Reasons for a link to be SS11/DS11?

Hi!

I have two PtP links with an issue I don’t quite understand. Both are ePMP 4500L in both ends, running TDD PTP. RFE UltraHorns.

One of them have uplink SS11 and downlink DS11. The other is opposite. They are both from the same tower (along with a few other ePMP4500L without this issue) - but their frequencies are far apart, and spectrum analyzer looks fine, RSSI is fine. Have tried every possible frequency, it’s the same all the time.

I can (and will…) accept if this is caused by interference/noise, but it’s the same whatever frequency I try, and at all times. The link works fine, but at least one of them has an occassional

aid=1 peer=10 disassociated. Reason: 53 (AP KEEP ALIVE RX STUCK)

once a week, or maybe every second week. This could be true for the other link as well, but I forgot to check before I rebooted it.

One of the links have been like that always, from day one. But the other one used to be DS/DS, but at some point it turned to SS/DS.

So sorry for making this so long. My question is really as simple as if this is normal if there is intereference/noise? Or could it be a hardware issue? Either radio or antenna/twistport-adaptor?

I own and operate IsoHorns and Cyber Antennas, RF Elements competitors.

Uplink refers to the access point’s (AP’s) ability to receive a signal from the client station.
Downlink refers to the client station’s ability to receive a signal from the AP.

DS11 represents two spatial streams using MCS 11 (1024-QAM).
SS11 represents one spatial stream using MCS 11 (1024-QAM).

To achieve reliable DS-11 802.11ax MIMO 2×2 performance, sufficient isolation between the two spatial streams is required. In a dual-polarity system, these spatial streams correspond to the vertical and horizontal polarizations. RF Elements does not publish port-to-port isolation figures on their datasheets. Independent measurements of TwistPort adapters show only about 12-19 dB of isolation across the 5 GHz band—which is by far the worst in the WISP industry.

RF Elements UltraHorn antennas use TwistPort adapters that do not have threading on the RP-SMA connectors. This lack of threading can result in loose or inconsistent connections, leading to high VSWR (Voltage Standing Wave Ratio), where the transmitted signal power is reflected back at the radio.

High VSWR can damage the sensitive receive components of the radio—particularly the low-noise amplifiers (LNAs)—which are responsible for amplifying very faint incoming signals, leading to loss of chains and elevating the noise. Excessive VSWR can also occur when operating the radio outside the antenna’s specified frequency range.

Although RF Elements UltraHorn antennas and TwistPort adapters are marketed as supporting the 5 GHz band, most of their antennas do not actually cover the full range from 5.15 GHz to 5.85 GHz. Instead, many models only support down to 5.18 GHz. Operating below 5.18 GHz can cause excessively high VSWR, which may damage the radio, especially over time.

Other common causes of high VSWR include manufacturing defects, user error (e.g., improper assembly), or environmental issues such as water ingress.

Professional antenna manufacturers specify the maximum VSWR on their datasheets and test every unit to ensure it meets or exceeds that specification. RF Elements does not publish VSWR values on the UH-TP-5-24 datasheet (or most other datasheets).

If you want to max out the capabilities of your Cambium radios, you need good antennas. I assume your link distances must be short if you are getting DS/SS11 with RF Elements 24 dBi horns. If you are on a budget, I recommend IsoHorns 0.6m CPE dish. If you want top performance or you are colocated with many other radios, I recommend Cyber Antennas 0.6m high performance dish. Both are amazing values and offer the best price-performance in the industry.

If you don’t have the budget for new antennas, just swap out the UltraHorns with any 5 GHz antennas and see what happens.

When you use the radio’s built in speed tests, do you get approx. the same results in each direction? When you look at the radio’s Monitor → Performance page, and scroll down to the bottom, what are the majority MCS %’s that packets are being sent at?

Having one side of the link always has SS MCS rates can indicate that one of the two chains on the radio is having issues.

While I have no doubt that there are better antennas, and that RFE might be a bad choice - it’s what I have to do with for now, and I have several other links with the same UltraHorns performing just fine over longer distances, so it can’t be just bad specifications. So I want to investigate further before splashing out for new antennas.

So to answer Eric; when using the radio’s built in speed tests, I get (for a 20 MHz link) 119/58 Mbps. This is on a 50/50 TDD link with co-location disabled.

On this particular radio, 92,5% of downlink packets are DS MCS 11, and 97,4% is SS MCS 11 for uplink packets.

Additionally for uplink, there is 2,4% SS MCS 10 - and actually there is a 0,2% DS MCS 6. For downlink there is 7% DS MCS 10 and 0,2% and 0,3% of DS MCS 9 and DS MCS 8. This link is only 2400 meters.

The other problematic link is quite a bit longer, 11 000 meters (but I also have one at 12 000 meters that performs flawlessly, both uplink and downlink have about 90% DS MCS 11)

This other problematic link has about 85% SS MCS 11 for downlink, but the uplink is DS MCS 9, 10 and 11. So this is opposite of the first link mentioned. For this link I even tried replacing all of the hardware at the SM end (radio, antenna and twistport adapter), but that made no difference. I have not tried swapping out at the tower yet, but will try if you think it’s worthwhile.

All these (three) mentioned radios are in the same tower, but frequencies far apart (and I have tried most of them!), and have about 50-60 degrees between them. And again there are a couple of other identical links in the same tower without problems.

It has been working fine for over a year - it’s just not giving me the speed I want and need…

It sounds like some chain issues with the radio. I’d open up a support ticket with Cambium and have them look it over and see if an RMA is necessary.

Thank you for your advice. I did create a support request today, and was recommended to replace the radios :slight_smile: Will keep you posted…

1 Like

So this week we swapped some radios. Tried to swap the SM first, since they are more accessible. One of the links instantly came back with DS11/DS11 after replacing the 4500L with a 400C. Great! Just swapping the SM in the twistport adapter, did not make any changes on adapter nor antenna.

The second link nothing changed… So we had to go to the tower anyways. After replacing the AP for that second link with a new 4500L, it came up as DS11/DS11 immediately. Same here, just backup config, restore config, swap out of and into the same twistport adapter and antenna.

So pretty sure it was hardware related… Now I have 7 PtP links from the same tower with UltraHorns and 4500L with DS11/DS11 95%+. Very happy with that! :slight_smile:

Was a bit afraid it was noise and intereference causing this, so very happy it was so easily solved, after all.

2 Likes