Print

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

[Patrick533]

I said above I had 2-ea 140MM case fans on my workstation. My mistake, I should have said 200MM(About 8-9 inches across) case fans, they don't turn very fast but they move A LOT of air. The DIR-827 is now sitting on top of one and has been for about an hour. The DIR-827 is COLD to the touch. It was very warm when I started, the house temp today is about 68F. It blew a bunch of warm air out for about 15 minutes. All of my equipment sits on wire shelving to let air flow pass so that never was an issue. The PCI bus chip on the WiFi card is rated for 80C ambient, I don't have my thermal equipment any longer otherwise I would have done a temp profile on the router and I am not going to even guess what it is but too hot to touch for long would be a good description of the internal heatsink.

No, I did not try 1.05, it was not on the FTP site, I only tried .04, .02 and .01. Currently running .02.

WiFi is back on and the logs have been cleared. Now the wait........ If it fails, it was not heat.

[FurryNutz]

Wire shelving? What this wire shelving made of? Metal I presume? Could be a cause of bad signal propagation and interference. 

v1.05 is on main Canada site if you want to test it later on.

I wonder if the addition of some chip set heat sinks would be a good idea, Like we did here:
http://forums.dlink.com/index.php?topic=56166.0

Keep us posted...

[Patrick533]

You said you have your DIR-827 running? Do me a favor and ping it a couple of times and maybe run a tracert to yahoo.com or something.

I am seeing what I consider high pings for a router that is 15FT away via CAT6.

I am getting 2-3ms pings with V6 on or off.

Thanks!

[FurryNutz]

I did this for someone in the DSLreports forum just a few days ago:
I can ping that IP and avg mostly in the 80ms's an I'm well north of you:
Reply from 74.125.139.139: bytes=32 time=82ms TTL=40
Reply from 74.125.139.139: bytes=32 time=83ms TTL=40
Reply from 74.125.139.139: bytes=32 time=82ms TTL=40
Reply from 74.125.139.139: bytes=32 time=84ms TTL=40
Reply from 74.125.139.139: bytes=32 time=82ms TTL=40
Reply from 74.125.139.139: bytes=32 time=82ms TTL=40
Reply from 74.125.139.139: bytes=32 time=83ms TTL=40
Reply from 74.125.139.139: bytes=32 time=83ms TTL=40
Reply from 74.125.139.139: bytes=32 time=83ms TTL=40
Reply from 74.125.139.139: bytes=32 time=83ms TTL=40
Reply from 74.125.139.139: bytes=32 time=87ms TTL=40
Reply from 74.125.139.139: bytes=32 time=85ms TTL=40
Reply from 74.125.139.139: bytes=32 time=82ms TTL=40
Reply from 74.125.139.139: bytes=32 time=83ms TTL=40
Reply from 74.125.139.139: bytes=32 time=84ms TTL=40
Reply from 74.125.139.139: bytes=32 time=82ms TTL=40

Ping statistics for 74.125.139.139:
Packets: Sent = 50, Received = 50, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 82ms, Maximum = 331ms, Average = 88ms


I'll do a test for you with tracert using CAT6 and IPv4...

[FurryNutz]

C:\Windows\system32>tracert www.yahoo.com

Tracing route to ds-any-fp3-real.wa1.b.yahoo.com [206.190.36.105]
over a maximum of 30 hops:

  1    <1 ms    <1 ms    <1 ms  192.168.0.1
  2    10 ms     7 ms     7 ms  ##.###.##.#
  3     9 ms     7 ms     8 ms  192.168.69.89
  4   106 ms     8 ms     7 ms  192.168.102.173
  5    17 ms    17 ms    24 ms  xe-8-2-2.edge2.Seattle1.Level3.net [4.53.158.5]

  6    18 ms    19 ms    19 ms  ae-14-51.car4.Seattle1.Level3.net [4.69.147.134]

  7    18 ms    17 ms    18 ms  YAHOO-INC.car4.Seattle1.Level3.net [4.79.106.26]

  8    23 ms    29 ms    21 ms  ae-7.pat1.gqb.yahoo.com [216.115.96.45]
  9    22 ms    21 ms    21 ms  ae-1.msr2.gq1.yahoo.com [66.196.67.3]
 10    22 ms    23 ms    21 ms  xe-8-0-0.clr2-a-gdc.gq1.yahoo.com [68.180.253.13
5]
 11    21 ms    22 ms    24 ms  et-18-25.fab5-1-gdc.gq1.yahoo.com [98.137.31.182
]
 12    21 ms    22 ms    21 ms  po-14.bas1-7-prd.gq1.yahoo.com [206.190.32.23]
 13    21 ms    22 ms    25 ms  ir2.fp.vip.gq1.yahoo.com [206.190.36.105]

Trace complete.

This look ok?

[Patrick533]

This one: 1    <1 ms    <1 ms    <1 ms  192.168.0.1

Mine is 2-3 ms, it used to be < 1 ms

I am running 192.168.1.1 for an internal IP, ever run into that being a problem(.1.1 vs 0.1)?

I have tried 2 different NIC's, locking in the speed, you name it. The internal network transfer speed is still right up there, it just takes a little longer to get going.

We are seeing increased pings across the board. That is the major reason the kids want to go with a high end router, this one is sloooooowing down. I compared the one I told them I like, the speed comes in the same as the DIR-827 almost 100%, number of simultaneous connections and everything.

Miss my DIR-825, I have heard they have added a firewall to hardware C-1 and above!

This is direct to the router, no switch or anything.

1     3 ms     2 ms     1 ms  dlinkrouter [192.168.1.1]
2     9 ms     7 ms    13 ms  XXX.XXX.XXX.XXX
3    10 ms    10 ms     8 ms  dtr03rvsdca-tge-0-1-0-3.rvsd.ca.charter.com [96.34.100.76]
4    29 ms    10 ms    11 ms  crr01rvsdca-tge-0-4-0-1.rvsd.ca.charter.com [96.34.96.232]
5    16 ms    15 ms    15 ms  bbr01rvsdca-bue-2.rvsd.ca.charter.com [96.34.2.66]
6    13 ms    15 ms    11 ms  prr01lsanca-bue-5.lsan.ca.charter.com [96.34.3.7]
7    15 ms    13 ms    14 ms  eqix.lsan.twtelecom.net [206.223.123.36]
8    21 ms    14 ms    14 ms  ont1-ar3-xe-0-0-0-0.us.twtelecom.net [66.192.251.90]
9    15 ms    17 ms    15 ms  66-192-151-102.static.twtelecom.net [66.192.151.102]
10  19 ms    24 ms    18 ms  host-66-59-236-253.static.linkline.com [66.59.236.253]
11  16 ms    15 ms    15 ms  host-216-66-249-9.static.linkline.com [216.66.249.9]
12  15 ms    15 ms    18 ms  host-216-66-249-93.static.linkline.com [216.66.249.93]
Trace complete.

Weird man!

[FurryNutz]

I haven't tested with .1.1. I mainly use 0.1 here. Would be a major change here to test .1.1. 

My test was thru a switch then to router. CAT 6 cabling.

Hows your ISP modem signals? Ensure the signals are good.

Ya they added Firewall to the 825s Rev C. Rev C is the most recent version. Only seen limited users with it.

What ISP modem do you have?
You wanna test .0.1?

[Patrick533]

I just pinged everything in the house, slower then I would expect. Turned off jumbo packets just in case, no change! Switched over to a gigabit switch, no change. I am running 2 D-link gigabit switches in daisy chain so they have enough hard wire in the gaming area, might try unplugging that too.

Since I attached the windmill to the router it has not crashed! 5ghz is up and running.

The cable modem signals are very good, when we moved in I nabbed the closest room to the cable entry point for my office, street>1 splitter>modem. Actually some of the signals are a little high but within spec.

I have Charter, downgraded to 30x4, I did have 100x5 but they raised the price to like 110 a month, not worth it. Brand new Arris TM902A modem, I get 2 telephone lines from them too but I did try a Motorola SB6120 and Zoom 5341J last week, no difference. We have 8 bonded download channels so that is usually ROCK solid.

I thought the kids had left but one of them is home gaming tonight. I will switch to the latest firmware and .0.1 in the morning before they get up. I might reset my NIC in my machine to defaults since I have changed NIC cards at least 3 times, just to make sure a setting did not get messed up.

So far no crash, this is the time it would usually have died at least 5 times.

More tomorrow, I'm IP/IT baked for the day. 

[FurryNutz]

Jumbo isn't recommended on these routers. Have seen a couple of Admins not recommend using Jumbo frames.

Arris = barf 

Is splitter of good quality? Can you remove it temporarily by chance?

Glad the fan seems to be helping. Makes me wonder if something is heating up your router. I took a temp on mine tonight and just normal operation, saw at most 108F on the right side of the top surface area. Most other temp is 80-90F. I might open mine up and see if adding some chip set heat sinks could help some.

Also see if you can remove that metal shelving for something wood or plastic?

Keep us posted.

Get some sleep. Nighty night. 

[Patrick533]

Charter won't let us use our own modems anymore, the only reason they let me use mine last week was for testing.

They have made at least 10 trips to my house to get the signals where they are today.

When they 1st installed cable my power level was at -15 and signal to noise was 30. They wanted to put an amp inline to band aid fix it. I told them to go out there and balance the system like it is supposed to be. After getting the California Public Utilities Commission involved with threats of fines, this is as close as they could come(and the daily modem resetting and phones NOT working stopped, let alone the daily TV outages). The upstream is a little too close for my liking but the SNR is great. Actually if I take the splitter out I will be out of spec on the upstream.

Interface Name    Freq                    Power      SNR      Modulation
Downstream 1   649.750 MHz   4.10 dBmV  41.30 dB     QAM256
Downstream 2   613.750 MHz   2.70 dBmV     41.10 dB     QAM256
Downstream 3   619.750 MHz   2.90 dBmV     41.00 dB     QAM256
Downstream 4   625.750 MHz   3.30 dBmV     41.40 dB     QAM256
Downstream 5   631.750 MHz   3.60 dBmV     41.40 dB     QAM256
Downstream 6   637.750 MHz   3.70 dBmV     41.40 dB     QAM256
Downstream 7   643.750 MHz   4.00 dBmV     41.40 dB     QAM256
Downstream 8   655.750 MHz   4.30 dBmV     41.30 dB     QAM256

Upstream
Interface Name    Frequency      Power            Type      Symbol Rate    Modulation
Upstream 1   23.76 MHz     36.00 dBmV   ATDMA   5120 kSym/s   QAM64

Status
System Uptime:    3 days 00h:05m:13s
Computers Detected:   0
CM Status:   OPERATIONAL
Time and Date:   Tue Dec 03 19:45:14 2013

I will put the DIR-827 on my wooden desk tomorrow for testing but that is a little close if it starts crashing again, damage could be imminent!

[FurryNutz]

Shame they wont allow BYOM.  

So you got my curiosity going this evening and I opened up my 857, Any of this look familiar to yours if you opened yours? I didn't open my 827 as it's currently online.  


I'm curious if the following mod would be any helpful in this 8 series router heat issue that some have experienced.


I see the same composite heat tab on the one chip as on the DAP-1533 and I presume other routers as well. I tried to move it and it's stuck on there pretty good. I guess thats a good sign.  

Proposed addition of metal low profile heat sinks:



I don't think the cover will allow for the bottom sink so I removed it for posterity.


Definitely no room for any mods underneath.  

Of course this would be in extreme cases maybe. It wouldn't be a hard mod to do though. It's easy to take the cover off and just apply the new heat sinks.

I might try this on the 827. I need to go to the computer store and find more heat sinks.  

Heat Sink Specs: Similar
High conductivity thermal adhesive tape
Size:14x12x5.5mm

[Patrick533]

All of these are pretty similar, just setting them on a fan has cooled mine. The way they designed it for convection cooling is a good design. The rising heat is made to pull in cold air from the bottom and that cools it off a little. A fan on the bottom blowing upward has done a much better job here. The laptop chiller idea certainly would work well and since there is a USB port on these, there is your power! The cooling on these would have been better managed with more ports for the heat to come out of. You look at my CPE modem, I can see and read all the chips through the case. Basically the case becomes nothing more then a plastic screen door instead of a fashion statement. I think the Motorola modems were the same way. It lets the heat flow much better, almost like no case at all.

But then my world comes crashing down and IPV6 has to go!

The Xbox 1 has IPV6 enabled by default, I guess for the last few days they have been really enjoying the IPV6 on the XB1 here. The XB1 basically circumvents all of my DNS filters(no pirate bay, **** sites, etc) and gave them FREE access on the net to do as they please. It is very hard to filter IPV6 because instead of having 1 IP and then a NAT, EVERYTHING gets it's own IP address randomly. So I turned off IPV6 last night and now I have some upset MEN "I don't trust on the internet"! There is more to that story then I can go into here but unless I run 2 routers, no more IPV6 HERE. At least until the MEN get out of my home and start paying their own way with a REAL job instead of living off me and Mom(rant complete).

So far the router with the fan is stable with the WiFi turned on, I will give it a couple more days. I changed the addressing last night to .0.1, there is really no difference and I also went back to firmware 04. I am still getting 2-3ms pings just off the router with IPV6 turned off. I am going to break out my old FIOS router and just hook it up to my computer to test pings and such later today. Most routers these days will do close to a gigabit from wan-lan, back in the day the FIOS routers was the only thing that would do 30Mbps. Being your pings to your router are sub 1ms, I get the feeling that this router did not make it through a California 90F summer and the fan is just prolonging the inevitable. BUT, only time will tell. 

[FurryNutz]

Try v1.05 as well.

Glads it not crashing anymore since you put on force induction cooling. 

Keep us posted.

[FurryNutz]

FYI, I took the cover off the 827 while running live and online, and that one chip with the composite heat sink read 128-130F.  

Wonder if thats ok. I think I'm going to apply a metal heat sink to that one chip...what do you think about this?

After applying the one heat sink directly on top of the other one, heat reading was around 112-113F.  This is not under any load so I'll connect some WiFi and see what happens and take some readings.

[Patrick533]

Well that is below the max that the manufacturer said it was OK to run at, but I am not sure what that thing on top is supposed to do, some new newfangled heatsink material. The manufacturer said 176F MAX, but that is going to cut the life WAY down on the part. When we do MTBF calculations, temp is one of the major contributing factors to solid state electronics death. Just 10 degrees C less can add a bunch of life to an IC on paper.

My IPV6 woes have disappeared, the IPV4 router I used to do that with showed the same ping times as the DIR-827, so long term cooling this thing is in my future since I am now running 2 subnet's, if there is room and I can find a low profile heat sink in this here box, I'm going to epoxy some aluminum on 1st just to get an idea if it is that chip or not. Don't really want to spend the bucks on a layer 3 managed switch...... Cooling the Dir-827 is now on the top of my list, but I have an employer jumping down my throat wanting an interview so it may be a day or two until I can start the project. I wish I had an IR thermometer but then again if it stops crashing, I guess that was it.