Anyone Using an Asus DSL-AC68U

Anyone got that email address, so we can all mass send an email to him with our requests.

@ixel
After i did a reboot the other day my sync came in at about 72, then after turning TCM off with the code you provided. My router gui was stating still 72, but when i run code in tcm to pull up the stats my sync was changed to 58,
Do you think with TCM disabled that it can effect sync? or do you think its dlm?
I don't mean to confuse more the things but seems asus sending (even for stats) telnet commands to the chipset ... example the spectrum... it may be a conflict there via different commands or one effecting the other ... but i am not expert on this with asus

Let know asus about it, at the email provided
Also will be good idea to send mass mails ;)
 
Although I couldn't find it listed, I modified the URL of 2049 and it downloaded 2160. Looking through it I've noticed that a process called tp_init might be responsible for communicating with the modem (router <-> modem), so as an experiment I will be killing 'tp_init' to see if it has any effect on the modem's performance (e.g. less errors).
 
Regarding 'tp_init', this process does indeed communicate from the router to the modem and is responsible for things such as fetching the statistics and possibly applying changes in settings to the modem (possibly!). I believe it's also responsible for other functionality such as 'dynamic line adjustment', to monitor and apply new settings as necessary. I'll need more time to determine if killing it does really help improve the modem's performance slightly or significantly but I doubt it would.

EDIT 1: No difference to FEC error count going wild. I'll continue comparing the CRC errors to see if there's much difference (I doubt it) with varying levels of INP and delay settings (up to 80Mbps, target SNRM 6.0 dB). I'll limit re-syncs/changes to once per day.
 
Last edited:
I can not see or find the 2160 source on any asus website either only the .com site has any source code and its for a yonks old version, as some also mentioned (i think) before theres no ADSL driver (ive not bothered to check again if that is the case or not, 700+Mb download for nothing soon gets old ;) ).


EDIT: Have possibly found it as Ixel briefly mentions...
http://dlcdnet.asus.com/pub/ASUS/wireless/DSL-AC68U/GPL_DSL_AC68U_30043762160.zip
Not downloaded to confirm yet, interesting they have not placed this on there website anywhere, so wonder if it is the real and complete thing.

Will download later and compare it to the old listed 2049 for differences.

Anyone able to confirm or deny so far if the DSL drivers are included???
 
Last edited:
I can not see or find the 2160 source on any asus website either only the .com site has any source code and its for a yonks old version, as some also mentioned (i think) before theres no ADSL driver (ive not bothered to check again if that is the case or not, 700+Mb download for nothing soon gets old ;) ).

The link I said earlier should get you 2160. It does have a tclinux.bin inside it, but the actual modem firmware isn't open source as I'd expect.
 
The link I said earlier should get you 2160. It does have a tclinux.bin inside it, but the actual modem firmware isn't open source as I'd expect.

Im not sure will have to check but im pretty sure that violates the licence all MT based chipsets and drivers are open AFAIK

They have removed the source codes, not sure if they will add them again
Strange seems have seen the posts here lol

Click my link on prior page still there :)

EDIT:

Old and new source is on the UK site now.......
http://www.asus.com/uk/Networking/DSLAC68U/HelpDesk_Download/
select others, then source.
 
Last edited:
Inspired by other peoples success with different settings to me I had another go with this modem yesterday.

I forced it to Annex A rather than using the A/I/J/L/M option, and disabled the SRA Rate Adaptation that was on by default. My settings are now everything disabled other than

Annex: A
Rx AGC GAIN Adjustment (VDSL): Stable
Bitswap (VDSL): Enabled
VDSL Profile: 17a Multi Mode
G.INP (G.998.4): Enabled

This is resulting in a stable line with minimal CRCs (11 in 3.5 hours).

However, I think there may still be issues. DLM knocked me back down from 67Mb to 54Mb this morning even though I had no disconnects yesterday (other than switching the modem) so I'm suspecting that there is a high level of Errored Seconds going on under the hood causing DLM to object.

Ill run it today and see if I get him by another DLM knock down tomorrow morning, but as far as the router stats screen goes, this seems to be operating properly at the moment.
 
I whipped it out earlier today as well.
Reloaded the 2160 with driver firmware, factory reset and setup from scratch.

Settings as per PlumpMonkey but I left Rx AGC GAIN as Disabled as I didn't find it did much when I had it set to Stable.

Connection currently Fastpath.

Uptime is 2 hours, downstream CRC count is 2616, downstream FEC 31834 no HEC. All this when the connection wasn't actively being used - I went outside to change my wheels over to winters and when I left it was 5 CRC and came back in later it was up at the 2616 mark.


With the HG612, I averaged 10,000 CRC per day and less than 300 ES+SES combined per day. So even if the CRC is higher as long as it's not a magnitude of 10 higher then I may be OK running the DSL-AC68U on fastpath. We'll see how today progresses.
 
Ahh yeah, that reminds me. The the other change between the last time I tested this and yesterday was that I now appear to be on Interleaved rather than Fast Path which might make things more stable.

Ill post back tomorrow with how its got on.
 
Just noticed two people inside the fibre cab over the road with a lot of equipment, I wonder what they're up to - can only assume it's possibly some upgrade work maybe ;).

EDIT: Because of the G.INP being rolled out, until my ECI /r modem has updated its firmware I'll be staying on that (which might also trigger BT's DLM into then believing my connection is G.INP capable and assigning me a retransmission profile). The ASUS works fine here, provided I have a light level of error correction, but I get similar hundreds of ES daily on my HG612 too. Either it's not a high quality line, I'm getting some noise (crosstalk perhaps) or it's related to using a different modem chipset in comparison to the DSLAM's chipset (ECI cabinet here, so Infineon).
 
Last edited:
Inspired by other peoples success with different settings to me I had another go with this modem yesterday.

I forced it to Annex A rather than using the A/I/J/L/M option, and disabled the SRA Rate Adaptation that was on by default. My settings are now everything disabled other than

Annex: A
Rx AGC GAIN Adjustment (VDSL): Stable
Bitswap (VDSL): Enabled
VDSL Profile: 17a Multi Mode
G.INP (G.998.4): Enabled

This is resulting in a stable line with minimal CRCs (11 in 3.5 hours).

However, I think there may still be issues. DLM knocked me back down from 67Mb to 54Mb this morning even though I had no disconnects yesterday (other than switching the modem) so I'm suspecting that there is a high level of Errored Seconds going on under the hood causing DLM to object.

Ill run it today and see if I get him by another DLM knock down tomorrow morning, but as far as the router stats screen goes, this seems to be operating properly at the moment.

ES can only be equal or lower than the number of CRC's I believe.
 
ES can only be equal or lower than the number of CRC's I believe.

Yep if things are being reported correctly. That AFAIK is correct. The other figure people have issues with, IE FEC people on fastpath should not even be getting. I think it is pretty clear by now the FEC figure can good as be ignored and to get the ES figures its best to get them from TC Console as Ixel has previously demonstrated. Thos based on his observations do seem to be pretty accurate.

People that were on fastpath which suddenly get interleaving applied i am guessing happens time to time when the DLM system gets confused when you make major changes to a setting such as the RX gain,SRA,Annex etc. It should not make a difference but ive seen a couple of reports now where people have made setting changes to suddenly get DLM'd. Typically unless you have seriously broken something DLM will adjust the line back up within 2-3 weeks. (at least thats my experience with it when i have been interleaved).

Further edit: one possible way to maybe stop that happening would be to disconnect the connection (IE unplug modem lead) make changes you want to the device, then plug cable back in for it to connect, DLM should then just see that as you unpluging things and plugging them back in, as long as you do not do it many times that should have no DLM affect.

Just noticed two people inside the fibre cab over the road with a lot of equipment, I wonder what they're up to - can only assume it's possibly some upgrade work maybe ;).

EDIT: Because of the G.INP being rolled out, until my ECI /r modem has updated its firmware I'll be staying on that (which might also trigger BT's DLM into then believing my connection is G.INP capable and assigning me a retransmission profile). The ASUS works fine here, provided I have a light level of error correction, but I get similar hundreds of ES daily on my HG612 too. Either it's not a high quality line, I'm getting some noise (crosstalk perhaps) or it's related to using a different modem chipset in comparison to the DSLAM's chipset (ECI cabinet here, so Infineon).

The work may have been a new card being installed, was your cabinet full already? You can sometimes check and get a rough idea at...
https://www.btwholesale.com/includes/adsl/adsl.htm
Use the postcode and address check and enter just your postcode in both, if it says FTTC is not available, does not have it listed or has an estimate of 0Mb or similar silly figure then it is possible your cabinet was full and they were adding a new card to it.
 
Last edited:
Back
Top Bottom