CAX80 Randomly Loosing Downstream Lock - MDD message timeout?

[CyborgMonkey]

Hi everyone, I recently upgraded my internet through RCN to gigabit (940/25mbps) and for that decided to get the RAX80 router which seems like quite a beast. Unfortunately though im running into an issue where the modem is randomly loosing it's downlink lock and the internet remains cuts out until I power cycle the device. The behavior at the device is the downlink light starts flashing while the uplink and online remain solid. Prior to using the RAX80 I was using a SB6183 with zero issues from the same cable outlet in the house. I've investigated the logs and I'm not seeing any obvious signs of issue besides a "MDD message timeout" which is always the last message before the downlink is lost. This happens just about every day or every other day and strangly occurs mostly in the evening or at night (when traffic is at it's lowest here).

I do not use RCN for TV, so the cable coming into the house is routed directly to the CAX80. It does pass through x2 male-to-male adaptors before reaching the router (x1 to extend the cable in the basement a short distance (<10ft) and x1 at the wall jack). Thinking it possibly was bad cable and/or connectors, I went ahead and bought brand new cable and new male-to-male connectors, plus rerouted the cable a much shorter distance in the basement. The cable coming into the house from outside was only put in a few years ago, so I wouldn't expect any issues there. Additionally I checked the outside box where the cable comes out of the ground and goes into a MoCA filter before going into the house, but I see no obvious signs of issue either. While I haven't yet experienced any issues since installing the new cable and connectors, I'm also not seeing much at the CAX80 to suspect there would be any improvment. All SNR values are basically the same with only a ~1dBmV power gain downstream. Upstream is identical.

HW version: 1.01

Router Firmware: V1.1.0.41

Cable Firmware: V1.02.08

Downstream:

Channel	Lock Status	Modulation	Channel ID	Frequency	Power	SNR	Correctables	Uncorrectables
1	Locked	256 QAM	16	621000000 Hz	1.6 dBmV	39.7 dB	0	0
2	Locked	256 QAM	1	531000000 Hz	1.8 dBmV	40.3 dB	0	0
3	Locked	256 QAM	2	537000000 Hz	1.9 dBmV	40.0 dB	0	0
4	Locked	256 QAM	3	543000000 Hz	1.8 dBmV	40.1 dB	0	0
5	Locked	256 QAM	4	549000000 Hz	1.8 dBmV	40.0 dB	0	0
6	Locked	256 QAM	5	555000000 Hz	1.9 dBmV	39.8 dB	0	0
7	Locked	256 QAM	6	561000000 Hz	1.5 dBmV	39.8 dB	0	0
8	Locked	256 QAM	7	567000000 Hz	1.7 dBmV	39.7 dB	0	0
9	Locked	256 QAM	8	573000000 Hz	1.5 dBmV	39.7 dB	0	0
10	Locked	256 QAM	9	579000000 Hz	0.8 dBmV	39.4 dB	0	0
11	Locked	256 QAM	10	585000000 Hz	1.2 dBmV	39.5 dB	0	0
12	Locked	256 QAM	11	591000000 Hz	1.3 dBmV	39.6 dB	0	0
13	Locked	256 QAM	12	597000000 Hz	1.1 dBmV	39.4 dB	0	0
14	Locked	256 QAM	13	603000000 Hz	1.4 dBmV	39.6 dB	0	0
15	Locked	256 QAM	14	609000000 Hz	1.7 dBmV	39.5 dB	0	0
16	Locked	256 QAM	15	615000000 Hz	1.5 dBmV	39.8 dB	0	0
17	Not Locked	Unknown	0	0 Hz	0.0 dBmV	0.0 dB	0	0
18	Not Locked	Unknown	0	0 Hz	0.0 dBmV	0.0 dB	0	0
19	Not Locked	Unknown	0	0 Hz	0.0 dBmV	0.0 dB	0	0
20	Not Locked	Unknown	0	0 Hz	0.0 dBmV	0.0 dB	0	0

Upstream:

Channel	Lock Status	US Channel Type	Channel ID	Symbol Rate	Frequency	Power
1	Locked	ATDMA	3	5120 Ksym/sec	23900000 Hz	30.3 dBmV
2	Locked	ATDMA	1	5120 Ksym/sec	36800000 Hz	31.3 dBmV
3	Locked	ATDMA	2	5120 Ksym/sec	30350000 Hz	30.3 dBmV
4	Not Locked	Unknown	0	0	0	0.0 dBmV
5	Not Locked	Unknown	0	0	0	0.0 dBmV

Downstream OFDM:

Channel	Lock Status	Modulation / Profile ID	Channel ID	Frequency	Power	SNR / MER	Active Subcarrier Number Range	Unerrored Codewords	Correctable Codewords	Uncorrectable Codewords
1	Locked	0 ,1 ,2 ,3	159	634000000 Hz	2.58 dBmV	40.3 dB	646 ~ 3449	294579275	109804738	0
2	Not Locked	0	0	0 Hz	0 dBmV	0 dB	0 ~ 4095	0	0	0

There are no upstream OFDMA channels.

The last time the downstream lock was lost, there was one downstream OFDM uncorrectable codeword. I have never seen any correctables or uncorrectables when the downstream lock is lost from any of the downstream bonded channels.

Does anyone have any ideas why my downstream link would keep loosing it's lock or what the meaning of the "MDD message timeout" is? Most of the research I've done seems to point to cable issues, but my setup is so simple at least inside the house, that I have a hard time believing that's the point of failure. I would like to make sure I have all my bases covered before calling a tech out.

Thanks!!!

[CyborgMonkey]

Issue happened again this morning . Left the error state going when I called RCN and the tech told me she could see the device operating in a "partial mode". I power cycled the device and then had the tech reprovision it. She confirmed that there were no signal strength issues from the diagnostic tooling on her end which also coincides with what I'm seeing. Since the reprovisioning I'm actually seeing more consistent and faster downstream speeds then before. Not sure what exactly happened before, but fingers crossed maybe this will work.

An interesting note from when I set the device up initally with RCN: After provisioing the device the tech had me pull the power from the device each time, despite the device restarting on it's own. The tech today did not have me do this, but offered some insight as to why the other tech might have had me do this... Apparently with the DOCSIS 3.1 standard, the cable modem could take an upwards of an hour to come back online after provisioning, and they had found a trick to this was to power cycle the device to get it to come back online faster. This makes no sense to me because if the device resets on it's own after provisioning, why bother pulling the power again? I have a feeling this might have corruped a provisioning perameter as i had to do this a couple times due to the first tech entering my MAC address into the system incorrectly.

If this doesn't work I have a feeling it's an issue with the CAX80, which really sucks for a $450 piece of equipment. Guess I'll have to call Netgear if it happens again. I see no point in paying for a tech to come out and tell me the CAX80 is the problem.

[FURRYe38]

Have the ISP check the signal and line quality and provisioning UP to the modem. 
Be sure there are no coax cable line splitters in the between the modem and ISP service box. 
Be sure your using good quality RG6 coax cable up to the modem. 

Might contact NG support to ask for addition help and information regarding this.

@vkdelta 

[CyborgMonkey]

Having a RCN tech come out on Thursday to check out the issue because I'm completely baffled at this point. I feel like the intervals are just getting smaller now too. Last on was almost exactly 12 hours.

There are no splitters between the line into the house and the device. There are only the two female-to-female coax connectors (753GFFF) to extend the cable <10ft to the wallplate on the first floor (done with brand new RG6). Alltogether there's roughly 30-35ft of cable between the cable coming out of the ground and my CAX80. ~20ft of that is RCNs cable they installed about 4-5 years ago when the service was first put into the neighborhood.

I also called the Netgear technical support which tried to tell me it was a Wi-Fi channel issue (it is 100% NOT a Wi-Fi issue). I have a technical engineering background and I know this is not the case. I stressed this point considerably and yet I'm not sure if they're scripted or what but they absolutely insisted on it. Increadibly fustrating.

So yeah, this is where I'm at. Guess I'll be doing a lot of modem reboots until Thursday...

[FURRYe38]

Well to verify its not a wifi issue, you can turn OFF the wifi radios and then check again. You can install a external wifi router configured for AP mode as a 2ndary wifi as well. I doubt the wifi is an issue as well. 

Make contact with the forum moderator as well. 

[DasWax]

I am having the same issue with my CAX80 since September. Every firmware upgrade I have done this still happens, Both Spectrum and Netgear blame each other. I got annoyed of this issue happening all the time and I am now back on my stock spectrum modem and my ASUS RT-AC68R....No issues again 🙂

I blame Netgear in the end....

[jesmith602]

@CyborgMonkey your upstream power levels look low. I would check them with RCN.

Here are all my CAX80 troubleshooting tips:

https://community.netgear.com/t5/Cable-Modems-Routers/CAX80-Cox-ISP-Challenges-recurring-disconnects...

[DasWax]

---

@jesmith602 wrote:

@CyborgMonkey your upstream power levels look low. I would check them with RCN.

Here are all my CAX80 troubleshooting tips:

https://community.netgear.com/t5/Cable-Modems-Routers/CAX80-Cox-ISP-Challenges-recurring-disconnects...

---

What's the point, as OP said its a DOWNSTREAM issue, not a UPSTREAM issue. Its already a known issue with Netgear that this firmware for the modem side is at fault and still no fix.

[FURRYe38]

Any progress on this? 

---

@CyborgMonkey wrote:

Having a RCN tech come out on Thursday to check out the issue because I'm completely baffled at this point. I feel like the intervals are just getting smaller now too. Last on was almost exactly 12 hours.

There are no splitters between the line into the house and the device. There are only the two female-to-female coax connectors (753GFFF) to extend the cable <10ft to the wallplate on the first floor (done with brand new RG6). Alltogether there's roughly 30-35ft of cable between the cable coming out of the ground and my CAX80. ~20ft of that is RCNs cable they installed about 4-5 years ago when the service was first put into the neighborhood.

I also called the Netgear technical support which tried to tell me it was a Wi-Fi channel issue (it is 100% NOT a Wi-Fi issue). I have a technical engineering background and I know this is not the case. I stressed this point considerably and yet I'm not sure if they're scripted or what but they absolutely insisted on it. Increadibly fustrating.

So yeah, this is where I'm at. Guess I'll be doing a lot of modem reboots until Thursday...

---

[CyborgMonkey]

Hey guys, I appreciate the feedback! So to pick up where I left off earlier, not soon after my last post I had an RCN tech come out and he basically confirmed everything looked great signal-wise. The one small modification he made was to recrimp some outside connectors between the splitter box for my residence and the neighbors in the front yard (it's like a green cylinder that sticks up from the ground between the houses), and also recrimped the connector where that same feed comes into the house. He claimed the tech that put it in did not make the copper signal wire lead long enough and explansion/contraction due to temperature could be triggering the disconnects. My thoughts, yeah sure maybe. Can't hurt i guess! He also removed a moca filter from the outside box on the house which I guess isn't needed anymore.

So after the tech left the issue occued again within a day. At this point I thought bad device, so I returned that CAX80 and got a new one. New one seemed to work slightly better, but still ran into the same exact problem. By this point I was personally contacted by Netgear about the issue (from this forum post actually) and they walked me through collecting full diagnostic logs from my CAX80. Netgear responded that they saw a lot of errors in the logs, but never really clairified to me what exactly the problem was despite my asking. At this time the downstream "unlock" / MDD messgae timeout was occuring daily or every other day. So s few days later I get a response from NG asking me to test out a new beta firmware (v2.1.0.5). The recommendation was to install this firmware, do a complete factory reset of the device and then set it up again using the app. Okay, fine. About 2 hours after installing the firmware and having everything setup, issue happens again. However! That was 10/22, since that first incident I have not experienced the issue again.

Now, as someone who works professinally in system testing/validation, I'm well aware of flukes and strange behavior of computer electronics. I'm really not yet conviced this beta firmware has actually fixed the issue. I'm still giving it until this weekend before making any more final opinions. The only other thing i'll say here is I do see MDD message timeouts in the web UI event viewer, but so far these have not caused a noticable impact on connection. Maybe there is some better recovery code now in the beta firmware for when these events occur? Not sure. I would expect for a regular household residence the signal isn't always going to be perfect and some minor dips and peaks in signal would occur every now and again, but nothing the CAX80 shouldn't be properly able to handle and recover from.

Last point i want to bring up is that the beta firmware i received was only for the router, not the modem. Now, I'm not an expert on how cablem modem/router gateways work, but I don't understand how when i'm having issues with the cablem modem portion of the device, why an update to the router firmware would correct this problem. Clearly the issue lies in the downstream communication at the modem iit would seem, not the router portion of the device. But who knows, maybe there's more interconnection to it than I'm aware of...

Anyways, I'll follow up again soon with any updates.

[FURRYe38]

Thanks for the update. Ya I have seen odd issues persist soon after a FW update, then for some reason stops. Maybe it took both the modem, router and cable services to kind of settle in after the initial problem happening again once, then maybe it's now working as expected. Ya, keep an eye on it over the weekend and see. Possible someting was being cause from the router side. Who knows what NG is going between the router and modem sections. 

[jesmith602]

@CyborgMonkey What kinda max speeds is the router getting?

[CyborgMonkey]

@jesmith602 I'm able to get the full "gigabit" speed that RCN offers, 940/20:

@FURRYe38  Yeah maybe this is the case that it needed some time to readjust. When the device entered the failure mode before it was always the MDD timeout message that was the last event I saw and it's timestamp was always just about spot on with when the downstream unlock would occur. It's good to see now that while these messages are still there (maybe they're common as it's only classified as a warning?), at least they aren't triggering the failure mode anymore. Anyways, will see how it goes!

[FURRYe38]

[90sScott]

Hey I know this is old but would you care to share the beta firmware? Have RCN in Chicago and am running into this same exact issue.

[jesmith602]

@90sScott the beta version they sent me has since been updated to 2.1.1.4.

[90sScott]

Odd firmware's been updated to that since run, any other suggestions you have?

This is literally driving me crazy

[jesmith602]

@90sScott Have you tried adding a 6db or 10db attenuator to make upstream power raise to above 40 dBmV? My problems all went away when I got the 2.1.1.4 firmware & added the attenuator. Based on your other post, has RCN been able to tell you why so many channels are unlocked? I have 8 upstream locked, and 32 downstream locked. The only other thing I'd suggest is to ask Netgear support to escalate to an engineer who can review your debug logs from an engineer's perspective, and do more detailed troubleshooting. Perhaps they have a newer beta firmware they can share?