So what's the difference between the white and black cell c modem?

[Lounger]

Read some more, especially my previous post in this thread.

I did, and it doesn't contradict what I wrote. In fact it says something similar. The point is, the buyer will not see any difference in performance atm.

 

[ginggs]

I did, and it doesn't contradict what I wrote. In fact it says something similar.

Not sure we're looking at the same thing, but the maximum throughput of the E1752 seems to be under 6Mbps, quite a few people have posted speedtest.net results with the E1820 over 10Mbps.

The point is, the buyer will not see any difference in performance atm.

At the moment, quite possibly, but who wants to buy a modem that's already obsolete?

 

[oronte]

Don't waste your money with the black modem, you'll almost never get those advertised speeds (and as the network gets more strained it's less and less likely. I only had good speeds on the day I got the modem since then I haven't seen a download go faster than 1mbps, even though my area is supposed to be fully covered.

 

[marlboro]

How sure are you that the black modem won't still give you better speeds than the white even if not reaching full potential? Those of you giving advise here like experts do you actually know the transmit/receive power, internal electronics design, etc. to compare for both modems? Or at least have you done tests from the same location with both modems and did you get the same results?

 

[Splice]

The white one is R600 or so. for 3G

The black one is R1500 and it's a .5 faster I think? 3G.5

I don't really care about speed anyone have hassles with the white one? Please discuss, thanks

Get the white one! The black one will probably go corrupt in the future!

 

[UnUnOctium]

How sure are you that the black modem won't still give you better speeds than the white even if not reaching full potential? Those of you giving advise here like experts do you actually know the transmit/receive power, internal electronics design, etc. to compare for both modems? Or at least have you done tests from the same location with both modems and did you get the same results?

More than likely that they both will have the same transmit power (standards are a beautiful thing). The 7.2 Mbps profile for HSDPA will give you a lower code rate and a smaller constellation size when compared to the 21.1 Mbps HSDPA+ profile. This translates to a lower BER for a given SNR.

Therefore, if you cannot get the full 7.2 Mbps for a given SNR (this remains approximately the same for both modems, explained later) then you can definitely not get 21.1 Mbps due to the profile being less robust than the 7.2 Mbps one.

Now, the difference in speeds from the modems will be small since the BS transmits at a fixed power level (probably around 30 dBm) and what then determines the SNR (CPE side) is the loss incurred by the channel (includes ICI, ISI, Path Loss) and the variations in electronics on the CPE modem. The channel loss is independent of the user's modem, so therefore the difference made between the 2 modems would only be in the quality of the electronics (front end + DSP). Now, my guess is (and this is a stab in the dark) that CellC use the same DSP for both modems just that the one is set to have a max profile with a rate of 7.2 Mbps and the other at 21.1 Mbps, I don't know if it's true (since I haven't opened either of the modems) but even if it wasn't, the noise added by the DSP would be minimal.

The difference therefore remains only in the front-end electronics, and this is tiny, maybe 1 or 2 dB.

Now if you reverse the argument (i.e. white modem better for when you can't get full speeds) it makes no difference since both modems support the same profile set for speeds below their maximum too.

 

[marlboro]

More than likely that they both will have the same transmit power (standards are a beautiful thing). The 7.2 Mbps profile for HSDPA will give you a lower code rate and a smaller constellation size when compared to the 21.1 Mbps HSDPA+ profile. This translates to a lower BER for a given SNR.

Therefore, if you cannot get the full 7.2 Mbps for a given SNR (this remains approximately the same for both modems, explained later) then you can definitely not get 21.1 Mbps due to the profile being less robust than the 7.2 Mbps one.

Now, the difference in speeds from the modems will be small since the BS transmits at a fixed power level (probably around 30 dBm) and what then determines the SNR (CPE side) is the loss incurred by the channel (includes ICI, ISI, Path Loss) and the variations in electronics on the CPE modem. The channel loss is independent of the user's modem, so therefore the difference made between the 2 modems would only be in the quality of the electronics (front end + DSP). Now, my guess is (and this is a stab in the dark) that CellC use the same DSP for both modems just that the one is set to have a max profile with a rate of 7.2 Mbps and the other at 21.1 Mbps, I don't know if it's true (since I haven't opened either of the modems) but even if it wasn't, the noise added by the DSP would be minimal.

The difference therefore remains only in the front-end electronics, and this is tiny, maybe 1 or 2 dB.

Now if you reverse the argument (i.e. white modem better for when you can't get full speeds) it makes no difference since both modems support the same profile set for speeds below their maximum too.

Lots of three letter abbreviations. Can you pull some more of those out of your sleeve? All I managed to gather is another "expert' advise, which comes down to: "More than likely"... "my guess is (and this is a stab in the dark)"... "I don't know if it's true".

 

[UnUnOctium]

Lots of three letter abbreviations. Can you pull some more of those out of your sleeve? All I managed to gather is another "expert' advise, which comes down to:

Actually, I do happen to be an expert, believe it or not. I can pull a million and one three-letter acronyms (not abbreviations). Whether you're too ignorant to actually try and comprehend is no fault of my own.

"More than likely"

I'm not working on CellC's network, I don't know if they're illegally bumping up transmit powers. It's the same "more than likely" that you'd say when you were asked if the moon has been around before you. You can give a highly probable guess but you can never be exactly certain since you might be the victim of a Truman-esque prank.

"my guess is (and this is a stab in the dark)"

This has more cost than performance ramifications, but the performance ramifications are linked to research and work done by Ludwig Boltzmann. Two semiconductors are so unlikely to ever have identical properties that the probability can be approximated as 0, this is in part because the effect actually originates from quantum physics but I'm not going there since I doubt you'd even know the difference between energy and power.

"I don't know if it's true".

This is related to the prior statement. I say this because I've never owned any of CellC's new USB modems to tinker and modify. I therefore cannot confirm for certain whether they would use the same DSPs (or even chipsets) for both modems. That has nothing to do with the operational principles and science behind it.

Now, I don't expect you to catch onto any of these concepts (matriculants aren't able to either, so don't worry) but I'm guessing Uncle Thabo said almost the same lines as you when he was offered research on AIDS and flat-out denied it.

 

[sajunky]

I fully agree with UnUnOctium. If your burst download speed is below 7.2Mbps, it is indication that 21Mbps modem uses more robust modulation methods, the same as the cheaper conterpart. It can be still faster than the cheaper modem due to the faster clocking of its electronics (esp. equaliser, DSP) giving lower processing noise, but it is expected to loose this advantage on the failed attempts of negotiating higher speeds. Also when doing mostly Web browsing the modem never get chance to negotiate maximum speeds.

We don't have detailed specification for these modems. We know that both modems have equalisers. The black dongle has receive diversity, but it doesn't say in which UMTS band. There is also an issue of problematic external antenna connection. White dongle do not have receive diversity, nor external antenna connector.

I am located on the cell edge with maximum download speed 1-2Mbps (also on Vodacom network with E272). From the beginning of this month I tested Vodafone K3565 (Huawei) modem, then for few days I could compare with CellC white dongle which gave me identical results. Not a conclusive results due to the short time before network became unstable, RSSI started changing and frequently disappear completely, also 2G became unworkable (fail to connect). Two different ordinary cellphones reported SIM registration errors (one CellC contract, the other Virgin prepaid). A week ago CellC gave up, now 3G signal is gone, but at least EDGE connection returned to the maximum speed (as before middle of June).

Now if I had to choose between CellC white dongle and Vodafone, I would pick Vodafone 3.6Mbps. It doesn't have equaliser, but it has receive diversity in both UMTS bands. It shouldn't be a challenge for UnUnOctium to explain us what is an advantage (if any) of the equaliser to the receive diversity. The real issue is that Vodafone has a hidden external antenna connector, so I can boost signal and get stable (maybe close to 3Mbps) download speeds.

 

[UnUnOctium]

I fully agree with UnUnOctium. If your burst download speed is below 7.2Mbps, it is indication that 21Mbps modem uses more robust modulation methods, the same as the cheaper conterpart. It can be still faster than the cheaper modem due to the faster clocking of its electronics (esp. equaliser, DSP) giving lower processing noise, but it is expected to loose this advantage on the failed attempts of negotiating higher speeds. Also when doing mostly Web browsing the modem never get chance to negotiate maximum speeds.

Thanks for the comment. Just have a few questions/comments of my own. Firstly, as far as I'm aware, processing noise isn't determined by the clock speed but rather the electronics quality (you might be confusing it with aliasing). They would obviously use the higher quality batches/yields for the 21.1 Mbps variants. Just to clarify my general point in my previous comment(s), since the speed is determined by fixed profiles and that both modems have the same fixed profiles up to their advertised speed, if a 7.2 Mbps cannot make its maximum speed, then a 21.1 Mbps will not make it either, only exception being if you make 7.1 Mbps or some nearby speed with the white one and the black one achieves 7.2 due to the higher quality electronics.

We don't have detailed specification for these modems. We know that both modems have equalisers. The black dongle has receive diversity, but it doesn't say in which UMTS band. There is also an issue of problematic external antenna connection. White dongle do not have receive diversity, nor external antenna connector.

As I have never opened/had these modems, I cannot confirm this on my own, but as far as I'm aware, neither of them should have receive diversity. This is because the 21.1 Mbps HSPA+ profile is the last one before MIMO/Dual-Cell is required. Any higher speeds and you need antenna diversity. Source: (http://en.wikipedia.org/wiki/Hsdpa#HSDPA_User_Equipment_.28UE.29_categories)

I am located on the cell edge with maximum download speed 1-2Mbps (also on Vodacom network with E272). From the beginning of this month I tested Vodafone K3565 (Huawei) modem, then for few days I could compare with CellC white dongle which gave me identical results. Not a conclusive results due to the short time before network became unstable, RSSI started changing and frequently disappear completely, also 2G became unworkable (fail to connect). Two different ordinary cellphones reported SIM registration errors (one CellC contract, the other Virgin prepaid). A week ago CellC gave up, now 3G signal is gone, but at least EDGE connection returned to the maximum speed (as before middle of June).

Now if I had to choose between CellC white dongle and Vodafone, I would pick Vodafone 3.6Mbps. It doesn't have equaliser, but it has receive diversity in both UMTS bands. It shouldn't be a challenge for UnUnOctium to explain us what is an advantage (if any) of the equaliser to the receive diversity. The real issue is that Vodafone has a hidden external antenna connector, so I can boost signal and get stable (maybe close to 3Mbps) download speeds.

I fail to see how a modem would not have an equalizer since the performance increase afforded by it is tremendous. I highly doubt it would be able to achieve those rates too without an equalizer, either way the Rayleigh/Rician channels will make the channel's frequency response non-uniform which would severely affect your received symbol's I/Q values. Usually LMMSE or ZF/LS equalizers are employed, especially the zero-forcing one since, while not as good as LMMSE, it provides a great improvement over a no-equalizer approach with a very small increase in complexity.

I doubt the Vodafone one would have receive diversity either (cannot confirm ofcourse) but it would make it non-standard if it did. If, all along, by receive diversity you meant the external antenna connectors please do note that if you plug in an external antenna, the internal, microstrip one is disabled.

On a side note: The advantages of antenna diversity are simple. The devices are able to transceive on each antenna independently. So a 2x2 system (2 transmit antennae, 2 receive antennae) allows you to double your rate, however a 3x3 or 4x4 etc will be less than triple and less than quadruple. This is because the 2x2 code has a rate of 1 (which no other STBCs have) and is therefore the optimal one (known as the Alamouti code).

 

[sajunky]

Thanks for the comment. Just have a few questions/comments of my own. Firstly, as far as I'm aware, processing noise isn't determined by the clock speed but rather the electronics quality (you might be confusing it with aliasing). They would obviously use the higher quality batches/yields for the 21.1 Mbps variants.

I don't comment on profiles, we already agreed to this point. However your further assumptions are not supported by experience.
I am aware of the fact that the white dongle works cool and black dongle runs hot and cause various problems in computers not supplying sufficient power. So the theory of electronics quality fails when confronted with these facts. It is rather faster processing clock or completely different hardware (or both) and I know that receive diversity require two separate receiving channels, demodulators and decoders. Quite a lot of extra hardware is involved.

As I have never opened/had these modems, I cannot confirm this on my own, but as far as I'm aware, neither of them should have receive diversity. This is because the 21.1 Mbps HSPA+ profile is the last one before MIMO/Dual-Cell is required. Any higher speeds and you need antenna diversity. Source:

Receive diversity was used for a while in E272 and K3565 for example. They are Release 5 devices. You seem to confuse it with 2x2 MIMO (introduced in R7) which involves dual transmission and receive channels. I would call it SIMO, but I don't use terms I don't understand.

I fail to see how a modem would not have an equalizer since the performance increase afforded by it is tremendous. I highly doubt it would be able to achieve those rates too without an equalizer

Equaliser is not specified in previous devices. I see many analog filters in these devices. There are present in RF and also in baseband circuits. So my understanding is that equaliser should replace some baseband analogue filters.

I doubt the Vodafone one would have receive diversity either (cannot confirm ofcourse) but it would make it non-standard if it did. If, all along, by receive diversity you meant the external antenna connectors please do note that if you plug in an external antenna, the internal, microstrip one is disabled.

Not at all. If I plug the external antenna to my E272 the internal antenna is disconnected, but there is also internal diversity antenna permanently attached. In E1820 the external antenna switches off the internal diversity antenna. It cause many problems - read this thread - there are things tested and verified by users:
http://mybroadband.co.za/vb/showthread.php/276198-Cell-C-E1820-External-Antenna-Connector

On a side note: The advantages of antenna diversity are simple. The devices are able to transceive on each antenna independently. So a 2x2 system (2 transmit antennae, 2 receive antennae) allows you to double your rate, however a 3x3 or 4x4 etc will be less than triple and less than quadruple. This is because the 2x2 code has a rate of 1 (which no other STBCs have) and is therefore the optimal one (known as the Alamouti code).

Once again you talk about 2x2 MIMO. Not all diversity schemes require hardware and network protocol involvement of the base station. There are many things we do not undestand, but we musn't confuse others.

 

[UnUnOctium]

I don't comment on profiles, we already agreed to this point. However your further assumptions are not supported by experience.
I am aware of the fact that the white dongle works cool and black dongle runs hot and cause various problems in computers not supplying sufficient power. So the theory of electronics quality fails when confronted with these facts. It is rather faster processing clock or completely different hardware (or both) and I know that receive diversity require two separate receiving channels, demodulators and decoders. Quite a lot of extra hardware is involved.

Indeed. It is definitely true that for the faster speed it will need faster processing. If the black ones do indeed run hotter than the white ones then the hardware isn't identical but my original point was that faster processing doesn't mean higher receive SNR. As with the receive diversity point, you do indeed have more hardware (depending on how many receive antennae you have) but the only extra bit is just the antenna and the front end (down-mixing, analogue filtering, any IF if you're superhet inclined). The demodulation for CDMA-based modulation systems is done in software by the DSP.

Receive diversity was used for a while in E272 and K3565 for example. They are Release 5 devices. You seem to confuse it with 2x2 MIMO (introduced in R7) which involves dual transmission and receive channels. I would call it SIMO, but I don't use terms I don't understand.

SIMO is indeed the correct term and then most likely what they're using.

Equaliser is not specified in previous devices. I see many analog filters in these devices. There are present in RF and also in baseband circuits. So my understanding is that equaliser should replace some baseband analogue filters.

Nop, the equalization is done in software, analogue filters will still be used for the front-end. Since they're using W-CDMA, the DSP despreads the signal with the spreading sequence assigned to it by the base-station. The base-stations transmit pilots so that the channel frequency response can be effectively equalized to ensure optimal detection, this is all done in software on the DSP. This can be done before or after despreading (makes little difference) but can only be done in software.

Not at all. If I plug the external antenna to my E272 the internal antenna is disconnected, but there is also internal diversity antenna permanently attached. In E1820 the external antenna switches off the internal diversity antenna. It cause many problems - read this thread - there are things tested and verified by users:
http://mybroadband.co.za/vb/showthread.php/276198-Cell-C-E1820-External-Antenna-Connector

I see. Very interesting arrangement though since the antenna gains will be immensely different.

Once again you talk about 2x2 MIMO. Not all diversity schemes require hardware and network protocol involvement of the base station. There are many things we do not undestand, but we musn't confuse others.

You asked me to explain antenna diversity SIMO is more of a stepping stone towards MIMO and MISO is rarely used.

P.S. the "Thanks for the comment" wasn't sarcastic, glad to meet a person who is also in the know. Do you do this for a living or are you also a student?

 

[sajunky]

P.S. the "Thanks for the comment" wasn't sarcastic, glad to meet a person who is also in the know.

Of course not, I didn't receive it that way. I fully enjoy this conversation, albeit of some pinch of annoyance from drawing out wrong assumptions, I must confess. I found your technical comments very informative.

Do you do this for a living or are you also a student?

We are always students. In home, school, our own life and work. When we stop learning, we are dying.

 

[Dolby]

I don't really understand wtf you guys are saying

But this black modem runs very hot and if its a power issue, would using a USB splitter for 2 ports cool it down?

 

[UnUnOctium]

I don't really understand wtf you guys are saying

But this black modem runs very hot and if its a power issue, would using a USB splitter for 2 ports cool it down?

No, it's designed to run like that. If anything (if it doesn't have cooling) more power = more heat

 

[Cell.C-Guru]

I don't really understand wtf you guys are saying

But this black modem runs very hot and if its a power issue, would using a USB splitter for 2 ports cool it down?

its normal for 3G modems to get very hot. Nothing to worry about

 

[RichardG]

I have 2 USB modems BLACK and WHITE. I am able to pick up 3G Signal using the white device but the black device doesn't pick up any 3G using MDMA, not sure if this could be to new base station soon to be switched on in January or something but my RSSI on the white dongle is -87 dBm and I am able to download at around 4 Megs roughly non-stop without any glitch sites loading super fast...

I would prefer anyone to go with the white dongle, I was pretty shocked that the white dongle has more power than the black dongle...

BTW I am not sure where exactly the base station I connect 2 as I have picked up only 1 with 3G the others have EDGE

 

[ginggs]

I have 2 USB modems BLACK and WHITE. I am able to pick up 3G Signal using the white device but the black device doesn't pick up any 3G using MDMA, not sure if this could be to new base station soon to be switched on in January or something but my RSSI on the white dongle is -87 dBm and I am able to download at around 4 Megs roughly non-stop without any glitch sites loading super fast...

Try the procedures in posts 2 and 6 in this thread and see if that gets you back onto 3G.
Also make sure you are using the latest MDMA.

 

[RichardG]

I have the latest version of MDMA, if i don't force the connection on the white dongle I will only pick up edge connectivity have tried both computers with different OS Windows XP and Windows 7, the white dongle has no issues whatsoever but the black dongle only picks up EDGE, but have noticed that the black dongle using MDMA forcing 3G is picking up a few single but without any LAC, CELL, RNC etc. I am thinking 1 of 2 possible causes. 1 firemwire/drivers needs to be updated - but have been on CELLC websites but its useless the drivers page hasn't been updated. 2 The base station is probably about 5KM+ and the white dongle can detect it but not the black, could be something to do with power...

I had no issues with the black dongle at emperors have tried all the methods as mentioned and I am not getting any glory in receiving 3G, not sure if the dongle could be faulty but many factors have to be taken into consideration..

If I knew where base station 101 is located could give me rough idea how far I am from it cause when using edge I pick up base station 105, 108

UPDATE:
Finally it seems I am able to pick up 3 G connectivity with the black dongle - soon after I posted i left the black dongle on the other notebook and by a miracle WCDMA switched on I have faith in CELL C - They do read the forum after all

 

[fred1]

Latency(ping) difference between 21.6 and 7.2mbps

Hi
Can anyone tell me the difference in latency please.
One to buy the special today and going to use it on xbox live.
Thanks!