Kitz Forum
Broadband Related => Broadband Hardware => Topic started by: snadge on July 26, 2020, 10:05:35 PM
-
can anyone tell me what the differences are between these two routers? the web-pages indicate them to be exactly the same, both Dual-Band Ac1300..
cheers
-
The 8324 does not have AC WiFi. Look on the specifications page and you'll see only the 8924 has 5Ghz WiFi
-
According to their web page it does?
https://www.zyxel.com/uk/en/products_services/Wireless-N-VDSL2-VoIP-Combo-WAN-Gigabit-IAD-VMG8324-B10A/
-
Edited my previous reply whilst you were typing.
Look on the specifications page and you'll see only the 8924 has 5Ghz WiFi
https://www.zyxel.com/uk/en/products_services/Wireless-N-VDSL2-VoIP-Combo-WAN-Gigabit-IAD-VMG8324-B10A/specifications
https://www.zyxel.com/uk/en/products_services/Dual-Band-Wireless-AC-N-VDSL2-Combo-WAN-Gigabit-IAD-VMG8924-B-Series/specifications
-
their webpage is wrong? its title has AC/N dual-band
https://www.zyxel.com/uk/en/products_services/Wireless-N-VDSL2-VoIP-Combo-WAN-Gigabit-IAD-VMG8324-B10A/
strange...
-
Yes it is wrong.
-
you wouldnt happen to have the ZyXEL 8924-B10A SoC specs on hand would you?
cheers
-
I can get router specs but not what SoC it is, all i know is it is a broadcom chipset, possibly 63168 but what specs are they set at? ie. RAM, eMMC, CPU speed/core count etc
-
looking at the 63168 specs tells me the 8324 maybe 63168 but the 8924 has specs that dont coincide with it? namely AC 3x3 and USB3.0 x2 - the 63168 specs stipulate only N 2x2 wifi and USB2.0 x1..?
-
This is a photo of the front side of the PCB of my VMG8924-B10A. The command "cat /proc/cpuinfo" reports "system type : 963168VX" so I can confirm that the SoC is indeed a BCM63168. I also know that the VMG8324 and VMG8924 share firmware images, so are very similar.
As I understand it, the only difference between the two, as said by Ronski, is that the VMG8324 does not have 5GHz WiFi and the VMG8924 does. This is reflected by the mini-PCIe port (bottom left in the photo) not being populated on the VMG8324 (this is how the VMG8924 is able to offer AC 3x3 when the chipset does not support it, since it is provided for with separate hardware). The VMG8924 also does not feature USB3.0 (the ports are blue confusingly, but only have 4 contacts instead of the requisite 5 + 4 for USB 3.0). Both the VMG8924 and VMG8324 have two USB ports which are provided from the one offered by the BCM63168 using a USB hub chip (I believe this is the one to the left of the lower USB port in the photo, but I can't read the part number on my photo and am currently not in a position to open it to confirm).
As for more detailed specs, there are two cores (see cat /proc/cpuinfo below) and 128MB RAM (see cat /proc/meminfo below). The eMMC is a TC58NVG0S3HTA00 which is 1Gbit (so 128MB). I don't know the CPU core frequency I'm afraid.
~ # cat /proc/cpuinfo
system type : 963168VX
processor : 0
cpu model : Broadcom4350 V8.0
BogoMIPS : 398.33
wait instruction : yes
microsecond timers : yes
tlb_entries : 32
extra interrupt vector : no
hardware watchpoint : no
ASEs implemented :
shadow register sets : 1
core : 0
VCED exceptions : not available
VCEI exceptions : not available
unaligned exceptions : 2488
processor : 1
cpu model : Broadcom4350 V8.0
BogoMIPS : 402.43
wait instruction : yes
microsecond timers : yes
tlb_entries : 32
extra interrupt vector : no
hardware watchpoint : no
ASEs implemented :
shadow register sets : 1
core : 0
VCED exceptions : not available
VCEI exceptions : not available
unaligned exceptions : 2488
~ # cat /proc/meminfo
MemTotal: 123664 kB
MemFree: 20392 kB
Buffers: 0 kB
Cached: 31164 kB
SwapCached: 0 kB
Active: 12532 kB
Inactive: 27176 kB
Active(anon): 8544 kB
Inactive(anon): 0 kB
Active(file): 3988 kB
Inactive(file): 27176 kB
SwapTotal: 0 kB
SwapFree: 0 kB
Dirty: 0 kB
Writeback: 0 kB
AnonPages: 8544 kB
Mapped: 7340 kB
Slab: 51320 kB
SReclaimable: 920 kB
SUnreclaim: 50400 kB
PageTables: 624 kB
NFS_Unstable: 0 kB
Bounce: 0 kB
WritebackTmp: 0 kB
CommitLimit: 61832 kB
Committed_AS: 18668 kB
VmallocTotal: 1032148 kB
VmallocUsed: 11060 kB
VmallocChunk: 966916 kB
-
you wouldnt happen to have the ZyXEL 8924-B10A SoC specs on hand would you?
cheers
No I haven't, but it looks like some one else has supplied the relevant info.
-
cheers guys/gals
so taking from the information above, WikiDevi mirror (https://wikidevi.wi-cat.ru/Broadcom) and on OpenWRT the spexifications for the 8924/8324 is:-
SoC : Broadcom 63168
CPU : Broadcom 4350v8 @ 400Mhz (dual-core)
RAM : 128MB DDR2
ROM : 16MB / 64MB / 128MB (unsure what the amount is for this model but all others are 16-128)
USB : 2.0 x2
LAN : Gigabit x4
WAN : Gigabit x1
WLAN : AC1300 (8924 only) / N300
-
The VMG8924 also does not feature USB3.0 (the ports are blue confusingly, but only have 4 contacts instead of the requisite 5 + 4 for USB 3.0).
The XyXEL VMG3925-B10B modem/router has a single USB port which is also colour coded blue but I believe is USB-2 standard, indeed confusing and a bit naughty on ZyXel's part.
I've checked the USB socket on my VMG3925-B10B and can only see the 4 connector pins, having said that, looking at my RPi 4B which has 2 X USB-2 ports and 2 adjacent USB-3 ports the pin count looks the same, can not easily see the extra 5 pins associated with USB-3 standard.
-
are their any routers (OR MODEMs) with better Broadcom chipsets that allow telnet, better features and / or similar connection speed etc..?
fancy trying a different one and see if my 16GB loss comes back UNLIKELY but i want new hardware
any recommendations guy? Im a bit behind with the times
-
The BCM63168 still performs better than any other chipset.
The BCM63138 is newer but synced at an almost identical rate (a few kb lower).
There's nothing Broadcom based that will bring that 16Mb back.
Crosstalk can only be clawed back via Vectoring which isn't even possible on your ECI cabinet.
You're only likely to achieve a bump in sync with a Lantiq chipset and manually tweaking the SNRM down.
With FTTP coming I'd recommend not spending money on any additional xDSL hardware.
-
The BCM68168 still performs better than any other chipset.
Which modem's got that chipset?
-
Which modem's got that chipset?
Sorry typo. That was meant to be the BCM63168.
The BCM63138 is the newer chipset.
-
I have a Plsunet Hub One with a Lantiq chispet WILL THAT HELP AT ALL?
THANKS
-
Do you mean help with getting the 16Mbps back? :no: no modem/router would get it back since cross-talk is a side effect of FTTC which can not be dealt with from the user side by any means, it is the owner of cabinets (OR for ex. in UK) to apply vectoring which can greatly reduce the cross-talk effect.
I am sorry but if the current hardware is not having modem / router functioning problems then you should not spend money on another one, the only difference between old and new one (except for better wifi or additional features) will not be other than +/- few Mbps due to using a different chipset (or different PCB design in case of same chipset), even the Plusnet Hub 1 you have could lead to a DLM kick in due to the chipset change but 100% it will not get the 16Mbps back.
-
it is the owner of cabinets (OR for ex. in UK) to apply vectoring which can greatly reduce the cross-talk effect.
Never going to happen on FTTC VDSL systems in the UK now beyond the very few trials already implemented, has been debated at length in other topics.
I believe vectoring is already active by default in G.Fast technology to the very few that were able to take advantage of that system, further uptake currently on hold.
(I'am sure I will be corrected if either of these statements are inaccurate)
-
Never going to happen on FTTC VDSL systems in the UK now beyond the very few trials already implemented, has been debated at length in other topics.
I believe vectoring is already active by default in G.Fast technology to the very few that were able to take advantage of that system, further uptake currently on hold.
(I'am sure I will be corrected if either of these statements are inaccurate)
You are totally correct, I asked before about why OR is not applying vectoring since some lines are relatively long + cross-talk users end up with decreasing speeds over time and I remember the answer was vectoring costs more money to be deployed nation wide and since OR focusing on FTTP plans there is no sense in pushing vectoring... so pretty much internet users are stuck with what speeds they get at the moment...
* I were pointing to that end users can not do anything about cross-talk whatever the hardware they have as it only can be dealt within cabinets.. I should have said that it is hardly going to happen by OR, so thanks tiffy for making it clear :)
-
just my zyxell say max rate 77MB down and 25MB up yet only gets 63MB down 6.3db SNRM
-
just my zyxell say max rate 77MB down and 25MB up yet only gets 63MB down 6.3db SNRM
I'm sure you know, but max rate can be way off unless you have fastpath. I can literally change it on my HomeHub 5A by tweaking the target SNRm to something completely unrealistic.
My interpretation is that max rate is an estimate based on a perfect line with the current line attenuation and target SNRm, whereas the real sync rate will come out lower if the bit loading is less than perfect due to line quality/crosstalk.
-
From the signature and I guess it is the current situation, line is Interleaved for some reason either as one of the DLM mysteries or something happened on the line that caused it (and maybe even if this thing is fixed but line is stuck on interleaved) I believe cross-talk does not trigger Interleaving.
Although somehow my line max did not change before and after fastpath was applied. (outside UK)