E1820 External Antenna Connector

[ginggs]

What is even more interesting 1geoff99 is that the location if the CellC modem is still important. I still need to have it close to a window. It is as if both the roof antennae and the modem's location is important. I have no idea why that would be. Any technical expert out there with a theory?

I'm no expert, but I can come up with a theory; the main antenna and the diversity antenna need to be able to "see" the same cell.

I did my tests with a Poynting LPDA-A0044 which is highly directional. In my case the antenna is aimed at 90 degrees to the window in my office, so it is very likely the main antenna on the modem in my office and the diversity antenna on my roof would pick up different cells. In fact, while doing the tests, I was watching the cell IDs when connecting the external antenna, hoping that the modem would associate with a different cell which would have been some evidence that the connector was working.

Regarding your TV antenna, if it is a VHF antenna I cannot see it having any effect; VHF only goes up to 300MHz, but a UHF antenna should have gain from 470MHz to 890MHz, so there would be gain at 900MHz

 

[ginggs]

I did some tests at my father's place in Somerset West yesterday. He still has a Poynting LPDA-A0021 antenna on his roof from his days on MTN and his own CAB-038. He also has a Huawei dipole antenna that he got for his Huawei E620 a couple of years ago. I alternated between testing the E1820 with no antenna and one of the antennas directly connected via the external antenna connector. The modem was on Cell ID 44221 and the signal strength was -85dBm for the duration of the tests which was about 20 minutes.

Antenna	Download	Upload	Ping	Link
None	3.48	0.13	96	link
Dipole	5.15	0.91	85	link
None	3.19	0.54	98	link
Dipole	3.66	1.04	90	link
None	3.85	2.01	90	link
Dipole	3.88	1.19	69	link
LPDA	2.41	2.73	73	link
None	3.81	2.56	87	link
LPDA	3.57	2.26	106	link
None	3.06	2.52	93	link
LPDA	3.91	2.74	93	link
None	3.88	0.13	90	link

Update with additional info:
Cell ID "seen" by the modem's internal antenna is 44221 (7.5km away)
Cell ID "seen" by the LPDA-A0021 antenna is 44262 (2.1km away)
Using the stock E1820 version 11.831.03.00.00 firmware

 

[wndrbees]

Comparison	Download	Upload	Ping
None-Dipole	0.275	0.714	0.219
None-LDPA	0.646	0.047	0.812
LDPA-Dipole	0.224	0.003	0.469

A t-test can be used to analyze whether there is a statistically significant difference between the data sets. A resulting p-value of less than 0.05 corresponds to a significant difference. Using a two-tailed t-test (equal / unequal variances where used based on a f-test) the following p-values where found on Ginggs' test results:

The null hypothesis is rejected for all but the upload speeds of the None-LDPA and LDPA-Dipole comparisons. Statistically there is no difference with the different configurations exept for upload speeds with the LDPA. (Note: sample size is small)

 

[1geoff99]

Maybe you need to have a signal strength worse than -85dBm to notice an effect?

 

[DXL]

Pictures PLEASE

Pictures PLEASE

One of the pigtail connecter that fits into the modem.

One of the modem connector port - when the anntennae fits in.

Thanks

DxL - TEAM

 

[Tun@]

I'm also possibly having to do the Ariel thing unless CellC adjusts something.
Before i go & waste money on a areil, I see at the bottom of that 3g wetpaint link that was posted earlier a post stating "it has a connector for external antenna, but it is disconnected inside from the modem's scheme...."
I'm not a Techie but could a multimeter confirm this true or untrue?
What would i need to do - connect a CRC9 connector to it with 2 wires & check for continuity?
Would that give me an answer?

 

[sajunky]

Anybody tried Cell-C HSPA+ modems on other networks in 3G/HSPA mode?

Acoording to 3G_modem_wiki E1820 is single band W-CDMA 2100MHz device. Information about E1752 varies from single band 2100MHz to dual band 2100/1900MHz (also missing 900MHz Cell-C band).

My understanding is that both Cell-C branded modems are customised to operate in W-CDMA 900MHz band. Are they also dual band 2100/900?

Related question is whether internal antenna diversity feature is active on the single band (and which one) or both. If it is linked to single band only, it is probably still 2100MHz, not 900MHz.

To explain this situation look on details of my (useless now) E272 modem. It is 3-band W-CDMA 2100/1900/850 with internal antenna diversity 2100MHz. There are two UMTS receiving channels: the main channel with tripple receivers 2100/1900/850MHz and the separate channel with single diversity receiver 2100MHz. The quickest way to convert this modem to 900MHz band would be replacing 850 MHz tranceiver chip with 900MHz chip and associated filters. Similar frequency, no changes in printed circuit board. In result no receive diversity on 900MHz, but still working 2100MHz with diversity.

How E1820 is modified? Originally It is single channel 2100MHz, then Cell-C version should be also single channel. It wouldn't work in 2100MHz network. Unless it is version of E1823.

Comment regarding external antenna issue as shown in:
http://mybroadband.co.za/vb/showthr...na-Connector?p=4877444&viewfull=1#post4877444

It behaves similarly to my E272. When the external antenna is plugged in, the mechanical switch inside the socket cuts off the internal antenna. Howether when the external antenna is disconnected and switch is stuck open (like it was in my case), the modem is not able receiving any signal, including UMTS 2100MHz in which it has permanently attached internal diversity antenna. It looks like all signal meassurement and channel control is based on information from this channel. No signal in this channel, no attempt for diversity! Now assuming the external antenna socket is wired to the diversity channel and switch gets stuck open, it affects only diversity feature and guess ... nobody notice it.

There are some speculative assumptions here. It would be perhaps better if we had technical specification for these modems.

 

[ginggs]

My understanding is that both Cell-C branded modems are customised to operate in W-CDMA 900MHz band. Are they also dual band 2100/900?

Yes, according to the Cell-C's E1820 packaging:

Cell C E1820 Key Features

• UMTS/HSDPA/HSPA+ 900/2100MHz
• GSM/GPRS/EDGE 850/900/1800/1900 MHz
• Equalizer and Receiver Diversity
• HSDPA data speeds up to 21.6 Mbps (28 Mbps MIMO Ready)
• HSUPA data speeds up to 5.76 Mbps
• Data, SMS and USSD service
• Micro SD Card slot
• Plug and play
• External Antenna Interface

I have tested a Cell-C E1820 with MTN and Vodacom SIMs and can connect to UMTS @ 2100MHz. Remember Cell-C have stated they will be using both 900MHz and 2100MHz, see here.

Related question is whether internal antenna diversity feature is active on the single band (and which one) or both. If it is linked to single band only, it is probably still 2100MHz, not 900MHz.

I have just tested a Cell-C E1820 with a Vodacom SIM and can see no difference in using the external antenna connector at 2100MHz either.

 

[sajunky]

Thanks ginngs. It makes sense to sell modem covering both bands. However diversity in both bands should be confirmed, it is a subtle matter, difficult to test. Anybody confirm the same for E1752? This is a modem I am looking for.

I have just tested a Cell-C E1820 with a Vodacom SIM and can see no difference in using the external antenna connector at 2100MHz either.

 

[ginggs]

However diversity in both bands should be confirmed, it is a subtle matter, difficult to test.

How so? When I put my E272 in its cradle the signal jumps up about 20dBm after a second or two, similarly when I used to plug the pigtail into my E620. Similar results using Poynting's inductive coupler with the E1820, but using the pigtail on the E1820 results in no visible change.

Would you please explain what 'diversity' means and how it can be tested?

Anybody confirm the same for E1752?

The E1752 has no external antenna connector.

Cell C need to clear this up now, they are advertising the 'External Antenna Interface' as a feature of their E1820.

 

[sajunky]

How so? When I put my E272 in its cradle the signal jumps up about 20dBm after a second or two, similarly when I used to plug the pigtail into my E620. Similar results using Poynting's inductive coupler with the E1820, but using the pigtail on the E1820 results in no visible change.

It is why my posting has been moved here. My case of E272 faulty switch shows these facts:
- signal indicator you see comes from the main channel
- modem is unable initialize connection if there is no signal in the main channel. There are above reports that is also unable to maintain connection when signal in the main channel completely disappears.
- E272 external antenna is wired to the main channel. Faulty switch shows limitation of the chipset or firmware.

Now Huawei went easy way avoiding warranty returns. I have no doubt that what was said in this thread is true. E1820 external antenna is wired to the diversity channel exposing the same bugs. Inductive coupler works because it supplies signal to both channels.

Would you please explain what 'diversity' means and how it can be tested?

Receive diversity means multiple receiving path. In digital communication data is assembled from small pieces. Each one carries with extra information like ID of such piece, so you know how to put everything together. There is also data used to reconstruct data from erroneous pieces or at least report error or missing piece of data, so you can request retransmission of such piece. But it only happens when signal is processed at the higher network layer. For now data is keept assembled from smaller pieces and passing through to the higher layer. At this layer data is assembled and requests are made for missing pieces. However it leads to increasing latency. Gamers are you watching?

It is why receiving diversity comes with resque. Data is assembled simultaneously from two or more separate receiving channels and if some pieces are not recoverable it can be taken directly from diversity channel (looking for the same ID) instead of waiting for retransmission.

Antenna diversity (receive) means two antennas receiving the same frequency. It doesn't imply multiple decoders. It says only that receiveing signal can be switched from any antenna based on signal strenght, signal quality or both. Also used in analog systems. Also used for transmitting the same signal with two or more antennas - different polarisation by example.

http://en.wikipedia.org/wiki/Antenna_diversity

How it can be tested? I don't know. Signal indicator shows only signal strenght of main channel. You don't have information on signal quality etc. I think the best D.I.V. method would be based on messuring latency. Not a speed, because it is masqueraded by use of large buffers..

 

[ginggs]

Before i go & waste money on a areil, I see at the bottom of that 3g wetpaint link that was posted earlier a post stating "it has a connector for external antenna, but it is disconnected inside from the modem's scheme...."

I seriously doubt Huawei would go to the expense of installing a gold plated connector and not wiring it up. With electronic devices one often finds the reverse though; there is place on a PCB for components which were never mounted, for various reasons. Sometimes it is possible to install them yourself and enable features again.

How it can be tested? I don't know. Signal indicator shows only signal strenght of main channel. You don't have information on signal quality etc. I think the best D.I.V. method would be based on messuring latency. Not a speed, because it is masqueraded by use of large buffers..

If it is so difficult to notice a difference then why bother? I'm going to stick to using the inductive coupler until we know more. Hopefully there is a firmware upgrade to come which will enable diversity or MIMO, or whatever it is that this external antenna connector is used for.

 

[Tun@]

OK - i'm concerned you seemed to get no improvement on the E1820 when you did your tests using that connector pin tho.
At the moment i'm about to grab a Poynting LPDA-A0044-7M & try use the universal adapter on it instead of CRC9 - Use a ADPT -026(guess thats what your calling the inductive coupler)

Any better ideas to trying that?

 

[morkhans]

OK - i'm concerned you seemed to get no improvement on the E1820 when you did your tests using that connector pin tho.
At the moment i'm about to grab a Poynting LPDA-A0044-7M & try use the universal adapter on it instead of CRC9 - Use a ADPT -026(guess thats what your calling the inductive coupler)

Any better ideas to trying that?

It would seem the inductive coupler is the only way to get an external antenna connected to the E1820 at this stage. So it's your best bet.

 

[ColdComfort]

I do think that the aerial plug does work, in my case I have no HSDPA+ signal inside my house without the antenna. But with it and using the antenna connection on the modem I can pick up a HSDPA+ signal that is usable but not nearly as good as the connection that I get if I am outside in my garden.

I then remembered that I had this adapter I used with one of my old cellphones to connect and my connection went from 2-3mb on speedtest to 6-8



I want to know if the other adapter that fits around the modem better is better than the one I am using now?

 

[morkhans]

The device you have uses the same inductive principles as the other model. It's just got a rubber band to make it fit snugly over the various sized modems. If it's working for you stick with it.

 

[Tun@]

ColdComfort what aerial/antenna is that using please...?
My signal is the same - fine in the garden but nil HSDPA inside.

 

[ColdComfort]

The device you have uses the same inductive principles as the other model. It's just got a rubber band to make it fit snugly over the various sized modems. If it's working for you stick with it.

Thanks I will then stick with the one I have if there is no real difference between them

ColdComfort what aerial/antenna is that using please...?
My signal is the same - fine in the garden but nil HSDPA inside.

I am using this model antenna LPDA-A0021-03

 

[Tun@]

Ok,same animal as i'm looking at grabbing,Cheers

 

[sajunky]

If it is so difficult to notice a difference then why bother? I'm going to stick to using the inductive coupler until we know more.

I agree you don't need to bother. With directional antenna you get rid of noise, multiple signal path and many other factors. I started a new thread about these modems asking for technical details. All I can tell that receive diversity is critical for all others who do not use external antenna. Unfortunately my posting has been moved here and being off topic created distraction.