Can Advertising range reach as far as disconnected range?

[JasonGeo]

Hi,

I am running some tests with the long range example provided by Nordic and observed this behavior which I can't account for.

https://devzone.nordicsemi.com/nordic/nordic-blog/b/blog/posts/testing-long-range-coded-phy-with-nordic-solution-it-simply-works-922075585

The 2 nRF52840 DK boards were connected in coded PHY and the distance between them were increased until a disconnection occurs. Maintaining this same distance (540m) when disconnection occurred, we try to establish connection again by waiting for coded PHY advertising packets to hit the receiver. But it was unsuccessful, no connection was made.

We then move towards the advertising DK a couple of metres hoping to re-establish connection again but was not successful.
We have to move towards the advertising DK until the distance between them was about 350m before the scanner is able to pick up the advertising and re-establish connection.

My question therefore is Advertising range able to reach as far as connected range? It would be fair to say advertising range would be slightly shorter than connected range by a few metres but not by hundreds of metres? 
Is there any significant delay between the time the receiver fails to "sense" the connection and the bluetooth stack reporting disconnection status?

Say if I have a sensor placed at a distance that is almost at the threshold of losing connection and is working properly. Then someone or an object is placed in between the sensor and the reader and affected the signal strength and the connection is lost. I will never be able to talk and recover that sensor anymore if advertising range is shorter than connected range, will that be the case?

Can you please demystify this for us.

[Simonr]

Hi

As the "Testing non connected states" mention in the blog post, an important step is to set up the scanner to scan all the time to make sure that you have the biggest chance to pick up the advertisements. If not, you might have to wait a while for the scanner to find an advertisement on longer ranges, as more will be missing at longer ranges. You should indeed be able to scan and connect to an advertising device on the Coded PHY at longer ranges than ~half the range you stay connected in. Are you sure you advertise and scan on the Coded PHY?

Best regards,

Simon

[JasonGeo]

Hi Simon,

Thanks once again for supporting the ticket.

As mentioned, the test was done with the same Long Range example from Nordic which you provided the link for. So the DK boards would be advertising and scanning on Coded PHY assuming I followed the instructions correctly.

I am abit confused with your second last sentence. Are you saying if the boards disconnect at 500m, I should be able to re-scan and connect from a distance of approx 250m to 500m?
Or are you saying I am able to scan and connect from a distance of 500m to 750m?

From my testing, I am definitely not able to re-scan and re-connect at more than 400m after disconnection at 500m. I can scan and connect at distance 350m or less.

[Simonr]

Hi Jason

With my second to last sentence I meant that if the device disconnects at ~500 meters you should be able to scan and connect to an advertising device near 500m (not more) as well. Sorry about any confusion. How exactly do you test up to 500m?

How fast do you increase the distance between the two test devices? If you move them too quickly, the disconnect reason might be a connection supervision timeout a while after the last transmitted message. Do you confirm that there is communication between the two devices at a distance close to the disconnect reason? You can try decreasing the connection supervision timeout.

Best regards,

Simon