ECI modem question

Having moved to BT Infinity from virgin (and really not wanting to change my router) I purchased a refurbished openreach ECI modem online. I am now just waiting for my service to start.

But after reading all the issues with G.INP (since I don't know if the firmware has been updated) will it even connect or should I just return it and eat the P&P?

Feeling pretty stupid at the moment....

It will connect but won't support G.INP

G.INP is only available if your cabinet is a Huawei anyway.

Send it back as unsuitable. You want an HG612, as that will give you full access to line stats for diagnostic purposes.

I thought downstream G.INP, which is currently all that OR provide, was now being rolled out on ECI cabinets. I could very well be wrong.

Finding out who manufactures your Fibre Cabinet is the first thing required here... do you know what cabinet & exchange you are connected too... you can find this at the following link by entering your address (Not Postcode, As its no reliable)

BT Wholesale Checker

If your cabinet is manufactured by ECI then you've no need to worry, if your cabinet is manufactured by Huawei, you still have little need to worry but may find better performance by using the correct hardware to match the cabinets hardware.

Well, no evidence of it yet

In reply to a post by Stubabe:

Feeling pretty stupid at the moment....

Don't!

It's a very messy and complex subject.

After using the BT Checker link given, my FTTC broadband is served by BT Cabinet 1 from Exchange Basildon. I am using an ECI Open-Reach Modem and how do I know where about is my Cabinet 1, and wherever it is, is ECI or HC cabinet, such as individual features ?? Do I need to physically "locate" them at all and how???

Many thanks

This link tells you how to find out http://www.kitz.co.uk/adsl/cabinet-lookup.htm

It's an ECI cabinet.

beat me to it lol

Thanks to everyone, it is a Huawei cab that's why I was asking. The cost to return the thing isn't worth it unless it wouldn't sync due to old firmware or something. I am getting the 40/10 service and the line checker suggests 30/7 min even for a degraded line. As much as I would like to geek out on them I can live without line stats (assuming nothing goes wrong!) The main point is I didn't want a [censored] HH5. I only went with BT against my better judgement due to massive quidco cashback and the current offers, otherwise it would have been zen or AAISP - moving house is expensive...

As it's a Huawei cab you'll be much better off with a Huawei HG612 which will give you G.INP. You won't get that with the ECI modem or a HH5A. You would with a HH5B but the wifi thruput is dire.

You should be ok with the ECI modem, and who knows in future you may opt to upgrade to an all in one device that supports full capabilities of the Huawei DSLAM (cab)

Meanwhile, its not really talked about much as most seem to abandon the ECI modems on Huawei cabinets, but G.INP is supposed to work with ECI devices or the HH5A (Which uses same or very similar chip) however this is on downstream only (though its not being activated on upstream anymore)

I would give the ECI a go and consider an all in one device in future maybe.

In reply to a post by mlmclaren:

but G.INP is supposed to work with ECI devices or the HH5A (Which uses same or very similar chip) however this is on downstream only (though its not being activated on upstream anymore)

I have seen you trot out this incorrect information before. As I've already said, G.INP doesn't work on the supplied ECI modem or the HH5A.

SIN 498, modems and routers must support retransmission in the downstream, but it is not a mandatory requirement in the upstream. Taking this into account and to maintain a high level of service, when we first rolled out retransmission to our Huawei estate we introduced interleaving on all lines where the modem/routers did not support retransmission in the upstream.

SIN498 Requires devices be G.INP capable on Downstream, it was only when upstream was enabled with G.INP that everything went wrong.

I've witnessed my neighbours HH5a see an increase of 9Mb/s* since G.INP Phase 2 was rolled out on our DSLAM.

* note this is 9Mb/s faster than when the modem was connected in fast path mode and before any G.INP was rolled out on our DSLAM.

Lol

As I said last time, mine doesn't. You should buy your neighbour a HG612 for Christmas, they'll love it.

and because yours don't work, it means everyones won't then??

You don't even have one and you're not a BT customer.

Whats your point.... this don't mean nothing!

I mean that you don't know what you're talking about.

Ok, well I could say the same about yourself really...

Considering the number of connections I monitor and have access to remotely and physically a couple of them are BT connections too... two using HH5's

Anyway, better things to be doing with my day than arguing with some arrogant [censored] on a forum...

ciao

The ECI modem does support G.INP on the downstream?

Not without the hacked firmware.

Are you making this up as you go along....

Hacked Firmware!

Unless you mean by changing a setting on the firmware by unlocking it then why would Openreach just not change this setting remotely... like when they released a firmware update!

In reply to a post by mlmclaren:

Are you making this up as you go along....

No

Hacked Firmware!

What about it?

Unless you mean by changing a setting on the firmware by unlocking it then why would Openreach just not change this setting remotely... like when they released a firmware update!

You better take that up with Openreach.

No such thing as hacked firmware.... theres unlocked firmware!

regardless, G.INP is capable on downstream for the chipsets inside the ECI modem and BT HH5a...

In reply to a post by mlmclaren:

No such thing as hacked firmware.... theres unlocked firmware!

I don't understand the distinction you're alluding to between hacked and unlocked firmware. Could you explain?

regardless, G.INP is capable on downstream for the chipsets inside the ECI modem and BT HH5a...

Are you saying the hardware in the modems would support it, but it's not turned on?

In reply to a post by BatBoy:

In reply to a post by mlmclaren:

No such thing as hacked firmware.... theres unlocked firmware!

I don't understand the distinction you're alluding to between hacked and unlocked firmware. Could you explain?

Hacked would mean something has been changed on the firmware and unlocked means you have gained access to factory firmwares features.

regardless, G.INP is capable on downstream for the chipsets inside the ECI modem and BT HH5a...

Are you saying the hardware in the modems would support it, but it's not turned on?

No your saying that, hacking something that isn't physically capable of something wouldn't make it all of a sudden capable.

With my conversation by someone very high up in BT, the ECI modem does G.INP on the downstream only - and all modems in circuit would have been flashed by BTOR already.

In reply to a post by lockyatlrg:

With my conversation by someone very high up in BT, the ECI modem does G.INP on the downstream only - and all modems in circuit would have been flashed by BTOR already.

Unfortunately, there's no way of proving that as no stats are available.

So your saying that G.INP probably does exist but theres no official statement to say it is.

So why have you been arguing that its not capable all morning?

In reply to a post by mlmclaren:

So your saying that G.INP probably does exist but theres no official statement to say it is.

No

So why have you been arguing that its not capable all morning?

Because it isn't.

Your proof is?

I've seen it active... and someone else seems to agree that its active.

In reply to a post by mlmclaren:

I've seen it active..

You've seen it active on a HH5A

..and someone else seems to agree that its active.

And someone else says the ECI modems "would have", not "have"

In reply to a post by lockyatlrg:

all modems in circuit would have been flashed by BTOR already.

The ECI modem is not the HH5A.

Edited by deleted (Sat 31-Oct-15 14:20:34)

In reply to a post by BatBoy:

..and someone else seems to agree that its active.

And someone else says the ECI modems "would have", not "have"

Actually...

In reply to a post by lockyatlrg:

The ECI modem does support G.INP

In reply to a post by BatBoy:

The ECI modem is not the HH5A.

The HH5a and ECI both have the same chipset.... both have been updated (if connected at right times)

In reply to a post by lockyatlrg:

all modems in circuit would have been flashed by BTOR already.

... to work with G.INP on DOWNSTREAM ONLY... however it did take BT a while longer to release the update for HH5a...

Specially ones sent out to customers after the the update was first seeded as they where still using previous versions of firmware (like I believe your's is)

I've seen it on my Huawei HG612 and my HH5B but not on my HH5A or my ECI.

Just passing on my experience without having to rely on my neighbour or someone high-up at Openreach

fair enough, I had guessed as much but whats saying the firmware on both HH%a and ECI are up to date.... AFAIK the ECI modems not connected during the rollout will not have been updated correctly and the HH5a as I mentioned also requires firmware to be pushed to it and doesn't seem to do automatic updates.

So when my neighbour had G.INP first enabled on is HH5a he had to wait for the firmware update, then that loaded on and he gained back his normal speed plus another 8-9Mb/s... however this hub later stopped working and a replacement was sent which has old firmware so he saw the loss in speed and rise in latency again for around 5 weeks and then the hub was updated and back to G.INP speeds....

By the sounds of things you have mainly been using a HG612 meaning your VRX chip based equipment probably hasn't been connected to a line long enough to get the new firmwares.

Also the TP-Link W9980 runs the same chipset as the ECI Modem and HH5a and they fixed the G.INP (Downstream) issue with an update, so theres a pattern here.

Anyway, I'm not discussing it anymore as there is no hard evidence to prove either parties theory.

Am I right in thinking most, if not all, ECI cabinets still don't have it?

To my knowledge yes you are correct, haven't looked recently though!

In reply to a post by mlmclaren:

when my neighbour had G.INP first enabled on is HH5a he had to wait for the firmware update, then that loaded on and he gained back his normal speed plus another 8-9Mb/s... however this hub later stopped working and a replacement was sent which has old firmware so he saw the loss in speed and rise in latency again for around 5 weeks and then the hub was updated and back to G.INP speeds....

Interesting, but only speculation and anecdotal. The VRX268-based HH5a - with stock firmware - does not provide detailed line characteristics; at least not via its standard web interface. Consequently, with the HH5a, it's impossible to say whether G.INP (ReTX G.998.4) is active on the line, or not.

Notably, those who've replaced the Openreach firmware on their VRX268-based ECI B-FOCuS /r modems - replacing it with OpenWRT plus an updated DSP BLOB said to support G.INP - have yet to confirm that G.INP is now active on their lines. Of course only those on Huawei DSLAMs would be able to test that out any way.

Maybe that sudden boost in speed on your neighbour's line - which he attributed to G.INP being activated on his HH5a - was just the spontaneous remission of a line fault? The fact that the sync speed appeared to rise and fall so dramatically, could simply indicate an intermittent line fault. Nothing to do with G.INP.

---

Edited by deleted (Sat 31-Oct-15 15:59:33)

In reply to a post by edwincluck:

Maybe that sudden boost in speed on your neighbour's line - which he attributed to G.INP being activated on his HH5a - was just the spontaneous remission of a line fault? The fact that the sync speed appeared to rise and fall so dramatically, could simply indicate an intermittent line fault. Nothing to do with G.INP.
---

I did check this by confirming line stats, but they where unchanged..

As I expect you know, 9Mbps or so difference in sync is typical of standard FTTC interleaving being turned on (lowers sync) or off (raises it again). The "Max/Attainable" tends to move in the opposite direction.

G.INP does reduce interleaving to a nominal figure, either 8/1 or 16/1 from what I see, and the sync gain above occurs.

Whether or not anyone can deduce the activation or not of G.INP on a line using the HH5A I have no idea, but think it unlikely. Standard interleaving being added or removed seems more likely to me.

The ECI Modem is an incompatible POS , it does not support G.inp, I used to use the ECI till just before G.inp ( version 1 up/down )was rolled out a short time after they rolled out version 2 G.inp ( down only) i recently during August tried the ECI modem, no changes in latency were observed for around 48hrs, then DLM clobbered the connection,leaving the modem in a walled garden like state,after re installing the i found that high amounts of interleave added and a big loss of sync speed, hence the POS was swapped back to the Huawei HG612 3b modem, i purchased from flea bay and flashed with hacked/unlocked fw so i can monitor stats, within 3 days DLM relented and totally reversed it's previous actions so back to G.inp and fast path on the upstream, and normal sync rates

And yet BT claim to of tested their modems for compatibility i doubt they made it out of the box

Edited by tommy45 (Sat 31-Oct-15 18:08:34)

This may have nothing whatsoever to do with the price of bananas but my unlocked HG612 which has been using G.INP for some months dropped and resynced without G.INP about a week ago after heavy rain. Since then I've switched it off / on every few days, waiting anything from a few seconds to a few hours between off & on but still no G.INP.

In reply to a post by JHo1:

This may have nothing whatsoever to do with the price of bananas but my unlocked HG612 which has been using G.INP for some months dropped and resynced without G.INP about a week ago after heavy rain. Since then I've switched it off / on every few days, waiting anything from a few seconds to a few hours between off & on but still no G.INP.

If there hasn't been any adverse effects, and your not seeing a high error rate then maybe DLM decided your line doesnt need it, though there is some speculation that G.inp had replaced fast path as the default profile

Is it on downstream interleaving or Fast Path?

Errm, downstream. Shot of stats page below.



Stats recorded 31 Oct 2015 19:18:54

DSLAM/MSAN type: BDCM:0xa459 / v0xa459
Modem/router firmware: AnnexA version - A2pv6C038m.d24j
DSL mode: VDSL2 Profile 17a
Status: Showtime
Uptime: 2 days 10 hours 20 min 41 sec
Resyncs: 0 (since 31 Oct 2015 19:17:51)

Downstream Upstream
Line attenuation (dB): 32.7 0.0
Signal attenuation (dB): Not monitored
Connection speed (kbps): 15413 1229
SNR margin (dB): 5.3 6.6
Power (dBm): 9.9 10.5
Interleave depth: 1 1
INP: 0 0
G.INP: Not enabled

RSCorr/RS (%): N/A 0.0000
RSUnCorr/RS (%): N/A 0.0000
ES/hour: 43.4 0


What G.INP did seem to achieve was to virtually nail the sn margin's feet to the floor. It barely moved.

John

That's Fast Path, so maybe DLM has decided you don't need G.INP anyway. Bear in mind that G.INP is an error-correction method. (Retransmitting duff packets instead of correction by heavy interleaving).

In reply to a post by tommy45:

The ECI Modem is an incompatible POS , it does not support G.inp, I used to use the ECI till just before G.inp ( version 1 up/down )was rolled out a short time after they rolled out version 2 G.inp ( down only) i recently during August tried the ECI modem, no changes in latency were observed for around 48hrs, then DLM clobbered the connection,leaving the modem in a walled garden like state,after re installing the i found that high amounts of interleave added and a big loss of sync speed, hence the POS was swapped back to the Huawei HG612 3b modem, i purchased from flea bay and flashed with hacked/unlocked fw so i can monitor stats, within 3 days DLM relented and totally reversed it's previous actions so back to G.inp and fast path on the upstream, and normal sync rates

And yet BT claim to of tested their modems for compatibility i doubt they made it out of the box

From what you have said above, the modem probably hasn't even been updated with latest firmware as it has not been connected during the stages of G.INP rollouts... it was said at the beginning in March that if the modems where not connected during the pushing of new firmware or the rollout of G.INP that it would not be capable.

You also mentioned that the modem worked fine for a few days and then DLM changed something that increased the latency, so IMO that doesn't mean the ECI isn't capable, it means the DSLAM changed the settings incorrectly..

However, you haven't told us whether or not if G.INP or Fast Path was enabled before connecting the ECI Modem then 2nd time, if it was Fast Path and was then switched to a standard interleaving profile that would probably mean that DLM detected an issue...

if G.INP was activated whilst the ECI Modem was connected but the firmware wasn't able to understand this just means that the modem from Openreach point of view is capable with G.INP, but the modem wasn't up to date.

The only people that can tell us if its capable or not is the people that stuck with the ECI's during the issues and didn't just disconnect and connect different equipment...

In reply to a post by mlmclaren:

In reply to a post by tommy45:

The ECI Modem is an incompatible POS , it does not support G.inp, I used to use the ECI till just before G.inp ( version 1 up/down )was rolled out a short time after they rolled out version 2 G.inp ( down only) i recently during August tried the ECI modem, no changes in latency were observed for around 48hrs, then DLM clobbered the connection,leaving the modem in a walled garden like state,after re installing the i found that high amounts of interleave added and a big loss of sync speed, hence the POS was swapped back to the Huawei HG612 3b modem, i purchased from flea bay and flashed with hacked/unlocked fw so i can monitor stats, within 3 days DLM relented and totally reversed it's previous actions so back to G.inp and fast path on the upstream, and normal sync rates

And yet BT claim to of tested their modems for compatibility i doubt they made it out of the box

From what you have said above, the modem probably hasn't even been updated with latest firmware as it has not been connected during the stages of G.INP rollouts... it was said at the beginning in March that if the modems where not connected during the pushing of new firmware or the rollout of G.INP that it would not be capable.

You also mentioned that the modem worked fine for a few days and then DLM changed something that increased the latency, so IMO that doesn't mean the ECI isn't capable, it means the DSLAM changed the settings incorrectly..

However, you haven't told us whether or not if G.INP or Fast Path was enabled before connecting the ECI Modem then 2nd time, if it was Fast Path and was then switched to a standard interleaving profile that would probably mean that DLM detected an issue...

if G.INP was activated whilst the ECI Modem was connected but the firmware wasn't able to understand this just means that the modem from Openreach point of view is capable with G.INP, but the modem wasn't up to date.

The only people that can tell us if its capable or not is the people that stuck with the ECI's during the issues and didn't just disconnect and connect different equipment...

it was connected prior to the 1st roll-out, And for some 2 weeks after it I know because i ended up with interleave on both up & the down stream & not issue 1, were the modem had not received the firmware update , (modem unable to obtain sync) needing a DLM reset so it could download the G.inp update
I suggest you read this G.INP - BT rollout 2015 has some in depth info

The issues that i experienced where different in that the modem was in sync, but no internet access, and after swapping to a compatible modem, the HG612 i was able to see that my ds had lost some 12mbps ,g.inp was no longer active and loads of lag,(interleave ) had been applied ,

Edited by tommy45 (Sun 01-Nov-15 14:43:43)