The reason I’m asking is that despite the excellence of the ADL product, using it for tactical avoidance can drive you into the very area you should avoid, because of the delay. It is possible to add some buffer by factoring in the winds and the cell movement obviously. But it should not, and I think Sebastian would agree with that, be used to find a way through an area of convection.

Apologies for repeating something that’s been said before and which is probably obvious to many, but as new forumites come, this is a safety information worth keeping in mind.

EGTF, LFTF

The cells were forecast to move north and this is what this one did. This is also reflected in the two images. In fact, it had already moved further north and there was no precipitation when we passed the green area. We were in VMC and we could see the ground but there was a cloud deck above us (also visible in the images) and visibility wasn’t great. The advantage was that we could be reasonably sure that there was nothing hiding behind these clouds.

EDFM (Mannheim), Germany

Sir_Percy wrote:

A protocol is just a piece of software that can be implemented …

First the ADL system uses sophisticated data compression. That was designed with the sole purpose of reducing transmission volume at the cost of massive memory and CPU usage for compression and decompression. The design assumed that the (rather powerful) iPad would do the decoding anyway. This is what happens if you use the ADL device with the ADLConnect app.
But now with the GDL90 protocol the ADL box itself has to decode and encode the radar data to deliver it in GDL90 format (which is in fact quite inefficient regarding data compression). The ADL110B/120/130 CPU is unfortunately just not up to that task. With a typical radar image that process alone might take 1-2 minutes and more.

Second the ADL device has to broadcast the GDL90 data continuously in parallel to its other communication with the ADLConnect app etc. The WiFi chip of the ADL110B/120/130 unfortunately does not support multiple parallel communication channels.

I see three possible solutions for the owners of those older ADL devices:
1) Trade it in for a new ADL140/150/200. This is usually ok unless the original box is fitted with some paperwork and exchanging the box might void that paperwork.
2) A second device or iPad background process could act as a proxy to broadcast the GDL90. This is possible but comes with a lot of detail issues.
3) I could come up with a “repowering” option for those devices. That would mean sticking in the ADL140/150/200 processor and WiFi chip into the older devices. I think that is the best option. But it will take some development time to get this done and it would take at least until early 2019.

www.ing-golze.de
EDAZ

Option 4) is to rewrite it all in assembler:

loop: ld hl, 5000h
ld a, h
or l
jr nz, loop

The above is just an example

Administrator
Shoreham EGKA, United Kingdom

Peter wrote:

Option 4) is to rewrite it all in assembler:

You probably won’t gain anything, modern compilers do a very good job generating highly optimized code. Sometimes you may have a chance if your processor supports vector instructions (i.e. SIMD), or, if the algorithm allows for that, by using software pipelining or other very sophisticated techniques, but even these concepts are exploited by modern compilers. In most cases you won’t beat compiled code.

EDFM (Mannheim), Germany
Sign in to add your message

Back to Top