I have been struggling to help a friend of mine living somewhere on a farm outside Stellenbosch with his slow speeds.
The strongest signal at his place is from tower ID 4231x, around -80 dBm inside and -70 dBm outside. Mostly he gets a few hundred kbit/sec from this tower. It is not known if the sporadic higher speeds are from a different tower.
We did some testing over last weekend. I went to his place and took my laptop and E1820. He has his own laptop and E367. We both did speedtests while getting around -70 dBm from the tower (4231x). Different speedtest.net servers all gave around 300 kbit/sec. To eliminate the possibility of local interference in the area, we did some "wardriving" and drove to a different location a few kilometers away where we could get a good signal again from the same tower. The new location and my friend's place are completely obscured from each other by a large hill in between. We suspect the tower is somewhere on this hill. The signal strength was between -60 dBm and -69 dBm. Here we did speedtests again on our respective laptops/modems and again got a few hundred kilobits/sec while uploads was around 1 mbit/sec.
To prove that his hardware was not at fault, during the wardrive on our way to the alternate location, we stopped near a different tower (44003, signal -70 dBm) and he did a speedtest again. There he got more than 4 mbit/sec download.
I sent this information to the Cell C rep and he promptly passed it along to tech support. A few days later a technician visited my friend's place as well as the tower. His assessment was that "the signal is too weak" at my friend's place. The technician said he drove to the tower and "at about 200 meters from the tower his speedtest gave 4 mbit/sec up and 1 mbit/sec down". He also said we tested "too far from the tower". He said my friend is welcome to ask for a refund given his report that he will attach to our reference number. This was around mid morning on Thursday.
Hoping that somehow the problem disappeared, my friend tested again on Thursday night and only got 100 kbit/sec on download.
Does distance affect download speed, even though the signal can be very good? I don't think so, at least, not in our case. We gathered from testing that the crappy tower is only a UMTS900 tower. So by using MDMA's band selection tool and only selecting the UMTS2100 band we could "deselect" this tower from being used. With this setting he is able to get -90 dBm outside in the garden from tower 44006 about 4 km away. Lo and behold, he gets 1.9 mbit/sec download from this tower! So, crappy signal, long distance, much better speed. I got 2.4 mbit/sec from the same tower (ID 44006) and I live less than 1 km from it.
We are struggling to make sense of all this. What did we miss? Why did the technician get good speed and ourselves repeatedly don't? How do we proceed from here?
The strongest signal at his place is from tower ID 4231x, around -80 dBm inside and -70 dBm outside. Mostly he gets a few hundred kbit/sec from this tower. It is not known if the sporadic higher speeds are from a different tower.
We did some testing over last weekend. I went to his place and took my laptop and E1820. He has his own laptop and E367. We both did speedtests while getting around -70 dBm from the tower (4231x). Different speedtest.net servers all gave around 300 kbit/sec. To eliminate the possibility of local interference in the area, we did some "wardriving" and drove to a different location a few kilometers away where we could get a good signal again from the same tower. The new location and my friend's place are completely obscured from each other by a large hill in between. We suspect the tower is somewhere on this hill. The signal strength was between -60 dBm and -69 dBm. Here we did speedtests again on our respective laptops/modems and again got a few hundred kilobits/sec while uploads was around 1 mbit/sec.
To prove that his hardware was not at fault, during the wardrive on our way to the alternate location, we stopped near a different tower (44003, signal -70 dBm) and he did a speedtest again. There he got more than 4 mbit/sec download.
I sent this information to the Cell C rep and he promptly passed it along to tech support. A few days later a technician visited my friend's place as well as the tower. His assessment was that "the signal is too weak" at my friend's place. The technician said he drove to the tower and "at about 200 meters from the tower his speedtest gave 4 mbit/sec up and 1 mbit/sec down". He also said we tested "too far from the tower". He said my friend is welcome to ask for a refund given his report that he will attach to our reference number. This was around mid morning on Thursday.
Hoping that somehow the problem disappeared, my friend tested again on Thursday night and only got 100 kbit/sec on download.
Does distance affect download speed, even though the signal can be very good? I don't think so, at least, not in our case. We gathered from testing that the crappy tower is only a UMTS900 tower. So by using MDMA's band selection tool and only selecting the UMTS2100 band we could "deselect" this tower from being used. With this setting he is able to get -90 dBm outside in the garden from tower 44006 about 4 km away. Lo and behold, he gets 1.9 mbit/sec download from this tower! So, crappy signal, long distance, much better speed. I got 2.4 mbit/sec from the same tower (ID 44006) and I live less than 1 km from it.
We are struggling to make sense of all this. What did we miss? Why did the technician get good speed and ourselves repeatedly don't? How do we proceed from here?