What's new

Radio engineers: Why doesn't the AX88U Pro's antenna array outperform the AX86U Pro's?

  • SNBForums Code of Conduct

    SNBForums is a community for everyone, no matter what their level of experience.

    Please be tolerant and patient of others, especially newcomers. We are all here to share and learn!

    The rules are simple: Be patient, be nice, be helpful or be gone!

snovvman

Regular Contributor
From this layperson's perspective, for the purpose of distance vs. throughput speed, why doesn't the AX88U Pro outperform the AX86U Pro? The AX88U Pro has four external antennas that are spaced farther apart when compared to the AX86U Pro's three external and one internal spaced closer together. I thought that diversity antennas would work better with the antennas farther apart (like the 88) and thereby leading to better radio performance, especially at distance. With all the speed comparisons I can find, the 86 performs just as well if not better than the 88. I know radio engineering is a lot more than just distance between antennas, it also depends on the client, but I'd like to better understand specifically how the antenna designs relate to performance and range. Curiosity... Thank you.
 
Last edited:
Unless you found a review comparing the two models in identical environment, with identical client devices, and also tested back-to-back, you have not seen any speed/distance comparisons worth using.

The GT-AX6000 is superior to the RT-AX86U (orig), The RT-AX88U Pro is superior to the GT-AX6000, in my testing, with better throughput at the same distance with the same clients.

While the RT-AX86U Pro may be similar, I have a feeling the RT-AX88U Pro has the edge (but I have not compared it directly to the RT-AX86U Pro).
 
Tested clients are almost always only two streams, so they generally only need to talk to two of the antennas anyway. More antennas will only help in very specific scenarios, as you might get signal bouncing in different directions if in a challenging scenario with lots of obstacles. It won't help with direct line of sight.

The antenna distance difference is negligible between the two routers, they will both radiate in the same doughnut shaped pattern.
 
Tested clients are almost always only two streams, so they generally only need to talk to two of the antennas anyway. More antennas will only help in very specific scenarios, as you might get signal bouncing in different directions if in a challenging scenario with lots of obstacles. It won't help with direct line of sight.

The antenna distance difference is negligible between the two routers, they will both radiate in the same doughnut shaped pattern.

Thank you, @RMerlin. I was wondering about the AX88U's 4 antenna diversity for 2.4 and if that would provide an advantage over the AX86U's 3, even if it is slight. I also noticed that the AX88U's antennas are about an inch longer than the 86. But as @Tech9 pointed out, they are just "gaming plastic".
 
But as @Tech9 pointed out, they are just "gaming plastic".

Get the router with most antennas, possibly the longest ones, preferably with color accents and LED lights. They are the best ones according to the marketing. Engineers use calculators and Wi-Fi antennas used in consumer products are typically 1/2 wavelength. You can do the math yourself and it will also give you an answer why your mobile phone doesn't have sticking out antennas and still communicates with the cell tower much further away and through walls. You can look for Starlink antenna internal photos as well, it communicates with satellites 340-620km away. And this thing is a 16dBi antenna inside Ubiquiti loco M5 bridge, it can shoot 5GHz signals kilometers away:

1716667223188.png


Home routers don't need external antennas, but they obviously sell the product well. I call it "Honda Civic Type R" networking with sticker tuning on top.
 
Get the router with most antennas, possibly the longest ones, preferably with color accents and LED lights. They are the best ones according to the marketing.

Lol they sure do move products with all that bling!

Engineers use calculators and Wi-Fi antennas used in consumer products are typically 1/2 wavelength. You can do the math yourself and it will also give you an answer why your mobile phone doesn't have sticking out antennas and still communicates with the cell tower much further away and through walls. You can look for Starlink antenna internal photos as well, it communicates with satellites 340-620km away. And this thing is a 16dBi antenna inside Ubiquiti loco M5 bridge, it can shoot 5GHz signals kilometers away:

View attachment 58980

What a great explanation and puts things in perspective. Cheers!
 
I did actually open some home router antennas and all of them were just a dipole as cheap as possible. For 2.4GHz the length is about 6cm, for 5GHz about 3cm. The best engineered antennas so far I have seen were on older TP-Link routers with flat antennas like this Archer AC2600:

1716667917301.png


Inside is a printed PCB 5dBm antenna like this:

1716668019638.png


...and some had excellent range, better than most.

Just black, no color accents, no LED lights... no BS sticker tuning. Just works well for the application.
 
Yes, there are better antennae and worse antennae, but unless you are building your own RF design, that is not important.

What is important is the RF design/hardware you can get/buy in total.

And Asus is above average there.
 
One of the more interesting internal antenna designs - this one shows a fair amount of RF and Antenna design expertise...

I present the AirPort Extreme AC from Apple... this was after the iPhone 4 antennagate mess, and was a result of the consolidation of all the RF teams under one roof...

This one is pretty interesting, as you can see the dual-band PIFA's, along with the shield that actually plays into the overall pattern.


Screenshot 2024-05-26 at 5.29.11 PM.png


I know a couple of guys that were part of that team back in the day - Apple threw a metric frak-ton of resources at that team, building out antenna chambers, hiring many RF folks, and made it a priority across their entire product lineup - from iPhones, to Macs, iPads, AppleTV, and of course, the Airport lineup.

While Apple has exited the WiFi AP game, that team is still there, working across product stack...
 
Last edited:
Yes, there are better antennae and worse antennae, but unless you are building your own RF design, that is not important.

RF design work is under appreciated - it's the total product, not just the antennas - it's everything from routing the feedlines on the circuitboard to shielding, and ultimately impacts the overall product design...

Having been there - that's why I usually suggest to leave the antennas along, as they are part of an engineered solution, and that's also a solution that comes into play with regulatory approval efforts.
 
I think the question being asked is wrong given there are people here that have tested the AX88U PRO and had it outperform the AX86U PRO in the specific environment tested. Not that either hold a crown, but one is out performing the other due to whatever obstacles and interference exist at the location.. Even with shared 4x4 MIMO config.

There is also a factor that hardware is tested and certified with different sets of amplifiers per local regulation, but I wouldn't bother trying to find the best absolute sample.. something will always be better.

Either way.. I would recommend the highest MIMO config just for beamfoaming purposes. Should be 4x4 for current BE gen.. 4x5 if you consider Mediatek.

8x8 configs are likely dead unless you buy a last gen QCA Pro networking 1200 platform which had teamed 4x4 + 4x4 on 5ghz. Current 1220 will be config'd as Tri-band with MLO for BE clients.
 
Get the router with most antennas, possibly the longest ones, preferably with color accents and LED lights. They are the best ones according to the marketing. Engineers use calculators and Wi-Fi antennas used in consumer products are typically 1/2 wavelength. You can do the math yourself and it will also give you an answer why your mobile phone doesn't have sticking out antennas and still communicates with the cell tower much further away and through walls. You can look for Starlink antenna internal photos as well, it communicates with satellites 340-620km away. And this thing is a 16dBi antenna inside Ubiquiti loco M5 bridge, it can shoot 5GHz signals kilometers away:

View attachment 58980

Home routers don't need external antennas, but they obviously sell the product well. I call it "Honda Civic Type R" networking with sticker tuning on top.
Actually, this isn't entirely true and I know this as someone that has been heavily involved in the development of a router with built-in antennas.
First of all, you end up with issues if the antennas are too close to the main PCB of the router. These range from the antennas not working at all, to really really poor range. This is why most routers (and phones) have the antennas placed as far away from the main PCB as possible. Remember the iPhone that you could only hold a certain way if you wanted good reception?
Also, depending on how you place the antennas inside a router, you can end up with vastly different results. We ended up moving one antennas about 1 cm internally and went from crap to great signal.
As such, it's much easier to use external antennas and they often cost less to produce as well, despite all the plastic and what not, as it's just a length of antenna wire or metal wire inside those fancy looking plastic sticks (as you pointed out further down), whereas PCB antennas are just that, PCBs, which requires a more costly process to produce, even if it's only marginally more so. The cheapest antennas are PIFA antennas, but they also have among the worst performance and are harder to make for dual- or mutli-band devices.
One of our engineers actually wanted to experiment a bit and made 100 PCB antennas based on sound antenna design guidelines, only two were good for the 2.4 GHz band and neither of them performed as good as the antennas we were buying from a company that specialises in making antennas.
So yeah, external antennas aren't "needed" as such, but both the antenna design and placing them right inside the router really matters.
 
There is also a factor that hardware is tested and certified with different sets of amplifiers per local regulation, but I wouldn't bother trying to find the best absolute sample.. something will always be better.
Unless something has changed, this is not true. These days the PA and especially LNA is often built into the radio, so there's no way to change them. Even so, no router maker would offer multiple hardware SKUS with different PA's, that would making routers prohibitively costly due to how many different combinations of components would be needed. Instead, this is all controlled in software/firmware and not on a hardware level.
 
Unless something has changed, this is not true. These days the PA and especially LNA is often built into the radio, so there's no way to change them. Even so, no router maker would offer multiple hardware SKUS with different PA's, that would making routers prohibitively costly due to how many different combinations of components would be needed. Instead, this is all controlled in software/firmware and not on a hardware level.

I meant FEM's.

AX86U (non pro) for example has multiple revisions of hardware since original release. Even down to 2.5G controller changing mid cycle.

ASUS as a company also tends to build routers in multiple locations. Vietnam, Taiwan, China.
 
which requires a more costly process to produce

Cheaper is always the preferred manufacturing option for consumer products. External antennas though also sell better to non-tech people, design marketing involved. Bigger antennas and more in numbers is seen as better, more powerful. Accents and LED lights - more gaming power! 💪
 
I meant FEM's.

AX86U (non pro) for example has multiple revisions of hardware since original release. Even down to 2.5G controller changing mid cycle.

ASUS as a company also tends to build routers in multiple locations. Vietnam, Taiwan, China.
Yeah, the FEM is made up by PA's and LNA's, but again, if you have a look, they're not region specific.

Changing components is nothing strange, most of this has to do with supply and if you remember, there was the pandemic and a lot of components ended up in short supply, so the hardware manufacturers had to source alternatives.

And this has exactly what to do with the matter of you claiming that WiFi routers are region locked by the hardware?
As this is simply not true. Also, Asus does not make routers in Taiwan any more. There are almost no companies that can produce routers in Taiwan today, largely due to the very expensive testing equipment no longer being located in Taiwan due to production having moved elsewhere and no-one wanting to invest half a million US$ in new test equipment.
 
Yeah, the FEM is made up by PA's and LNA's, but again, if you have a look, they're not region specific.

Changing components is nothing strange, most of this has to do with supply and if you remember, there was the pandemic and a lot of components ended up in short supply, so the hardware manufacturers had to source alternatives.

And this has exactly what to do with the matter of you claiming that WiFi routers are region locked by the hardware?
As this is simply not true. Also, Asus does not make routers in Taiwan any more. There are almost no companies that can produce routers in Taiwan today, largely due to the very expensive testing equipment no longer being located in Taiwan due to production having moved elsewhere and no-one wanting to invest half a million US$ in new test equipment.
I never claimed region locked. I'm saying the FEM may influence overall performance per design. Theres 2-3 different layouts in the current US FCC revision.

I've had 2 AX86U (non pro) models here with completely different throughput over wireless when placed in same location.

When AX58U launched, the Best Buy variant exclusively came from Taiwan only. Everyone else in US got the "RT-AX3000" marketed model (Same FCC ID) but supplied from Vietnam. Europe's supply was Chinese.

When AC86Us were being produced in China and then Vietnam, The special GT-AC2900 also only came from Taiwan for the US market..

2 different GT-AX6000's I've tried (subjectively had subpar throughput in my home) were also manufactured In Taiwan.. and this is relatively recent model. I was hoping the first one was defective.. Both Amazon purchased.

Saying ASUS doesn't produce hardware in Taiwan is just straight out false.. Might be the case for Europe.. but there are plenty of examples of US spec models being produced there.

Edit: I wouldn't be surprised if the "RT-BE7200" Best Buy SKU is exclusively Taiwanese made. Isn't marketed under the RT-BE88U moniker.
 
Last edited:
I never claimed region locked. I'm saying the FEM may influence overall performance per design. Theres 2-3 different layouts in the current US FCC revision.
So what does this mean then? "There is also a factor that hardware is tested and certified with different sets of amplifiers per local regulation".
That reads like hardware region lock to me.

I've had 2 AX86U (non pro) models here with completely different throughput over wireless when placed in same location.
I'm not at all surprised, the routers we made also had slight variances in performance, since WiFi isn't an exact science and there are variations in the parts that make up a router.
When AX58U launched, the Best Buy variant exclusively came from Taiwan only. Everyone else in US got the "RT-AX3000" marketed model (Same FCC ID) but supplied from Vietnam. Europe's supply was Chinese.
That doesn't correspond to the FCC data which lists it as made in the PRC. That said, the DUT that was sent in by Asus says Made in Taiwan. Asus would've had to submit more test data to the FCC to be allowed to sell the different SKUs if they were made in different countries. Likewise, you have to submit permissive changes to the FCC if you change components. That said I still don't believe they're made in Taiwan, but rather assembled in Taiwan. The PCB would've still been made in the PRC and the routers assembled in Taiwan, as Asus simply doesn't have any router production in Taiwan these days.
When AC86Us were being produced in China and then Vietnam, The special GT-AC2900 also only came from Taiwan for the US market..
This one is old enough to have been made in Taiwan, at least some of them.
2 different GT-AX6000's I've tried (subjectively had subpar throughput in my home) were also manufactured In Taiwan.. and this is relatively recent model. I was hoping the first one was defective.. Both Amazon purchased.
I have a really great 5 GHz band on mine, but the 2.4 GHz radio is meh in comparison, which is rather peculiar. Again, assembled in, not made in. This is common practice in a lot of countries, like the US for example, where components are imported and them assembled to a finished product.
Saying ASUS doesn't produce hardware in Taiwan is just straight out false.. Might be the case for Europe.. but there are plenty of examples of US spec models being produced there.
Yeah, sure... I've only lived in Taiwan for over 14 years of my life and worked both as a tech journalist and at several hardware manufacturers in Taiwan and nearly at Asus for that matter, but that's another story. I think I have some insight into the local router board makers. But again, as above, I'm sure they assemble them in Taiwan.
 
So what does this mean then? "There is also a factor that hardware is tested and certified with different sets of amplifiers per local regulation".
That reads like hardware region lock to me.


I'm not at all surprised, the routers we made also had slight variances in performance, since WiFi isn't an exact science and there are variations in the parts that make up a router.

That doesn't correspond to the FCC data which lists it as made in the PRC. That said, the DUT that was sent in by Asus says Made in Taiwan. Asus would've had to submit more test data to the FCC to be allowed to sell the different SKUs if they were made in different countries. Likewise, you have to submit permissive changes to the FCC if you change components. That said I still don't believe they're made in Taiwan, but rather assembled in Taiwan. The PCB would've still been made in the PRC and the routers assembled in Taiwan, as Asus simply doesn't have any router production in Taiwan these days.

This one is old enough to have been made in Taiwan, at least some of them.

I have a really great 5 GHz band on mine, but the 2.4 GHz radio is meh in comparison, which is rather peculiar. Again, assembled in, not made in. This is common practice in a lot of countries, like the US for example, where components are imported and them assembled to a finished product.

Yeah, sure... I've only lived in Taiwan for over 14 years of my life and worked both as a tech journalist and at several hardware manufacturers in Taiwan and nearly at Asus for that matter, but that's another story. I think I have some insight into the local router board makers. But again, as above, I'm sure they assemble them in Taiwan.

Well it says "made in". Maybe ASUS should update the sticker and packaging SKU? 😉 I don't doubt that Arcadyan moved out, but any possible low production runs?

And it means what I said.. I'm saying there is likely different HW sets depending on manufacturing date or supply.. Multiple FCC changes. The performance in my case was significant... Even vs the GT-AX6000 when compared to one specific unit.

Doesn't matter much now since I have a QCA 1220 design out performing everything else here via 5G.

Edit: Thinking about it, the Taiwanese GT-AC2900 had a slightly different top cover design vs Chinese model..

I'd assume ID molds are actually made in Taiwan and loaded with Chinese manufactured electronics assuming there is no production there.. I guess "Made in" would be technically be half correct.
 
Last edited:

Sign Up For SNBForums Daily Digest

Get an update of what's new every day delivered to your mailbox. Sign up here!
Top