3000L Half Throughput Link test

Hi all. We now have around 8 x 3000L deployed for PTMP - some with asymmetrical 90 degree RF Element Horns and others with 90/120 cambium emp2000 antennas. It seems no matter what we do we always seem to only get half throughput as to what is expected vs Linkplanner on all of these radios. We have ruled out interference and we have very very little anyway where we deploy. I feel like we are missing something simple. For the latest test we had a epmp1000 with omni antenna that was suffering hugely in peak with too many clients (around 40). So we put up a new access point which is a 3000L with epmp2000 90/120 deg antenna. We moved around half of the clients onto the new AP. All working great and solved the old radio which is back to its normal self. Only issue is if you are connected to the Omni (epmp1000) you get around 300Mbps aggregate to each subscriber (happy with that) but if you swap a radio to the new radio/sector you only get around 150mbps aggregate. the 40 odd clients are all Force 180 or Force 200 and understand that this is backward compatibility but we see this same issue on other sites that run new 3000l radios and force 300-25’s and 300-16’s and so have ruled that out for the moment. Both running 40Mhz channels. So it feels like whenever we deploy these 3000L’s they seem to only run single chain or something is amiss. As mentioned feels like something simple we are missing. Interested in any comment from you guys - thanks

all your client computers have the same firmware version as your ap? Are all your clients working in N or do you have any sm in AC? remember that if your ap is Ac and you have a cell of sm in N and AC the ap will modulate in N leaving out the Ac of your teams thus losing “the benefits” of Ac

1 Like

Thankyou for the reply. All radios are same software version. As mentioned in my post we are not just seeing this in the above situation. I am not looking for the benefits of AC I am looking for the same results but getting around half and I have same situation with using AC to AC. My OP explains the situation and there is something wrong overall with 3000L that we are missing. Appreciate the reply and suggestions nevertheless- thankyou

I think you missed the point. If you have a sector full of F180s, F200s then you will not see the same throughput as what link planer shows as it assumes AC modulation unless you specify the client SM and very accurately place the SM on the same gps coordinates of the client location. Removing any clutter (trees, bushes, low hills) from the total path and ensuring you most accurately align the SM antenna to achieve the required signal and SNR levels. Real life vs linkplaner is different, sometimes very different. Some places linkplaner will say a F180 is fine but you really need a F200 or F300-19 or F300-25. In short linkplaner gives you an idea of the possibility but does not tell the whole story. On the other hand, cnHeat uses actual clutter and geographic measurment data to tell you where your signal is, how strong it is and what the clutter losses are in a 2D map with radio elevation specified.

Then we start to talk about test method, wireless link test works as an indicator but to be accurate you need to ensure no other data is passing at the same time and the test duration and setup you choose also determines max throughput.
Windows downloading an update inbthe background can skew the numbers. Do you have an Iperf server setup at the tower? If not then there really is no way to test the throughput of the cables and switches with the wireless link. This can also get complicated with QOS and minimum service speed contracts.
Client side devices also play a roll, testing over wifi (a round robin shared system) makes traffic bursty and the SM will have queue depth troubles as the queue is always empty. This sounds good but in reality an empty queue means the SM doesnt need higher modulations to pass the data.

We have 3000Ls with F180s and have no problems reaching the SMs max throughput for the location IF we do our best to select the best location on the building, are careful with out alignment and we do not have over powering interference.