EnGenius ECW520 Cloud7 2x2x2 Test and Review

The EnGenius ECW520 is one of the most affordable WiFi 7 access points that the company currently offers and its aim was to offer enterprise-level features, while keeping the price tag underneath $200. And that’s perfectly fine, but we do have the 2x2x2 ECW526 available as well and recently, EnGenius has also released the affordable Cloud Lite ECW516L, so where does the ECW520 position in the market?

engenius-ecw520-access-point
Engenius ECW520 WiFi 7 Access Point

The Cloud Lite access points are obviously more limited both in hardware abilities and in terms of software support, but the ECW520 and the ECW526 are incredibly similar. Both rely on the Qualcomm platform, the antenna gain is similar, with the former actually being a bit better on the 2.4GHz radio, but there is a major difference in terms of port. The ECW526 comes with a 10GbE PoE+ port, while the newer ECW520 is equipped with a 2.5GbE PoE+ port.

EnGenius ECW520
Zyxel.com Check Device

The three radios (including the 6GHz) are offered on both access points and so are most of the newer WiFi7 features, so there’s quite a bit of overlapping between these two devices. Then again, there may be more fundamental differences that are less than skin-deep, so let’s take a closer look at the ECW520 and see how it perform in relation to its main competitors.

The Design and Build Quality

In terms of design, the EnGenius ECW520 is closer to the ECW130, than the newer, slimmer models. Sure, it’s a bit thick, but it’s also far smaller, standing at 15.8 x 15.8 x 3.5cm.

engenius-ecw520-wifi7-access-point
The size of the EnGenius ECW520.

The access point does come with a bracket, so it’s easy to mount it on the ceiling, but there are also a couple of holes within the metallic alloy bottom, so it’s possible to wall-mount the device as well. Also, similarly to the other EnGenius access points, the top is plastic, so the metallic bottom takes the role of dissipating the heat away from the components. So far, it has proven to be a reliable system. As for LEDs, there’s a single one which is far from being intuitive. It’s a combination between green and blue to let you know the status of the AP..

engenius-ecw520-ports
The ports area.

Anyway, on the rear side, we get to see the ports area where’s we can find the recessed Reset button, a 12V DC port and a single 2.5GbE PoE+ port.

The Thermal Management

Now let’s talk about the heat management of the access point. I did open up the device and, while compact, EnGenius appeared to have made the ECW520 fairly efficient. But this needs to be confirmed by a thermal camera.

And, after running a few tests to simulate a resource-heavy environment, the access point does get quite hot going above 52C (125F), so we may see some sort of throttling if there is a lot of constant pressure put on the network.

engenius-ecw520-thermal
The thermal management.

EnGenius ECW520 Teardown

As usual, it’s very simple to open an EnGenius access point, so in case you need to remove dust or repair the device, the manufacturer made the process simple. All I had to do was remove the four screws from the bottom of the device and the top part will pop off easily. Then, we can see the antenna array and, after taking out the screws, we can detach this part as well, but not before the six antennas are removed as well.

engenius-ecw520-pcb
EnGenius ECW520 Teardown.

And it’s not as easy as I expected it to be. That’s because the antenna connectors are glued to the board for some reason. It’s not easy to detach them, so I don’t really see the point, but the engineers thought otherwise. In any case, we can see the storage chipset on the bottom of the PCB, while the rest of the main components can be found on the top side, underneath the aluminum covers. As a side note, the EnGenius ECW520 has similar hardware to the Xiaomi BE6500 Pro.

Hardware Comparison

EnGenius ECW520 EnGenius ECW516L Zyxel NWA130BE Zyxel NWA110BE
CPU quad-core 1.5GHz Qualcomm IPQ5322 tri-core 1.8GHz Mediatek MT988DV (A73) quad-core 1.5GHz Qualcomm IPQ5322 (Cortex-A53) quad-core 1.5GHz Qualcomm IPQ5322 (Cortex-A53)
RAM 1GB DDR4 Nanya NT5AD512M16C4-JR 1GB DDR4 SDRAM Nanya NT5AD512M16C4-JR 222801NOEF 1GB Micron 3UR77 D8BPK (DDR4) 1GB Micron (41R77 D8BPK) DDR4
Storage 256MB NAND Winbond (25N02KWZEIR2434) 256MB NAND MXIC X241952 MX35LF2GE4AD-Z43/ 128MB serial NOR Winbond 25Q128JVSQ 2251 512MB (MXIC X233662 MX35UF46E4AD-241)/ 16MB NOR flash (MXIC MX25U12832F) 256MB NAND Winbond 25N02KWZEIR
Switch Qualcomm QCA8081 PHY Mediatek MT7995AV Qualcomm QCA8385 Ethernet PHY Qualcomm QCA8385 Ethernet PHY
6GHz Radio Qualcomm QCN6274 802.11be + 2x 50083714 + 2x 42674013 modules Mediatek MT7976QN2415-BWELBPTA4836 3×3 Qualcomm QCN6274 802.11be 2×2:2 Qualcomm QCN6274 802.11be /td>
5GHz Radio Qualcomm IPQ5322 802.11a/b/g/n/ac/ax 3×3 Mediatek MT7976QN2415-BWELBPTA4836 4×4:4 Qualcomm QCN6274 802.11be 2×2:2 Qualcomm QCN6274 802.11be
2.4GHz Radio Qualcomm IPQ5322 802.11b/g/n/ax 2×2 Mediatek MT7977IAN 802.11b/g/n/ax 2×2:2 Qualcomm IPQ5322 802.11b/g/n/ax 2×2:2 Qualcomm IPQ5322 802.11b/g/n/ax 2×2:2

Single Client Tests (6GHz & 5GHz)

I couldn’t help but compare the throughput I managed to record testing the EnGenius ECW520 with the Cloud Lite ECW516L. And that’s because both these access points come with similar price tags.

engenius-ecw520-test1
EnGenius ECW520 – Wireless Test – 5GHz and 6GHz – WiFi 7, WiFi 6 and WiFi 5 client devices – Upstream

From the looks of it, the ECW520 offers a better throughput near the client, but after going a bit farther away, even at 5 meters away, the performance is quite similar.

test1
EnGenius Cloud Lite ECW516L – Wireless Test – 5GHz and 6GHz – WiFi 7, WiFi 6 and WiFi 5 client devices – Upstream

And that’s true both upstream and downstream. Notice how these two access points failed to maintain a connection father than 20 meters (or about 70 feet).

test2
EnGenius ECW520 – Wireless Test – 5GHz and 6GHz – WiFi 7, WiFi 6 and WiFi 5 client devices – Downstream

Things do change slightly after checking the signal attenuation graph.

test3
EnGenius ECW520 – Wireless Test + Signal Attenuation (-dB) – 5GHz and 6GHz – WiFi 7, WiFi 6 and WiFi 5 client devices – Upstream.

The WiFi 7 client is the winner when near the access point and the loser at the higher-range tests. The other clients did fairly well and it was interesting to see how lowering the channel bandwidth on the 6GHz radio displayed a better throughput just a bit farther from the client.

test4

Again, this remains true both upstream and downstream. When compared to other wireless access points, the ECW520 sits between the ECW230S and the FIT 4×4 Lite, as long as the 5GHz radio is used set at 80MHz.

test5
EnGenius ECW520 vs other wireless APs – 5GHz, 80MHz.

After switching to the 160MHz width, the ECW520 sits between the ECW536 from the same family of access points and the U6 Pro.

test6
EnGenius ECW520 vs other wireless APs – 5GHz, 160MHz.

When using the 6GHz radio and the 320MHz channel width, the access points sits between the U7 Pro Max and the U7 Pro.

test7
EnGenius ECW520 vs other wireless APs – 6GHz, 320MHz.

Lastly, I decided to include a longer-term performance graphic where we can see some curious fluctuations, especially when using the 5GHz radio set to the 160MHz.

test9
EnGenius ECW520 – Long-term performance.

Could they be heat-related? Not sure, but the access point definitely runs hotter than other devices from EnGenius.

Single Client Tests (2.4GHz)

Now let’s switch to the 2.4GHz radio performance. We can see that it’s a decent throughput using either a WiFi 6 or a WiFi 5 client device and we do get usable throughput even when the distance is above 20m from the access point.

test10
EnGenius ECW520 – Wireless Test – 2.4GHz – WiFi 6 and WiFi 5 client devices – Upstream.

The signal attenuation graph simply enforces the previous results as we see close to 40Mbps even when the attenuation is -88dB.

test11
EnGenius ECW520 – Wireless Test + Signal Attenuation (-dB) – 2.4GHz – WiFi 6 and WiFi 5 client devices – Upstream.

When compared to other access points, the EnGenius ECW520 sits between the TP-Link EAP773 and the EAP670 which is a fair performance, especially since it’s better than what I saw with the ECW512L.

test12
EnGenius ECW520 vs other wireless APs – 2.4GHz, 40MHz.

The Multi-Client Tests

I don’t test networking hardware that often anymore, but I still preferred to go fairly in depth when testing the EnGenius ECW520, so I once again relied on net-hydra, a set of tools developed by Mr Jim Salter and which are available for anyone on GitHub. The initial test involves simulating 1080p streaming on five client devices.

And yes, only a couple of them are identical, the WiFi 6 laptops, but I suppose a diverse cast may show a more realistic performance that can be expected in most households.

test13
EnGenius ECW520 – Multi-client stress test using 5 client devices – Concurrent 1080p streaming (5GHz, 80MHz). The lower the score, the better.

The results show that one WiFi 6 client and the WiFi 5 ZimaBoard 832 performed the worst, while the MacBook Pro did a bit better, the last two clients doing decently well. Not perfect, but usable.

test14
EnGenius ECW520 – Multi-client stress test using 5 client devices – Concurrent 4K streaming (5GHz, 80MHz). The lower the score, the better.

Switching to the 4K streaming graphic, we can see a similar performance across the same clients, with two clients doing well, while the others showing unusable latency.

test10
EnGenius Cloud Lite ECW516L – Multi-client stress test using 5 client devices – Concurrent 4K streaming (5GHz, 80MHz). The lower the score, the better.

If I were to compare it to the ECW512L, it’s still well better considering that all client devices went above 100ms pretty much from the beginning. Now, I decided to include intense browsing and run it alongside the initial 1080p streaming.

test15
EnGenius ECW520 – Multi-client stress test using 5 client devices – Concurrent 1080p streaming and intense browsing (5GHz, 80MHz). The focus is on the 1080p streaming.

It’s obviously better to limit the number of clients if this is something that often happens in your network, but it’s still good to know that we do get sub-100ms latency on three clients for about 95% of the time. Not ideal, but usable to a certain extent.

test16
EnGenius ECW520 – Multi-client stress test using 5 client devices – Concurrent 1080p streaming and intense browsing (5GHz, 80MHz). The focus is on the intense browsing.

The intense browsing graphic shows that one client is pushed close to the limit, but the rest behave nicely. Moving forward, I ran the intense browsing traffic alongside 4K streaming and I got the following results.

test17
EnGenius ECW520 – Multi-client stress test using 5 client devices – Concurrent 4K streaming and intense browsing (5GHz, 80MHz). The focus is on the 4K streaming.

The 4K streaming graphic shows that it’s not really a good idea. Sure, we get to see a latency near 150ms for the four clients that are near the access point, but it’s not really usable. Still, once again, this is a better performance than what I saw with the ECW512L.

test13
EnGenius Cloud Lite ECW516L – Multi-client stress test using 5 client devices – Concurrent 4K streaming and intense browsing (5GHz, 80MHz). The focus is on the 4K streaming.

That being said, let’s also download a 10MB file continuously, but limit the number of clients to five, each with its own type of simulated traffic.

test18
EnGenius ECW520 – Multi-client stress test using 5 client devices – Concurrent 4K streaming and downloading (10MB files) + intense browsing (5GHz, 80MHz).

Of course, I started with a couple of downloading clients, one staying near 300ms most of the time, the other going to 600ms almost immediately and staying there for the entire duration of the test. There is impact on the other clients as well, rendering them unusable so I decided to limit the number of downloading clients to one.

test20
EnGenius ECW520 – Multi-client stress test using 5 client devices – Concurrent 4K streaming and downloading (10MB files) + intense browsing (5GHz, 80MHz).

And the results are peculiar. Sure, the downloading client stood near of below 250ms most of the time, but the two 4K streaming clients had a sharp rise for 5% of the time. The intense browsing client had a high latency, but within reasonable limits.

test16
EnGenius Cloud Lite ECW516L – Multi-client stress test using 5 client devices – Concurrent 4K streaming and downloading (10MB files) + intense browsing (5GHz, 80MHz).

As a comparison, it’s not that much different than the ECW512L’s performance.

test21
EnGenius ECW520 – Multi-client stress test using 3 client devices – Concurrent 4K streaming and downloading (10MB files) + intense browsing (5GHz, 80MHz).

Next, I limited the number of clients to three, but downloading a 10MB file is still too taxing, so I decided to download a 1MB file and switch the 4K streaming to VoIP.

test22
EnGenius ECW520 – Multi-client stress test using 3 client devices – Concurrent VoIP and downloading (1MB files) + intense browsing (5GHz, 80MHz).

And the result is actually decent. Of course, I also included the latency when all five clients were downloading a 10MB file continuously just for fun.

test23
EnGenius ECW520 – Multi-client stress test using 5 client devices – Concurrent downloading of a 10MB file (5GHz, 80MHz).

The MLO Performance

Moving on the MLO performance, we get to see some interesting things.

test24
EnGenius ECW520 – Wireless Test + MLO Performance – Qualcomm FastConnect 7800 Adapter (v3.1.0.1323) – Upstream

We immediately notice that regardless of whether set to 320MHz or the 160MHz channel bandwidth, the 6GHz radio doesn’t get very far, but it does have the best near-the-access-point throughput. I did go through all the available MLO configurations and it is possible for the throughput to go above 2Gbps in certain instances, but that’s just upstream.

test25
EnGenius ECW520 – Wireless Test + MLO Performance – Qualcomm FastConnect 7800 Adapter (v3.1.0.1323) – Downstream

Downstream, the performance is a bit more subdued, but still fairly decent. If you want a good throughput both near and farther from the access point, I suggest going for the 5GHz and 6GHz combo, both set to the 160MHz channel width. Otherwise, 2.4GHz and 5GHz is a decent combination as well. At the end, I have included the data I collected when running FLENT.

The Standalone Mode

The EnGenius ECW520 does have a standalone mode, but it’s barely useful. You get some basic monitoring info and what is called Local Setting. It includes a few network-related settings and nothing more, so you do need to rely on the Cloud for the proper experience. This could be a problem in the future if EnGenius decides to not support this model anymore.

engenius-ecw520-standalone
The standalone mode.

The Cloud Platform

I did mention the ECW512L several times, but while this one get the Lite version of the Cloud, using the ECW520 will give you full access to the EnGenius Cloud platform. The areas of interest are, as expected, the Dashboard and Access Points for the monitoring tools of which there are quite a few.

engenius-ecw520-cloud
The EnGenius Cloud platform.

But for configuration, there’s the SSIDs and radio sections where you can set the general settings which will be applied for all access points that will be added to the network. You do get the option to overwrite these settings at the device level with ease and we also get the iconic Diag tools which do come to life if you get the Pro subscription. Then again, even in its current default state, it’s still plenty enough for most network admins.

engenius-ecw520-diag
The Diag tool.

The Conclusion

OK, so should you consider the EnGenius ECW520? SMBs will definitely have a lot of fun with it. The price is fairly low, while getting most of the advanced WiFi 7 features plus the versatility of the Cloud platform. It’s easy to monitor, configure and the Diag tools will be of great help to sys admins. Is it better than the ECW516L? Not really. I think the ECW520 is the better option.

0 0 votes
Article Rating
Subscribe
Notify of
0 Comments
Newest
Oldest Most Voted
Inline Feedbacks
View all comments
0
Would love your thoughts, please comment.x
()
x