Well, I am a little out of ideas. I am more of a user than a coder. @lleachii's proposal to check the voltage of the radios / card sounds like something that could be done at least. Especially, in light of the post you found about irregular current on this Unielec board.
The linux regulatory database in principle stipulates the TX/RX power limits (per dBm) for each channel. If the documentation about CRDA is still correct, then following picture shows the parts of code that are (or can be) involved in setting the actual dBm:
To me, a limit of 24 dBm instead of 27 dBm seems plausible in following cases:
A) you are not using a current version of OpenWrt with up to date regulatory data --> but you told us you used snapshot versions, so this is unlikely
B) In some countries, there are regulations about devices being mitigated by 3 dB of the maximum allowed transmit power in certain DFS channels, if there is no DFS power control scheme in play. Quoting Wikipedia:
--> but since these are different channels, this should in theory not apply here.
@lleachii , you're still being a condescending, unapologetic ass, and will be ignored henceforth. There's a difference between knowing something, and knowing how to teach that something. You may have the former, but you're lacking the soft skills for the latter. I don't deal with, and won't deal with, know-it-alls.
@ThiloteE , checking such would be feasible, if the card were easily accessible. Problem is, enterprise-grade servers are meant to be as compact as possible. Accessing the card for testing would be risky, if not downright dangerous, trying to isolate the correct pins in such a cramped space is not something I'm comfortable doing.
Now, I've been using 22.03.4 (the latest build, r20123-38ccc47687). Should I try the nightlies again to see if there's any difference? Or would there not be a net gain?
Given this is a testbed right now, I'm willing to tinker a bit to see what I get from the software. I'm probably going to get a different adapter for the card, just to rule that out, too.
I did see that submitted issue when i went searching for answers before I made this post, but since it didn't directly reference the MT7915, wasn't sure it applied. Having seen also a post stating calibration data may be missing, maybe the two are linked?
My $0.02 is that given you have the luxury of this being a testbed at the moment, go ahead and build firmware from snapshot (master) and see how things look for your issue at hand. FWIW, I run snapshot builds on my RT3200 WAPs and rarely hit major issues. I would only recommend this for technical folks like yourself who are willing and able to get hands dirty if a bug does arise.
May just do that later, see what I find. Kinda taking a day off, haven't been feeling well. Do want to get back to figuring out this card, don't want to give up on it yet. Like the idea of running a virtualized AP.
If you are able to and have the willigness and the time to compile your own OpenWrt (FAQ: how do i compile OpenWrt?), then you could try to test a few commits that are in the pipeline but have not reached OpenWrt yet.
Scrolling through the feeds, there are two I found to be very interesting and which might be relevant here:
The symptoms of an underpowered 3.3v line on your machine will be instability in the hardware. OpenWrt will not show a reduced upper-limit value that you can set a radio's power output to because the card is an a 3.3v rail that can't supply enough current. The card wouldn't know, in fact wouldn't be able to know, that the amount of current that could be supplied wasn't enough until after it was set to a high power and tried to draw that current. At that point the symptoms that would appear are dependent on the main board's design, but all revolve around the voltage on the 3.3v rail sagging. Instability on the PCI bus, on the card, and (depending on how isolated that 3.3v rail is from the other rails) possibly on the whole machine.
He wasn't doing anything of the sorts, please let's keep civil.
I'm trying to sort out the definitive symptoms that will be easiest to triage. The one that should be easiest to pin down is OpenWrt only allowing up to 23dBm power output regardless of the channel selected. Can you confirm this it the case?
He was being condescending, and I stand by that. You only saw the convo here, you did NOT see the PM from him. And I've read some of his other posts in reply to others, and honestly am shocked at his lack of tact and care in response to others' questions.
Moving on.
Yes, the hard upper limit allowed in the config seems to be 24dBm, on both SNAPSHOT and 22.04. Given your explanation (which was better than LLEACHII's, and more useful), I expected as much. Still, also given another post, I wanted to make sure that the card wasn't suffering from lack of power. Your posts in particular on the referenced forum thread are what encouraged me to seek that answer, as it did seem you were able to stabilize the card on your hardware build. So, in your experience, what are some concrete symptoms to look out for that the card may not be able to properly draw from the PCIe bus power? Understand, I'm trying to learn, but at the same time, I do not deal well with those that treat any question as stupid. There is no such thing.
Then do so. Please don't escalate further. Continue to talk about the wifi problems. By now we are talking more about how you and we want to be treated than the actual problems. It's very distracting and easy to get sucked into. I remind you of OpenWrt rules: "Please be nice to each other", which goes both ways.
You may wish to respond to something by disagreeing with it. Thatâs fine. But, remember to criticize ideas, not people. Please avoid:
Name-calling.
Ad hominem attacks.
Responding to a postâs tone instead of its actual content.
Knee-jerk contradiction.
Instead, provide reasoned counter-arguments that improve the conversation.
I personally will stop here, because this problem seems to be above my expertise and I don't think I can contribute further to solving the wifi problems
Thanks for reminding all of the forum rules. Though, I am finding it increasingly difficult to keep silent on this one. There are certain individuals within this forum that have a history of replying in such ways that are condescending and insensitive. I would even go so far as to suggest some are passively provocative, as is easy to do under the guise of âanonymityâ in closed doors behind a computer screen.
Iâve seen many a new forum member face some of the behaviors described above, at times from members of the âSolution Institutionâ level members. Not sure all the reasons, but frankly it doesnât matter. Each and every person who posts here should have a right to ask a question without having to feel as though they are dumb. We ALL were new to OpenWrt once upon a time, even it if was years ago. But ignorance of OpenWrt is no excuse to assume someone needs a link to a Wikipedia page explaining what a multimeter is. That is just rude, and some would call it insulting.
If you donât understand my reference to the multimeter link, take a look through the posts in this thread that have since been deleted (hit the orange pencil said posts have since been permanently deleted ).
Frankly, Iâm glad some of this came to light. Itâs a good reminder that even being one of the longest standing members of this forum doesnât exempt one from being a decent, non-condescending human to another human. Kudos to @netracerx for standing up for that.
I'm done here. I'm not wasting my time on this forum any longer. It's apparent there are those that care more about being the elite on this forum than actually doing good for the community they're apart of. I come back and lleachii's posted and deleted to "please stop trolling"... which is sad given his behavior. There are gatekeepers here on the OpenWRT forums, and their want to keep OpenWRT to themselves is keeping OpenWRT behind other solutions (like DD-WRT, which is now doing what I need). I want to help, I want to be a part of this community, but if I have to deal with the likes of lleachii and his supporters, you can have it. I'd rather deal with Brainslayer, at least DD-WRT works and is a much more polished product. So, have fun keeping this to yourselves. You can have it.
Mr. R - if you want help, you are welcome. I've removed every message that offend you and your colleague, despite neither of you actually saying why measureing voltage and current with a multimeter was concending. Now you are just attacking me and dragging the software into it?
Feel free to seek help.
Feel free to inquire. I just ask that you stop tagging and removing me. I've noted to moderators before of those who add and remove tags in the time frame. I've now turned off global notifications because of it.
Thanks. Hope this helps.
EDIT: Additionaly, I should note your other colleague has noted he intends to have an honest discussion with me via Private Message when time permits for him to address my manner of writing and my actual answer, I welcome you to say whatever you feel in PM to me too - in an adult and responsible manner.
).