I think the 10 antenna would be a better compromise
When can we expect the board to release for sale? Or is there a way to get one of those samples?
Since Chinese New Year is coming up next week, I guess we will have to wait at least until mid of February.
From what I heard on Amazon it’s coming around April iirc
Hoping to hear more about the development of the 10-antenna version rather than this downright disappointing version of the product.
It really feels like a bait-and-switch that Banana PI ran to the press/media with 14 antenna version 4x5 setup to get people to buy BPI-RPs, when we got a last-minute and quiet switch to a 3x3 6-antenna design. Why is the SKU called “BPI-BE14” instead of “BPI-BE6” since it no longer has 14 antennas?
I will continue to hold off purchasing (now validated to hold off) a BPI-R4 until the situation has been rectified by getting the full 4x5 design for the 5 GHz and 6 GHz bands with a 10-antenna design.
Not sure what you try to achieve with this “hold off purchasing because i want 14 antenna”. If you think that raises the fear of losing you as customer, … i dont think so. if the 14 antenna version is the only reason for you to buy a banana pi r4, maybe get an AP then.
1 Like
Because a 14/10 antenna version would be better than most APs? Ubiquity for example right now is only shipping out 2x2 which would still be worse ironically compared to this setup. I just would prefer to have the original or more fully-featured design and go out with a killer setup as they originally showed off to the press.
But to your point, yes, I could buy an AP as the Wifi 7 AP market matures if the BPI team chooses not to do a 10-antenna version. But the value would probably be there, and I wouldn’t mind tinkering with OpenWRT (more for roaming) if it means I’m dealing with essentially best-in-class hardware. Wouldn’t it be nice to have that original full-fat 4x5 setup on 5 and 6 GHz bands at least as an option? It seems like 2/3rds of the community does based on the polling from the other thread…
Why run out and show a 4x5 setup if the end product you’re going to ship to customers is cut down? That’s all I’m asking. I was excited about what the original product would’ve been and was waiting for it. I think it’s understandable why one could be disappointed to see the actual product is quite different after 6+ months of waiting with essentially a different product in mind.
1 Like
I’m afraid it is because for most 14 antennas is pretty unwieldy and they do not look forward paying more for antennas than the board itself.
Simon said they’ll look into 10 antenna version after they fix thermal issues and openwrt one
ok, 14 antennas would technically be better.
but can someone please tell me or point me to a link which shows the difference in
speed, performance, stability,qos, amount of clients
between 6 and 14 antennas? because I actually think most people just want more antennas without any actual benefit.
1 Like
this device is still hard a work in progress. the software is not yet finished at all. this is not an end-user-device yet. it is still hard in work. if there will be a 6/10/14-antenna version, no one knows, as long as there is no concrete statement from sinovoip. when they marketed it as 14-antenna-device, thats bad, if there wont be any. but tbh, i guess there might be something coming, but probably not in the beginning. when sinovoip sees the requests and that there is money to earn, they probably will do it. if not, then probably not. crying in the forum doesnt help. contact them, tell them your concerns and see what they answer. everything else is just wasted resources. and it annoys me to read stuff like “im not going to buy XYZ when ABC is not included”. I did such stuff too, when i was 12 or so. it wont help you nor anyone else.
Well, if we go back to MediaTek’s blog post from June 2023 when they “partnered” with Banana Pi to develop a 4T5R module with the Filogic 880 (the original 14-antenna version).
MediaTek’s unique 4T5R platform allows for improved Receive Diversity and supports MRC (Multiple Receive Combining). This boosts 6GHz band throughput in typical indoor scenarios and extends maximum range, assuring incredible performance as well as more reliable and effective mesh networking when 6GHz is used as backhaul.
This was supposed to be the unique selling point to buying a BPI-R4 was that we were supposed to have asynchronous antenna diversity on the receiving end to boost range and “performance”. How does this perform in reality? Well, we need to have a 14 or 10-antenna version (likely the latter since it uses antennas for 5 GHz and 6 GHz reducing complexity compared to 14 antennas) to see how it actually works.
So yes, it was marketed as a “14-antenna board”, even by MediaTek, and like I said before the SKU name still is “BPI-BE14” (as per the title of this thread) instead of “BPI-BE6” which is really what this should be called now.
1 Like
The name does not reflect the number of antennas. The original board was called BE19 Banana Pi BPI-R4 Wifi 7 Module design
1 Like
Ahh c’mon, this board is for enthusiasts , all was known from the beginning . If there be 10 or 6 anthennas then who cares? Not intressted? Then buy a common router or ap of another company . For me is fun to play with it and go beyond other people use case for it. The cheapest competitor is tp-link be900 and it.cost four times this board and ans history ahowes with qualcom 807* this platform is still not matured to work with hw offload and good performance.
I do understand your point of view, but I’m really wondering what would be the real difference except from showing of. if the diagram I see some Diplexers, would that mean the antennas swapping frequency all day, because that’s not good. but if it would mean I can have multiple frequencies using the same antenna with (barely) any performance loss or something I’m all for because it means my device is less complex to install.
I myself haven’t bought a BPI-R4 yet, I’m waiting for the wifi board and case design to be ready. So when I start with it, I can immediately play with all the features.
1 Like
Hello Bart_K,
you speek my mind: speed, performance, stability, qos, amount of clients.
And additionally:
heatsink, power consumption, temperature overview, planned release date, price, dxf files …
→ all of this points compared with the taller Wifi 7 board (10/14 antennas). I’m a very curious guy and I would like to understand the pro points which speaks for the board. I like this litte board, but this is not a scientific view. I’m here since over a year, some days ago, for the first time I had to report a post. It was about this board (the words are poor language). It shows me, that this topic causes emotions! I think understanding the other site is the key for acceptance!
Generally:
The board is signd with V00. The bigger boad has also V00. I would not sell my soul, but may flirting with the devil to know where we are. Do V00 mean prototype?
Since the information about “OpenWrt One/AP-24.XY” are out, I’m in love with the MMCX connector:
Shortening the cable is also on an U.FL cable possible. This could be done on the SMA side. But the MMCX is more reliable.
I know, there are many points against it. But also points which speaks for it (especially DIY)! For the BPI-R3 I would not mentioned it, becase of the limitet space. But now with the additional board the situation changed.
So maybe next time 
?
Board Design:
I’m hoping Sinovoip stays at the concept of an external WiFi Board. And I mean not because of “we have to do it” there is no space on the main board! I thought I had read on the OpenWrt One/AP-24.XY - project, that the Wifi chips should be on the main board (like a roof should be on a car). As a reliability test guy, i know how bad it could be to have contact problems. As this guy I have to say → bigger main board and all components soldered together 
.
But as a maker i want to have a choice 
!
I congratulate for staying at the same main board size like the other router boards 
.
What i whished that card would be:
Last days I read much about pure OS, lineageOS and the fairphone. Comparing the different strategies → to make the world better in different aspects.
But beeing relistic, could the next future WiFi cards be the new part of the Banana Pi business strategy: → reducing development costs because of component reusability for instance for R4 AND R4 mini.
→ I’m talking about a sandwitch design with space between the layers for cooling! I not only mean the “mini” → I also mean the big router models.
→ layers/cards for a fan, RJ45 expansion, wifi …
A major point on R3 mini, OpenWrt One/AP-24.XY and R4 are the amount of RJ45 connections. A additional board with 3 (mini) or 6 more RJ45 jacks would make a difference and could reused for the next router board generations…
It is just a suggestion!
The BPI-R3 mini was a huge step in case design. The end customer was the target? But what makes R3 mini so special? It’s just a rhetorical question (I like Mediatek combined with OpenWRT). Could more w-Lan antennas or an additional SFP cage change the game? I think so (esspacially for the SFP cage, because of Internet over fiber optics). The SFP was a breakthrough for BananaPI. If you get this in a little square box than you have something which is more special. And like the wifi board, it could be recycled in the next generation.
The compeditors in this area are :
Summary: → I’m a fan of MMCX antenna sockets and because of recycling I would like as much as possible single components witch are arranged one above the other.
April for a 6 antenna board? Close to pointless for me. Too little too late.
@howardthedm Welcome to the reality of design, manufacture, certification & business. There is no such thing as a magic wand.
1 Like
@sinovoip,have you got driver correctly loaded? I get a timeout before the 3rd wifi interface should come up (2 i see in ip a). I guess we need firmware for the second frontend mt7977IA for this to solve.
Chinese new year… Answer may take some time …
I suppose you have to rename mt7996_eeprom_233.bin to mt7996_eeprom.bin to have the driver pick that up instead. In the end this should be done using dt-overlay again with distinct dtbo for each wifi module type (BE14 vs. BE19) needing user-supplied EEPROM data.