Pepwave SOHO MK3 - what wifi radio is used?

I am just in the process of installing my first Pepwave surf soho router. So far I am doing ok with the installation but am stumped on one thing. I would like to know what modem a client uses (2.4 GHz or 5 GHz) when they log on. This seems such a basic piece of knowledge that I am thinking I am overlooking it someplace but even after going over everything multiple times I am unable to find a way to display this information.
Could someone please point me in the right direction ?
The issue is coming up because I have a 300 MHz cable link (yes I know it’s faster than the soho can run) but am only getting about 60 - 70 Mb/s on a speed test. This is reasonable for a 2.4 GHz wifi link but not for a 5 GHz link and I would like to figure out how the clients are connected.
I tried enabling only the 5 GHz radio for the SSID in question but there was no improvement.
A 5 GHz link on an Asus R68AC gets me a download speed of around 270 Mb/s from the same computer and same speed test.

1 Like

Hello @jens,
Welcome to the Peplink Forum.

Have a look at the reply to @Michael234 on this previous post, shown are images on where to find what Wi-Fi frequency/standard your clients are using.


This should help you know what clients are connected to with your Wi-Fi.

As to why you are not getting the speeds you expect, first see what your devices are connecting on as it may give a clue.
Happy to Help,
Marcus :slight_smile:

1 Like

Is the feature request being considered and if so, is there a time line?
For a single home user, inControl is overkill and awkward. It could be as simple as replacing the ‘Active’ icon in the client list with an icon incorporating a speed indicator.

1 Like

Jens,

To get an answer to any question, you have to use the right terminology.

Wireless router clients do not connect to modems, they connect to one of two WiFi radio frequency bands.

Also, the speed of modems is not measured in MHz it is measured in bits per second. Actually millions of bits/second or Mbps. You Internet connection speed is 300 Mbps.

As for a speed test, you should first test with an Ethernet connected computer as Ethernet is faster than WiFi 99 percent of the time. Your Surf SOHO should run at somewhere around 120Mbps. All that said, as long as the speed is over 30Mbps most people will be very happy.

A speed of 60-70 is indeed reasonable for the 5GHz radio spectrum. MUCH depends on your neighbors, your choice of WiFi channel, your choice of channel width and your signal strength when you do the test. Plus, the client device also makes a very big difference. With all those variables, there is no one good or bad speed, even on the 5GHz frequency band. You may be getting your information from amateurs rather than professionals.

Finally, when do a speed test look at the dashboard of the Surf SOHO router to see the cpu usage. If its near 100% that would be a problem. Some Asus routers also show cpu usage.

Michael

2 Likes

Yes indeed I slipped and used MHz instead of Mb/s but you did understand my meaning correctly.
I agree that direct connection to the SOHO is the most reliable way of testing speed but this is not possible in my case. Instead I compared the SOHO throughput to the Asus throughput.
Speed was a big issue for me before I purchased the SOHO - it took some time before I justified the overall drop in throughput of the SOHO by saying to myself I would get a more secure and reliable network. 30Mb/s might have been acceptable 10 years ago, 75 Mb/s is acceptable to me (barely) but of course I have to share with other users so it will be rare that I see 75Mb/s.
I agree that channel width (80 MHz) and channel choice (auto) make a difference. I have duplicated conditions as best as I can between the two routers. Client device is identical and distance to the router at about 15 ft is the same.
I do not expect to see Asus speed on the SOHO, I do however expect to see 120 Mb/s

1 Like

Yes indeed I slipped and used MHz instead of Mb/s but you did understand my meaning correctly.

Yes, this time. The hardest part of solving a problem is often creating a fully complete description of the problem.

I agree that direct connection to the SOHO is the most reliable way of testing speed but this is not possible in my case.

You don’t have an Ethernet connected computer?

Instead I compared the SOHO throughput to the Asus throughput.

But is it Apples to Apples? As I said, look into channel width on each router. Look into the channel used by each router. Look into the signal strength. Look into the interference from nearby networks, cpu usage … yada yada yada

If you can run inSSIDer
https://www.metageek.com/products/inssider/
it has a very useful display of channel usage. Right now. Current. Last 5 seconds channel usage. So, you can really see which channels are heavily used by your neighbors. It helped me.

I get about 70Mpbs on 5ghz with a strong signal and very little interference from neighbors on the channel I am using. But, I am using the narrowest possible channel (20MHz) because there are many nearby networks. I might be able to up the speed by widening the channel - or - that might interfere with a neighbor and slow me down.

2 Likes

Why would anyone need more than 15Mbps?

A 70 Mb/s speed is reasonable with a 20 MHz bandwidth but as indicated, I am running 80 MHz bandwidth.
No, I do not have a computer with an ethernet port other than a Raspbery pi which would bring with it a whole new set of issues.
BTW, cpu usage on the SOHO during the speed test was between 65% and 85%.

1 Like

Why would anyone need more than 15Mbps?

Because we can …
This reminds me of the famous words of Bill Gates (that he denies having said) “640K ought to be enough for anybody.”

In answer to your subsequent post (I can’t answer it in sequence because as a new user I am not allowed any more post for 18 more hours)

Ie. 8 Mbps = 1 MB/s before overhead *ie in reality you get half that

You are mixing up units. In your earlier post you said that 70 Mb/s = 560Mb/s instead of 70 MB/s = 560 Mb/s (notice the capital ‘B’)
All the speeds we are talking about are in bits/second … the SOHO speed, the cable ISP link, the Asus router. There is no conversion factor involved anywhere.

1 Like

802.11ac/HT80 = 70 MB/s = 560Mbps not including overhead.

These are all good numbers. Your looking good.

How does 70 Mb/s turn into 560 Mb/s? And if 802.11ac is limited to 70 Mb/s then how can the Asus router manage to give me 270 Mb/s?

1 Like

https://www.checkyourmath.com/convert/data_rates/per_second/megabits_megabytes_per_second.php

Ie. 8 Mbps = 1 MB/s before overhead *ie in reality you get half that

Ooops. Please give a screen shot to help troubleshoot… Sorry!

Test the Asus router again, your WiFi neighborhood may have changed.

Was Asus also using 80Mhz wide channels? And the same channel that the Surf SOHO is using? You have to compare apples to apples.

What speed tests are you running?

It should go without saying not to trust a single speed test. Four different tests from four different companies will give you four different results even if run right after each other.

You also never mentioned your firmware release.

And, does your WAN profile have realistic settings for upload and download bandwidth?

And, are you the ONLY wireless client when you do the tests?

Finally, when you run a speed test, go to Status -> Real-Time Usage reports.
There you will get another opinion of the max speed in each direction.

As for the numbers, Mbps is mega BITS per second. That is the most commonly used speed measurement. If you should see MBbps that is meg BYTES per second. Annoying as hell that everyone does not use the same measurement.

2 Likes

Maybe your WIN10 laptop is doing an update in the background? I had that same problem recently.

1 Like

I have set up two computers to see if the throughput would change, maybe I was maxing out the computer or the link to the router somehow. Alas, the available overall throughput did not change substantially - in other words the current 85 Mb/s throughput was shared between the two computers.
I set up one computer on 2.4 GHz with a 40 MHz bandwidth and the second with 5GHz and 80 MHz bandwidth to test if the wifi radios were being overloaded. The answer is no … available throughput did not materially change.
I set up both Asus and SOHO routers for the same 5 Ghz channel with both set to 80 MHz bandwidth and have seen up to 300 Mb/s on the Asus router with around 80 Mb/s on the SOHO.

This concludes my testing unless somebody can come up with a substantially different way of testing things. I am satisfied that the SOHO is not capable of the advertised throughput.
It took considerable contemplation deciding to go with the reduced speed of the SOHO over the Asus but these contemplation’s were done on the assumption that the advertised throughput was accurate. I would have not purchased the router had the speed been rated at 80 Mb/s. I have not yet decided what to do next but chances are that the router will be used as a fallback/spare. It is likely not cost effective to ship the unit back to the seller in the USA (I am in Canada and return freight will be in the $50 range going across the border)
Yes, the SOHO has some interesting features that would have been nice to incorporate into my overall network setup but I am not prepared to reduce my speed to about 25% throughput compared to an Asus router.
Overall, especially after reading so many rave reviews, I am very disappointed.
I will reach out to 5Gstore.com and raise a trouble ticket to see if they can figure out if I have a setup/configuration error.

Sorry about the delay but new forum members have limited posting privileges and was unable to post until today.
Other bits - I am running Linux so no background updates.
Firmware 8.0.2
WAN speeds are set to 20 Mb/s up and 300 Mb/s down which is what my ISP is telling me.
Hopefully all questions are answered.

1 Like

Sorry to hear that.

You observed 80-85Mbps down - was there anything going up?

The SURF is conservatively rated at 100Mbps combined up/down throughput, but many users have observed up to ~120Mbps.

Using ~120Mbps as an example that means you could do 100Mbps down and 20 up. Or 60Mbps down and 60 up. Or 120Mbps down and 0 up. Etc… 120Mbps is enough to support ~12 HDTV streams (1080p) and surf the net on top of that.

You can monitor download/upload speeds in the Status>Client List - it shows both up and down transfer rates.

I also suspect that performance may have been reduced slightly due to numerous security firmware updates. To me that is a good thing - that’s why I bought it - security was my #1 concern. Not only that, but it is normal and common to every vendor that updates their firmware on a regular basis. It’s certainly better than the alternatives. More on that later.

I am still on 15Mbps DSL and its still good enough for 2-3 HDTV streams - yeah that’s pushing it but it works.

Cable is often not as fast as advertised - usually you only get half what they claim on average due to the shared nature of cable or next month when they sign up 10 new customers on your shared neighborhood loop. Cable is also less secure than DSL (at the moment)…

Naked Security – 14 Jan 20

‘Cable Haunt’ vulnerability exposes 200 million cable modem users

I recently cancelled a planned upgrade to cable because of the above issues with it.

I found that the faster consumer grade routers are plagued with security problems, never get updates, and run old insecure Linux distros/ are easy to hack. This may be common knowledge in the security community but most home users don’t know that - I didn’t up until recently.

To me it’s just not worth it - especially in todays threat environment. Everyone is running KALI Linux where I live apparently - you can probably load it on a “smart TV” now or a WiFi enabled toaster - they have a lot of distros!

I got deauth’ed and hacked/harassed for years by a bad neighbor before I figured out what was going on. The SURF MK3 was a key part of my security solution.

The SURF has put an end to these issues for me and the idiot neighbor that “mysteriously” got caught by law enforcement was forced to move, got a nice shiny new restraining order to go with their criminal record and enjoy some house arrest, fun legal fees, limited future employment options, etc… None of the neighbors here bother me anymore for some reason.

The SURF is powerful and easy to manage. Many business grade routers cost 2-3 times as much as a SURF and use CLI / Linux commands instead of an easy to use GUI. At 15Mbps I am literally only using 1/8th of it’s capacity.

My old consumer grade “gaming” router was hard coded to close the admin window after 1 minute for “security purposes”. That meant I couldn’t even keep it open to keep an eye on it. It would time out when I was trying to enter long WiFi passwords or change settings, check logs, etc… It was like a bad joke. Now it’s my new doorstop.

Personally I would drop cable ASAP, get a secure DSL connection and keep the ASUS as a backup and stick with a secure business class solution. For me this is kind of like the tortoise and the hare - I decided that it doesn’t matter how fast you go if you don’t get there.

Read the www.routersecurity.org website. Most home consumer routers are like swiss cheese from a security standpoint. Some have WPS permanently enabled even if you uncheck it in the menu. Some have hard coded back doors, etc., etc…

Your use case may differ from mine.

**Edited for clarity

1 Like

For those following this matter, 5Gstore has replied to my concern and has asked that I switch the WAN speed from AUTO to manual (I selected 100 Mb/s duplex). Download speed has now increased to around 95 Mb/s so a good improvement. CPU utilization peaks at 100%.
Still a bit short of expected results but heading in the right direction but cpu utilization has me a bit concerned that speed is now topped out. I will see if 5Gstore has any other tricks up their sleeve.

happysurfer, speed tests are broken down into download and upload speed. All tests have negligible upload traffic during the download speed and all resources should be available for download so I should see around 120 Mb/s

2 Likes

1 Like

That was enabled on the initial setup and all speed tests were done after this was enabled.

1 Like