Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[jamesbob]

I have a a broadband package from Zen.

I have been using the AVM Fritzbox 7530 for a long time with almost no problems.

I upgraded a few weeks ago to the AVM Fritzbox 7530-AX - marketed as the 7530 but with wifi-6.

The 7530AX led to regular DSL disconnects and reconnects. I took a closer look and saw the 7530AX synced slower and was running with what looks like a huge number of CRC errors:

7530AX

Attainable           60238 (down)  8673 (up)
Current               59132             8550
SNR dB               6                    6
line attenuation   21    27
ES 532
CRC. 5/min  294/15mins

7530

Attainable           66240 (down)  8814 (up)
Current               64496             8584
SNR dB               7                    6
line attenuation   21    27
ES 27
CRC.  0/min  0/15mins


The new 7530AX is syncing slower and has an alarming number of unrecoverable errors (at the fritzbox, not the exchange, whatever that means).

The older 7530 syncs faster and reports ZERO CRC error rate, but does report a small number of seconds with error (inconsistent?).

Interesting that the 7530 also has a download SNR of 7 not 6.

No other changes have been made to the house wiring, local LAN or anything else. Even the DSL filter, power cable and DSL cable are the same. The only thing I'm swapping is the model.

Question

Is this a faulty 7530AX unit? Is the difference explained by the modem chipset - Intel/Lantiq for the 7530 but Broadcom for the 7530AX?

Should I therefore go for a 7590AX which has an Intel DSL chipset too, but not the same as the 7530?

Could there be another explanation? What really stuck out for me was the period DSL drops and the apparently high number of unrecoverable CRC errors vs zero with the older modem.

[tubaman]

Different chipsets will sync at different speeds and produce different numbers of errors and that is entirely normal. Whether Lantiq or Broadcom will work best on your line you'll only know by trying them.  What I'd say is not normal here is getting drops and the number of errors you are seeing from the Broadcom chip when the ECI is coping ok. It could be a faulty modem or could be a design and/or firmware issue on the 7530-AX.
On a side note, unless you are transferring lots of data between devices on your LAN I'm not sure what advantage Wi-Fi 6 will give you over Wi-Fi 5 as both will happily outpace your VDSL connection speed, however that's probably just my 'if it ain't broke don't fix it' side kicking in. 

[jamesbob]

Different chipsets will sync at different speeds and produce different numbers of errors and that is entirely normal. Whether Lantiq or Broadcom will work best on your line you'll only know by trying them.  What I'd say is not normal here is getting drops and the number of errors you are seeing from the Broadcom chip when the ECI is coping ok. It could be a faulty modem or could be a design and/or firmware issue on the 7530-AX.
On a side note, unless you are transferring lots of data between devices on your LAN I'm not sure what advantage Wi-Fi 6 will give you over Wi-Fi 5 as both will happily outpace your VDSL connection speed, however that's probably just my 'if it ain't broke don't fix it' side kicking in. 

Thanks for replying Tubaman. Both fritzboxes report the exchange end as "Infineon: Version 13.2.6" - how do I interpret this? Does it also suggest non-BCM modems are better matched?

The idea behind wifi-6 is to better divide the service between about 12 connected wifi clients, 3 of which do streaming (video, zoom, etc) at the same time. I'm also told wifi-6 will extend the range for wifi-6 clients. I wanted to try it to see if these benefits were real.

Does anyone know from an electrical engineering perspective why so many errors would be detected by one modem and not the other. Is the line itself intrinsically poor?

[meritez]

Interesting, no longer supports ADSL1
https://boxmatrix.info/wiki/FRITZ!Box_7530_AX

Definitely a Broadcom chipset

[tubaman]

Thanks for replying Tubaman. Both fritzboxes report the exchange end as "Infineon: Version 13.2.6" - how do I interpret this? Does it also suggest non-BCM modems are better matched?
...

That's tells us you are connected to an ECI cabinet ( ), which is what I suspected as your downstream SNR is 6dB, implying no G.INP.
The matching chipset to cabinet thing is a myth and it really does depend on your particular line which will perform better. What is odd in your case is that the Broadcom is performing so badly in comparison to the Lantiq chip.

[jamesbob]

What is odd in your case is that the Broadcom is performing so badly in comparison to the Lantiq chip.

Is this enough information to tell the manufacturer that this modem is faulty?

Especially as the 7530 does not see any CRC errors.

[j0hn]

Is this enough information to tell the manufacturer that this modem is faulty?

Especially as the 7530 does not see any CRC errors.

It's not faulty. It just doesn't perform as well as the 7530 on your line and your cabinet vendor combination.

Your line may run better with Lantiq chipsets rather than Broadcom. Your cabinet and the 7530 both have Lantiq chips.

They will be able to swap it back to the 7530 or might have another solution, but the 7530ax is usually what they hand out.

The 7590ax looks pricey.

[Alex Atkin UK]

That's tells us you are connected to an ECI cabinet ( ), which is what I suspected as your downstream SNR is 6dB, implying no G.INP.
The matching chipset to cabinet thing is a myth and it really does depend on your particular line which will perform better. What is odd in your case is that the Broadcom is performing so badly in comparison to the Lantiq chip.

I wouldn't say its a myth given it DOES apply on some lines, it just doesn't apply to all lines.  It would only be a myth if matching the chipsets never made any difference.

[Edinburgh_lad]

I presume he means it's a myth in relation to matching chipsets - modem and cabinet - which I agree is a myth. For example, on my line, with no interference present (on the line), I have no problems running Fritzbox 7530 connected to Huawei cabinet. However, when I have interference on the line, then the connection drops, but I didn't notice it dropping significantly more than when I run a Zyxel. And so, the quality of the line comes into play as well as lots of other factors, for example quality of firmware etc. (for example, my Draytek router dropped the connection more than the Fritzbox, even though they have the same Lantiq chipset). Plusnet router was the worst of them all.

[tubaman]

My understanding is that when FTTC first rolled-out, and Openreach supplied the modems, they went to great efforts to ensure that users connected to Huawei cabinets received Huawei modems and those connected to ECI cabinets received ECI modems. This was done as it was the belief at the time that matching modem chipset to cabinet chipset would work the best. Over time it has been demonstrated that this matching does not always give the best results, but we still see many people thinking that it is what they should do. Hence my saying it is a myth, one definition of which is "an unfounded or false notion".

[jamesbob]

I've been having DSL problems apparently after upgrading to my new AVM Fritzbox 7530AX, upgraded from the 7530 without AX.

The problems may not be related to the modems but to something happening on the line. The 7530AX has a Broadcom DSL chipset whereas the 7530 has a Lantiq which matches the Infineon DSLAM apparently.

The following DSL spectra are taken during the regular episodes of high CRC errors and DSL line drops.



Here the DSL is down and we can see there is zero signal for the send direction. Question - what could be causing this?




Here the DSL is back up a few moments later and we have some signal for the send direction.



Here the DSL is back up and has been running for longer. We can see the signal for. the send direction extends a little further into the higher frequencies. What would cause this?


I am hoping experts on this forum will look at the spectra and say "aha, that looks like X or Y is a likely cause".

My feeling or hunch is that this is rainwater causing the problem as the worst performance happens when it rains, but I haven't kept a diary to be totally scientific about it.

[tubaman]

Assume this is effectively a continuation of this thread - https://forum.kitz.co.uk/index.php/topic,27378.html - where the thought appears to be that unfortunately the 7530AX just isn't suited to your particular line?

[Moderator note: Both topics now merged into one (without adjusting the subject lines).]

[jamesbob]

Assume this is effectively a continuation of this thread - https://forum.kitz.co.uk/index.php/topic,27378.html - where the thought appears to be that unfortunately the 7530AX just isn't suited to your particular line?

Hi tubaman .. it is the same 7530AX but I'm trying to see why for some periods of time the modem is fine, and then for some periods of time the modem doesn't like the line.

The lack of a send signal whilst having an almost full receive signal is intriguing and might suggest a mechanism - eg rain, or a loose connection. A full loss of signal up and down would be less of a clue to the mechanism.

The SNR also drops below 6dB for up/down during these episodes if that helps.

[parkdale]

You may have already done a quite line test? 17070 option 2, is there any noise on the line? test again when raining.

[Edinburgh_lad]

I had a similar issue but it was present regardless of the router used, so on Fritzbox, Draytek etc. This temp. loss of signal occurred sporadically. I don't know exactly what's caused it because even though I had a specialist team here following my CEO complaint, they never told me the reason. During the investigation, however, they found the issue with a lightning protection module, which after replacing did improve the connection in that loss of signal didn't occur as often (there's a wee wire in it and it was broken in our case; probably following a lightning strike), but also did some underground work (loose joint?). They also said something about SHINE being present in my case but claimed their equipment wasn't picking it up, though the issue clearly shows on the ISP's system and on my Fritzbox, but then it's Openreach so they'll even claim that pigs have wings.

What I find interesting is that in your case, you have no pilot reference, but it occurs on my spectrum.

Also, what version of Fritzos are you running?