FritzBox 7530 with PN Full Fibre - Frequent PPPoE timeouts

For the past couple of months, I have been experiencing frequent PPPoE timeout disconnections on my PlusNet Full Fibre 300 service.. The service is otherwise very good quality with no dropped packets and low latency according to TBB tests.

These happen at fairly random times of day, sometimes two or three in a row within an hour, sometimes just one a week.

I have a 4G USB stick on the FritzBox which kicks in successfully each time, so I don't completely lose internet connection while the FritzBox renegotiates.

I reported the issue to PlusNet last month and they were their usual unhelpful selves, first insisting I use the PlusNet router, then wanting me to factory reset the FritzBox both of which would involve redoing a load of reserved addresses and port forwarding rules which I wasn't prepared to do.

At their request, I did a factory reset on the ONT which seemed to improve matters for about a week then the problem returned.

FritzBox was on version 8.21, since updated to 8.25. The update made no difference.

The Fritz log typically shows: (today's instance)

09.06.26 07:40:23 Failover protection: The primary internet connection is available again.
09.06.26 07:37:44 Could not establish IPv6 internet connection: No response from DHCPv6 server (SOL).
09.06.26 07:37:23 Internet connection established successfully. IP address: 80.189.nnn.nn, DNS server: 212.159.13.49 and 212.159.6.9, gateway: 172.16.14.26, broadband PoP: acc-aln1.tlh
09.06.26 07:37:18 Internet connection established successfully. IP address: 192.168.8.107, DNS server: 192.168.8.1 and 192.168.8.1, gateway: 192.168.8.1
09.06.26 07:37:18 Failover protection: The primary internet connection is currently not available. The backup connection will be enabled.
09.06.26 07:36:48 Timeout during PPP negotiation.
09.06.26 07:36:48 Internet connection cleared.

Sometimes, I get the same PlusNet IP address, sometimes a different one.

The ONT in the living room (front wall of house) is connected to the router via a professionally installed Ethernet Cat 6e wall socket to a patch panel in the loft and back down to a wall socket in the study where the router sits(back of house). Total cable run from ONT to router is about 18 metres.

All patch leads are Cat 6e. I have changed both ONT to wall socket and wall socket to router leads. No difference.

Any suggestions about how to pinpoint the cause of the timeouts and where to go next with PlusNet?

As a starting point can you save your current configuration through the FritzBox interface? This page suggests you can. If so you can save the config, do the factory reset, let Plusnet check things out then reload your config.

PlusNet closed the problem after recommending the factory resets, saying if the problem persisted after resetting I should buy a new router. i.e. a complete denial that it is anything to do with them. So there doesn't seem to be any point in resetting unless there is a clear indication of what is going wrong and why.

What I was after here is some idea of what causes the PPPoE timeout and / or the "Connection Cleared" status that initiates the attempt to reconnect.

Since no-one has leapt in with experience of a similar problem, I'm guessing it isn't a common thing, so I'm on my own.

Oh well...

I've had this a few times on a Zen/Cityfibre connection. Mostly during the night (29 disconnects ) but don't know if it was some sore of maintenance but I did have it during the day once.

I assume you don't have any range extenders like the 3000AX?

Thanks, no, no range extenders as such.

I use hard-wired wireless access points around the house wired back to the router via various switches. Each operates independently of the others. I am pre-mesh. (Mostly pre-WiFi - I can see a cable and know what it does).

For info, I have now bitten the bullet and done a factory reset on the FritzBox. Slightly painful, but now seems to be up and running again. I await the next PPPoE timeout.

Should also have mentioned the FritzBox is also acting as a DECT base for two FritzBox C6s, and using A&A for VOIP, which make it difficult to switch to the PlusNet router for testing..

However, I am thinking of pre-configuring the PlusNet Router with the FritzBox address ranges and port forwarding so I can plug it in as a spare if necessary and switch the VOIP to route to my mobile when the FritzBox isn't available. But that's another story.

If you won't swap in the router Plusnet have provided to use with the service in order to rule out the Fritzbox as the issue then I'm not sure what you're expecting them to do. Maybe Fritzbox can expose a more verbose level of logging that will shed some light on how it thinks it's losing the PPP session.

Edited by jpm (Wed 10-Jun-26 12:35:41)

In reply to a post by Axolotl:

PlusNet closed the problem after recommending the factory resets, saying if the problem persisted after resetting I should buy a new router. i.e. a complete denial that it is anything to do with them.

In a sense they are right. The reason they supply a router is to avoid responsibility for the oddballs like you [and me] who want to use our own routers. For the brief period that I had to use such an ISP, I accepted that if support was required, I would have to use the supplied router, which was a right pain, because I could not even set my own private address range for my own network.

If you are keen to use your own router, then go to an ISP which does not supply a router and pay a little more.

In reply to a post by Axolotl:

For info, I have now bitten the bullet and done a factory reset on the FritzBox. Slightly painful, but now seems to be up and running again. I await the next PPPoE timeout.

do keep us updated on this reset

In reply to a post by Axolotl:

However, I am thinking of pre-configuring the PlusNet Router with the FritzBox address ranges and port forwarding so I can plug it in as a spare if necessary and switch the VOIP to route to my mobile when the FritzBox isn't available. But that's another story.

Again do this and see if the PPPoE disconnect go away. You may need a new router.

also if the fritzbox allows- go back a firmware edition.

There is a hard truth that, sometimes routers get old and can't do the job anymore. There is also a sitation where you just ask too much of it and it will burp and disconnect and reconnect. I had this happen on a low end tp link archer router. Had a new phone with wifi calling and the archer played up and i couldn't keep a connection

When I used to use a fritzbox on my former Plusnet FTTC connection I had a similar problem. I believe I made sure the "permant connection" button was ticked and changed to PPPoA and it seemed to sort it. Dont know if that will help your case though.

After I reset the FritzBox, nothing happened for about a week, then the disconnects started again. This time in the small hours .

Latest at 04.30 ish lasted about six minutes before connection was re-established. That was followed by another brief disconnect at 09.40. In each case, the same PlusNet iP was assigned. That was Yesterday morning. Today, no disconnect so far.

When I have time, I will try the other suggestions that have been made. One at a time, so if any work, it will be obvious what the cause was.

Since it takes up to a week before the problem occurs, there may be a long gap between postings.

My hand has been forced.

Today, the connection dropped at about 04.55 with the usual PPPoE timeout error, and has stayed out ever since.

Nothing I've tried with settings and restarting the FritzBox has made it connect again.

I have switched to the PlusNet 2 router and am slowly naming and configuring my network devices on it.

Obviously I have lost the VOIP service for now since the PlusNet router doesn't have a DECT or phone capability.

In reply to a post by Axolotl:

Obviously I have lost the VOIP service for now since the PlusNet router doesn't have a DECT or phone capability.

Not wanting to detract from your bigger problem, but if your 'VoIP' service depends on the ISP supplied router, it is not a true VoIP service

I think you've got things the wrong way round.
The Fritzbox router provided the VOIP service with it's built in ATA and DECT base station. Plusnet do not have any provision for a VOIP service so the ISP supplied router (Hub 2 ) will do nothing.

VOIP use requires either a router equipped with an onboard ATA or a separate ATA via Ethernet. In this case it's only "dependent on the provided router" because that router has an onboard ATA; not because it's ISP provided. I have a self-bought Fritz!Box and use it in exactly this way. It was not provided by, nor is it related to, my ISP. Yet you might say I'm dependent on it because I don't possess any external ATA. But I'd only have to buy one to solve it if I needed to use a different router. Not that I'd be likely to; the Fritz!Box is probably the best router I have ever had - never mind it's telephony hardware - and I'd replace it if needed with another one from the same stable, phone or not.

I think the other replies have said it for me.

Update:

The exact same problem has occurred with the PlusNet supplied Hub 2. I commissioned it about 11.30 yesterday. 23.41 the line drooped for 13 minutes before reconnecting, then again for 30s at 06.58.

I have raised another ticket with PlusNet. They are saying Openreach tests show no fault and have got me to move the Hub 2 adjacent to the ONT to rule out my internal ethernet cabling.

This I have done as of 10.38 this morning (19/6/26.) any bets on how long it will stay connected?

In reply to a post by Axolotl:

Update:

I have raised another ticket with PlusNet. They are saying Openreach tests show no fault and have got me to move the Hub 2 adjacent to the ONT to rule out my internal ethernet cabling.

right, besides the silly comment by pn, have you tried a new "short" cable between csp-router.And preferably one that has been tested

If by csp you mean the ONT, that is the set up now that I have moved the Hub 2 to adjacent to the ONT.

Not aware that I have any access to the csp which I thought was the external box that connects the drop fibre from the pole to the internal ONT?

The TP-Link switch software diagnostics for the switches that the cabling system connects to shows all cables as their correct lengths and "passed" when the router was in another room away from the ONT but I don't know how significant that is.

Now only the short cable between router and ONT is involved, I have no way of testing that.

It is a different cable from the one that connected the ONT to the wall socket for the cabling system. That was a 0.25 metre patch lead. Now it is a 1.5 metre lead previously used for PC to router connection.

Do the lights on the ONT do anything when the connection drops?

That's exactly correct about the VoIP provision.

When you say it's patched via the loft, is that literally an ethernet cable from one patch socket to another? Or are you doing it via a switch? You don't want any equipment other than a patch panel between the modem and ONT.

Can't tell. ONT is hidden behind curtains. Also I am not usually looking at it when the connection droops. I only become aware of the dropped connection later when I see the logs or the TBB quality monitor goes red.

It is patched direct ethernet to ethernet.

From ONT to wall socket. From wall socket wired system to patch panel socket A from Patch panel socket A to socket B short patch lead, socket B to wall socket in other room wired system, then short patch lead to router from wall socket.

No switches involved in that link.

However, updated situation is that I've raised another ticket with PlusNet because the problem came back even with switching to the supplied PlusNet Hub 2 router.

PlusNet have asked me to move the router to adjacent to the ONT so the wiring system is now out of the equation. Now have just ONT to router port 4 short ethernet cable.

Worked for about two days then the problem has come back. Three disconnects today (Sunday) 04.40, 06.58 and 21.45 from memory each lasting a few seconds.

Back to PlusNet support on Monday

In reply to a post by Axolotl:

It is patched direct ethernet to ethernet.

From ONT to wall socket. From wall socket wired system to patch panel socket A from Patch panel socket A to socket B short patch lead, socket B to wall socket in other room wired system, then short patch lead to router from wall socket.

No switches involved in that link.

However, updated situation is that I've raised another ticket with PlusNet because the problem came back even with switching to the supplied PlusNet Hub 2 router.

PlusNet have asked me to move the router to adjacent to the ONT so the wiring system is now out of the equation. Now have just ONT to router port 4 short ethernet cable.

Worked for about two days then the problem has come back. Three disconnects today (Sunday) 04.40, 06.58 and 21.45 from memory each lasting a few seconds.

Back to PlusNet support on Monday

Have you tried a reboot of the ont ?(clutching at straws)

ONT was factory reset as part of the original problem report to PlusNet. The problem seemed to go away for about a week but then returned.

I have not reset the ONT during the latest bout of drops because PlusNet support didn't ask me to this time around.

Slight progress.

Today, (Monday), I was actually there when one of the many disconnects occurred and the PON light on the ONT (which I have learnt is a Zyxel ONT, was red while the outage was happening..

As I understand it that puts the ball firmly in Openreach's court.

I have reported this to PlusNet (including multiple disconnects over the weekend) and finally Openreach have agreed to send an engineer to investigate. Scheduled for tomorrow pm.

Watch this space.

Good luck.

Explain that you have witnessed the PON light go red at times when the fault occurs (a kind of white lie)
Be nosy, ask what tests they running and ask them to explain/interpret the test results they get.
A light loss reading at the ONT should be pretty much the first thing they do. The reading ought to be between -14db and -21db. It can be higher, but you are then in realms of knowing what the readings are at the splitter first off.

It is more than likely you’ll get a contractor … don’t let them fob you off …. if they really cannot locate the fault (its resolution being shown by an improved light loss reading after the fix) then ask that they ‘further’* the fault back for Openreach network fibre people to look into.

*further is to pass the the fault to a different people or team when you are unable to rectify yourself.

Thanks.

Now I am awake, I realise that today is Tuesday, and the engineer is confirmed for tomorrow (Wednesday) afternoon.

I'll update with the outcome.

Update on today's Openreach visit.

An actual Openreach engineer came at 14.00, tested the light levels at the ONT, concluded they were low, but within spec. He then proceeded to check at the CBT and the CSP and found a bad splice in the CSP.

He showed me a photo of the splice that was glowing red. Apparently that is a bad thing...

He repaired that, got a much improved light reading at the ONT then restarted the ONT and couldn't get a PON light at all, so decided to replace the ONT.

At this point, farce took over. He couldn't access Openreach systems to register the new ONT, so I now had a new connection that couldn't be used because Openreach's system didn't recognise the ONT.

He put in a request to register the new ONT via some central system, but that was queued because Openreach was experiencing general problems with all their systems today...

He thought the change should go through some time overnight tonight but couldn't say for sure and agreed to keep the problem open until he could retest and confirm tomorrow.

We agreed I could revert to the FritzBox 7530 router so I could use the 4G failover while the ONT was out.

I set that up about 16.30, and the ONT burst into life while the engineer was still on site.

I lost connection again a couple of times while he tested the connection, which passed.

I've been connected since with no drops. nearly 5 hours.

I'm not holding my breath.

At least the Openreach man has witnessed the thing not working.

He showed me a photo of the splice that was glowing red. Apparently that is a bad thing...

It is. There will always be some escaping of light at a splice, (in this case from the light source at the ONT) but it shouldn’t be glowing so very brightly. This is evidence of the two cleaved ends of the fibres are not ‘butting up’ plushly together … this the light escapes.

I have my fingers crossed that this is the fixing of it. 🤞

The hiccup with the new ONT registration not going through is just one of those things.

Please update this thread in a couple of days either way.

In reply to a post by Axolotl:

I lost connection again a couple of times while he tested the connection, which passed.

I've been connected since with no drops. nearly 5 hours.

I'm not holding my breath.

At least the Openreach man has witnessed the thing not working.

Hopefully it will now last. With the OR guy seeing the los siginal himself, allows him(or) to esclate further if needed. Hopefully there is no more bum splices in the spine.

Edited by Taras (Thu 25-Jun-26 09:03:53)

Well, 48 hours now without a hitch.

Looks promising.

TBB BQM graph attached.

My Broadband Ping

In reply to a post by Axolotl:

At this point, farce took over. He couldn't access Openreach systems to register the new ONT, so I now had a new connection that couldn't be used because Openreach's system didn't recognise the ONT.

Have just seen this - catching up after a 27 hour outage - a whole area was down. Then overnight Tuesday/Wednesday numerous systems went down and faults could not be registered, lines tested &c and they started to come back on line during Wednesday afternoon and the registration of the ONT was probably connected with that.


At least your 4G backup worked - mine would not connect and stay up.