Lenovo Thinkbook E16 connection drops with Hap ax s

Hardware
1

Hi there,

I have a problem with a Hap ax s (set up as a AP). All is working ok, but my Lenovo Thinkbook E16 with Realtek 8852CE wifi chipset has wifi drops through use of wifi.

Does anybody have the same problem? So any other device connecting to the Hap ax s has NO problem.

2

I don't think your Hap ax s is the issue here.

I have made very bad experiences with Realtek WIFI Chipset in combination with AMD Hardware. Exactly these drops through use of WIFI, too. I have tried many many things, to solve exactly this problem, because the company I was working at, had bought a batch of new Laptops with AMD Hardware and Realtek WIFI Chipsets. Lenovo and HP could't help to solve this problem and blamed the APs. We even have tried to change more exotic parameters like PCI-E energy saving behaviour etc. without any success. Only slight improvements and less drops.

In the end we switched the WIFI Chipset to Intel WIFI Chipsets ourselves and with this change only, all problems with drops went away.

I guess thats why most AMD based Laptops use Intel or MediaTek WIFI Chipsets now.

3

I haven't had any problems with other Mikrotik AP's, so it's very strange.

I cannot replace the wifi-chip just for this combination.

4

Not sure the age on the RealTek stuff, but one issue I have hit is that MikroTik tends to update available frequencies as regs change, and older chipsets my not be able to use some of the new ones, and I have had to exclude some of the newer stuff from use to get stability.

This may be irrelevant here, but might be worth a few minutes to check.

5

It is a Thinkpad E16 from januari 2025, so not that old.

6

The MikroTik WiFi devices do, if not told otherwise, use whatever this chipset says it can do coupled with the Regulatory domain in use, and could differ from other AP's . . . still might be worth a minute or two to compare. Other than radio specifics/defaults, I would expect the ROS side of things to behave the same on all platforms due to the common code.

IE, if you create a limited config declaring just about everything (don't allow anything to just "fall through" and default) and test that, does the problem persist? If so, it's looking more and more like a chipset/driver issue.