AX5043 RSSI register update rate

Hi, out there,
By any chance, has anyone an idea of the rate at which the RSSI register is updated depending on the settings used, of course. Some time ago I posted the following message in another discussion but didn’t get any answer. Maybe this time I’ll be luckier…

… for some time I’ve been looking for data as regards the RSSI time of reaction. In other words, knowing that the RSSI register data reflects the signal behind the channel filter, a sensible way of updating the RSSI register would take into account the selected IF bandwidth which depends on the DECIMATION register. So far so good. However, prior to RSSI updating, there could also be some kind of filtering, be it simple averaging or something more complex. Of course, there are a couple of registers “TMGRX-something” and more specifically TMGRXCOARSEAGC, TMGRXAGC and TMGRXRSSI (all three with M and E) that can be involved in a way. However, even those set to their minimum value, the RSSI seems to be very slow compared to a given ASK bitrate. For example if set for 4 kbit/s ASK, the RSSI will hardly notice a 1,5 ms pulse, say 20+ dB down from the level obtained with a 10 ms pulse. … is there anything else I should consider in order to resolve this apparent discrepancy ? To put it another way, what settings are needed for detecting 1,5 ms pulses using the RSSI register ? Thank you in advance for any useful information.
Kind regards.

To put it another way, what settings are needed for detecting 1,5 ms pulses using the RSSI register ?

This is the information what I do have and might be useful regarding your question.

The sensitivity point depend on the Bandwidth and the PHY Rate. It can be as low as -138 dBm for PSK Rates in a 0.2 Khz band and -138 dbm @ 100 bps, 868 MHz FSK, FEC
The sensitivity point is evaluated when 90% or more packets are successfully received. It can receive data from -115 dBm up to 0 dBm.

Dear Martin,
First of all, thank you for your time and your commitment to help.
Having read your answer, I had the feeling that I hadn’t been fully understood, so I’ll try to be more specific. My application isn’t related to data transmission per se. All settings related to a modulation scheme are irrelevant, but the necessary IF BW matters a lot for optimum sensitivity. I’m using a proprietary ASK SS transmission and knowing the signal strenght is essential. Hence I have no choice but read either the RSSI register or the AGCCOUNTER register. I’m not using the data output at all. I have some experience with the AX5043 but in this very application I have an issue. The IF BW settings lead to a limited Baudrate but, of course, the AGC and, consequently, the RSSI are generally slower and even necessarily asymetric in terms of attack/decay time. My basic question is: how to obtain an AGC/RSSI time response which is comparable with the minimum symbol duration for a given IF BW. I presently don’t know if this is possible. There’s a register called AGCGAINx that is said to control the AGC time constants. Maybe I could investigate in that matter. Have you any comment on the above ? TNX.
Best regards.