Setup question: Boosting 8ta with LPDA external

Zook

Active Member
Joined
Feb 15, 2012
Messages
80
Reaction score
7
I have a CellC 21mb speed stick (Huawei e1820) and am trying out 8ta. I have a promo bundle, so social@8ta has confirmed that I'm on the 8ta network. Tracert seems to confirm that.

Signal is slower than CellC for my location (Edenburg, Rivonia, Jhb), and it drops more often, so yesterday I got one of these:
http://www.poyntingdirect.co.za/pK-...or-cellular-data-cards-include-7-m-cable.aspx
with the Poynting all-USB induction adapter, 2m pole and bracket combo that will make my old TV aerial hide in size-induced embarrassment. I tried various points and directions outside, very close to zero effect. Now, the aerial claims 11db so obviously I'm doing something wrong.

Masters of widgetery and lasers and things, what wisdom can you impart? Is there a 1-page definitive howto on howto poynt the poynter at the correct towery thing?

Zook.
 
Use MDMA to give you accurate signal measurement and then do a 360 scan by turning your antenna a few degrees at 2-3min intervals. Make sure to turn the logging on so you can capture the tower and cellids.
 
Thanks, morkhans. I did try with MDMA, but was using fairly broad brush strokes while turning, thinking that 30-45 degrees would give me some kinda rough improvement that I could finely adjust. I guess not.

Questions:
Does the vertical angle affect things and should I adjust that only after getting the best value horizontally? The bracket allows up-down variation so obviously it's something to consider.
How directional is this thing? Does it give you zero improvement and then suddenly the lights come on, or does it ramp up as you get closer? (E.g. 70% improvement within 20 degrees)
 
30-45 degrees should work but give a little time like Morkhans says there 2-3 mins to let the signal show thru at each stop.
Easiest i found was to find where the tower is by laptop & car & noting cellID then roofing it from there.
I just set the vertical at 90deg
 
That's interesting - didn't think of missioning around in the car. I did find http://opensignalmaps.com/ but they don't seem to have 8ta towers included, or they weren't visible near me.
 
I didn't pick up from your post. Do you use pigtail connector or inductive coupler? With pigtail this modem do not report signal strenght from the external antenna, but from the internal one.

However I had problem with 8ta setting up antenna using different modem, Vodafone K3565. It was working fine with CellC SIM, but with 8ta I was unable to find direction, similar to your case. I haven't tried horizontal polarisation, maybe you should try.

Re: Disconnections. It happens with 8ta frequently. Sometimes it stays connected for hours, sometimes disconnects every few minutes. Start of the session is painful. It need couple reconnects, before it learns how to keep connected.
 
sajunky: inductive. Weirdly even with that, zero effect that I can tell. Put it on, take it off, move things around...*sounds of crickets*

T-Man: 25351, pretty constantly today and yesterday. I think I remember a different one a day or so ago.

It's not horrendous. 10-second avg peak today 1570. But that's barely better than I got with iBurst in 2005 (!) which is now pretty cheap and doesn't have all this attached paraphernalia. On iBurst I added an external omnidirectional antenna and it went from 2 bars to 5, boom. Here I'm squinting to see if there's any change :)

When pointing the antenna, is the idea to have the pointy bit of the triangle aiming at the tower? (Assuming I knew where that was.)
Also, it's 'external', but some seem to have had success putting it in the roof. How much do tiles affect these things?
And for extra points, would a 3m pole instead of a 2m pole have much difference? Obviously depends on the area, but just wondering if anybody has had success with that.

I might try a friend's dongle to see if I can get any variation.
 
If you have a Android phone, download G-Mon and go drive around a bit. Set the logging interval to 20 sec.

It maps signal strength and CellId etc into a CSV file and a Google earth KML file.

While driving around LOOK where the towers are. The 8ta/3G ones are generally on very high towers, often located at the Telkom exchanges.

One thing with 8ta towers though, you can be very close to it and still only get a signal of -107dBm. 8ta has far fewer towers than the other providers. With the other providers you can generally see/measure when you're close to the tower.

O yes, I also noticed that people staying on a koppie or hill generally have very bad reception. They generally have LOS to many towers, but not one of them is focussed onto them. It's as if the phone/modem can't decide which one to use.

If you use G-Mon, please email me the data files. I collect them.:D PM me for my email.

My 2c worth.
 
Sorry T-Man, have iPhone although if I borrow an Android phone I'll shout. Meanwhile I'll try with laptop and MDMA.
 
The 1 at the end show that you're in a 120 sector to the North of the tower.
2 show southeast and 3 south west.

That's useful - so at least I can try a bit more carefully in the South-ish directions. Thanks!
 
Ok, I owe you guys some results. So far, it looks like I have the correct tower (1.3km away and direct line of sight from the roof, if so) and antenna aim, and adding and removing the antenna sleeve from my modem results in...6db difference! Speeds vary and often without the sleeve I can get approximately the same speed, but it'll raise the average a lot and hopefully result in fewer drops.

Will keep testing and tweaking to see if it's a reliable improvement.

Many thanks to morkhans, Tun@, sajunky and certainly not least, T-Man.
 
The 1 at the end shows that you're in a 120 sector to the North of the tower.
2 shows southeast and 3 south west.

Source?

Ok, I owe you guys some results. So far, it looks like I have the correct tower (1.3km away and direct line of sight from the roof, if so) and antenna aim, and adding and removing the antenna sleeve from my modem results in...6db difference!

Either your sleeve is not positioned optimally or you're pointing at the wrong tower. You should get at least 10 dB more than not using the antenna.
 
... You should get at least 10 dB more than not using the antenna.

depending on the length of the cable from the antenna to the coupler - AFAIK loss is about 0.5dB per meter
 
7m cable and the coupler is inductive, so that could account for some loss. Do people get 10+ even with that setup?

Hope not, and hopefully there's more improvement to be had.
 
I've got a similar setup, but am using the Uniterm Direct antenna and a 7m +3m cable and inductive coupler - Had USB modem in 3G router and sat on roof, resetting the router via cellphone after every ~20degree turn, to ensure modem has to again determine best signal - worked quite well!

To me the dB of the antenna was not the big drive, the idea was to get an unobstructed signal
 
Yes i had 7+5m cable & got 20dBm improvement but that was from -90dBm down to -70dBm area.
I don't think you'll see as good if your starting at -70dBm without areil
Also with inductive coupler + 1820)
 
Last edited:
7m cable and the coupler is inductive, so that could account for some loss. Do people get 10+ even with that setup?
Inductive coupler loss is about 6dB, so your results are OK. People get better improvement, if there is better signal at the antenna location than on the ground level. It is not always the case.
 
Inductive coupler loss is about 6dB, so your results are OK. People get better improvement, if there is better signal at the antenna location than on the ground level. It is not always the case.

Yes, the "worst" improvement would be if you have your modem on the roof where the antenna is and compare its signal with/without the antenna at the same location. But if you're inside your house the improvement is usually 10dB+. With a proper external antenna connection a friend of mine gets a 22 dB improvement.
 
Top
Sign up to the MyBroadband newsletter
X