Why it is that the short pigtails are meant for 6G? They seem to be the same kind of coaxial cable. I can’t imagine the 6G radio to be sensitive to the extra inch of length. Laptop NICs regularly have to deal with longer antenna cables. It must be for convenience of routing the wires inside of the metal enclosure.
First of all, I don’t have that much practical experience with this topic.
I haven’t made any measurements myself! In general, shorter is usually better.
“I can’t imagine the 6G radio being sensitive to an extra inch of length.”
I think it is relevant, but the question is how much of a difference it makes.
→ The decision to place all the 6G antennas on one side is something we could discuss. It seems like it might be influenced by cable length. They may have had to choose between the best antenna placement and shorter cables (because of the simple enclosure).
There has been some discussion about the different aspects of Wi-Fi (cables, antenna type, and placement).
I will not link all chats, but this thread is a good starting point for exploring the topic further:
Here’s why I doubt the antennas are the best choice for everyone:
Therefore, the “equipment” for the BPI-R4 is aimed at people who might want an easy introduction to this topic, such as programmers and end consumers. It will definitely work!
A good example of how “easy” the included equipment is can be seen in the antenna cables. They don’t even use RG178 cables…
If you want a significantly better Wi-Fi 7 router, you’ll need to build your own case with good cooling for the BE14 module, use special cables, and potentially use special antennas with optimal placement. This will take extra time and some additional money.
Following your advice, I reassembled the router with short pigtails all connected to 6G and the corresponding antennas on one side. I saw 5G figures go from 1.05 to 1.15 Gbps. 6G didn’t improve and I’m still not too happy with the throughput of 2 Gbps tops in 802.11be 320 Mhz. Although, that could be my wireless NIC (Qualcomm NCM865). Swapping antennas and cables for better ones seems promising too.
Those pigtails are cheap, on market there are better one for not much more money. I’ve mixes outputs little bit , on one side i have on sides two dual frequency with middle one 6gig, and same on other side. Besides that placement of antennas hokes aren’t good - two of ther are directly over pcb screws. Not to mention about leading those cables… Madness.
Thank you very much. BTW, the item page defaults to the SMA Female adapter on the other end, which I think would not work - we’d need SMA Male to go into the antennae.
So I read the thread on choosing antennas for BPI-R3 and decided to try ones from Antsym. Only problem is, those are 2.4 or 5.8 GHz. I could use the same 5.8 GHz antennas for 6 GHz band too as they seem to be an overall improvement over the stock ones. Still, would someone recommend a proven brand to look for 6 GHz options?
Those would be the best one, three with 10cm and three with 15cm, but this case project is bs and You have to put all cables in shitty way, i hate this .
I don’t think the bottleneck right now will be pigtails and higher dBi antennas.
We don’t even have a proper stable driver yet
If you really want top performance you should probably wait for BE19000, with 14 antennas.
But again, I doubt it will matter much unless you have dozens of concurrent users.
So I tried those antennas by Antsym (5.8 GHz, single band, allegedly 8 dB) combined with RG178 pigtails. My Tx speed (from the router to the client) has stayed roughly the same at 2.05 Gpbs on the 6 GHz band 320 MHz and 1.56 on 5 GHz 160 MHz. However the Rx speed went up from 920–930 Mbps to 1.05–1.12 Gbps on both bands. All tested with the stock BPI image in close proximity. Not a dramatic improvement by any means, but an improvement nonetheless.
Pigtail length of 10 and 15 cm was barely enough for the official metal case. I’d suggest 20 and 25 cm ones for easier routing.
This chat was established for someone who wanted to know how long the pigtails should be.
The problem is, you mostly buy the pigtails and antennas together with the BE14 board.
That’s why it’s good to think about it beforehand (If you even care about). First time “right” …
For me, it’s not about the bottleneck. And yes, BE19 is better.
What changed my mind about “one antenna for all” is this post by Dale:
Betonmischer tested it in close range:
It depends on your use case what is enough for you!
This is NOT the chat for BE14 or BE19 wifi module.
It was an aside reply to a comment posted here as I hadn’t heard anything about the BE19 until the person mentioned it so was curious. It’s as tangentially related to this topic as your comment “Sounds interesting … why do you want do build an x86_64 system? I thought x84 is dying. ” was in the thread you directed me to for the Wifi 7 Module design. That’s kind of how discussions work.
If you really want top performance you should probably wait for BE19000, with 14 antennas.
But again, I doubt it will matter much unless you have dozens of concurrent users.
It was an aside reply to a comment posted here as I hadn’t heard anything about the BE19 until the person mentioned it so was curious. It’s as tangentially related to this topic as your comment “Sounds interesting … why do you want do build an x86_64 system? I thought x84 is dying. 
” was in the thread you directed me to for the Wifi 7 Module design. That’s kind of how discussions work.