On 10/13/2022 2:39 PM, wAYNE wrote:
On 10/12/22 9:22 AM, Paul wrote:
On 10/11/2022 7:40 PM, wAYNE wrote:
On 10/11/22 4:57 PM, Killadebug wrote:
On Tue, 11 Oct 2022 12:29:13 -0400, wAYNE wrote:
Ok, managed to get everything working again although not quite to my >>>>> satisfaction.
Apparently, I had to go into the router and "unblock" my desktop, since >>>>> it uses a wired, and the laptop uses the WiFi. Once that was done, now >>>>> both PCs can simultaneously connect.
I'm not happy about having to enable WiFi on the laptop to get
connectivity. Any thought on why the wired no longer works would be >>>>> welcome. As I mentioned, it seems to not work wired on either Win or >>>>> Ubuntu, whichever I boot into. Switching out cat5 cables did not
remedy. Could it be a connector, motherboard, or bios issue?
A separate question relating to Wifi frequency. I'm on the old Dell >>>>> Inspiron 1545 laptop. In the Fios router, it seems to have both 2.4 Ghz >>>>> and 5 Ghz. I set up for 2.4. Should I stick with that, or try the 5 >>>>> Ghz connection and, if so, what's the advantage if any? As I said, I >>>>> haven't used or configured WiFi in years.
Thanks in advance.
% Ghz will give much better thru put, I would use that if your laptop is >>>> close to your router. % Ghz gets progressively weak the more walls it
must go thru.
As far as your wired connection gos, di you try a different network port >>>> on your router? You also may want to check the wired port speed settings >>>> on your router for the port that is giving you trouble. I always hard
code the port speed and duplex. Auto-negotiate is kinda of picky that is >>>> why I had code.
Yes, I did try different router ports but no difference. I'm now thinking that this is the charger board in the laptop as it contains the power jack, USB port and ethernet port. It is a small board that can be replaced and I have a replacement on
the way. I just got finished swapping out the defunct motherboard and it's possible that the charger board is causing trouble.
That's one weird board. The Ethernet port connector looks
like a trouble-maker. Like the wires are SMT and underneath
the connector body.
https://www.dcxpower.com/dell-inspiron-1545-1546-pp41l-484aq03011021c11c21-484aq13csc-484aq200sb-484aq230sb-dr1-0853012sc-dc-power-jack-socket-connector-vga-rj45-lan-usb-port-in-charging-board-p-194.html
Paul
I've been keeping at least one 1545 going for about ten years now. It's cheaper than buying new. Although I've replaced motherboards twice, this is the first time since I've owned one that I had trouble with the port. I found out today that I
could have used an ethernet to USB adapter, which I didn't know about, but the charger board is already on the way. Of course, I'll have to repeat the disassembly process like I'm changing out the motherboard, but, at this point, I'm used to it.
The most annoying thing about the 1545 was the CMOS battery replacement. Just to change it required almost complete disassembly, but I remedied that by using a small dremel tool to cut a square into the casing so that, when the CPU/ memory cover is
removed, the battery can be changed out as well if needed. There is no visible modification once the CPU cover is back in place and infinitely much easier battery changes! Of course, the modification has to be done while the motherboard is out of the
unit; otherwise surely the metal shavings would short something.
There is a way to test Ethernet wiring.
One of the Marvell NIC chips, has the ability to check the
continuity of Ethernet wires and transformers. (The keyword is
VCT in the documentation...)
On this page, in the lower left corner, is a download button
for the Marvell whitepaper on Virtual Cable Tester.
https://www.teklib.com/library/vct-marvell-virtual-cable-tester-technology-for-gigabit-networks-white-paper/
Name: Virtual_Cable_Tester_Marvell_WP.pdf
Size: 743522 bytes (726 KiB)
SHA256: 4CA96A44A49ED3EE54055F275DCB89BFEB01299338ABFF1C3B9FE5420EFDE689
To use the tester, you do the following:
1) PC with particular Marvel VCT NIC is powered and conducts the test.
2) Run Ethernet cable to a second piece of equipment.
3) The second piece of equipment is *unpowered*. This prevents
HeartBeat activity from screwing up the pulses.
4) Cable has a max of four pairs (i.e. GbE). The status of
the pairs can be Short, Open, Good (characteristic termination).
Marvell ----\||/--------------------\||/----+-------+ Conceptual
VCT /||\ /||\ | Foreign diagram of
1of4 pair \||/ \||/ Rterm XCVR method
----/||\--------------------/||\----+-------+
I only own one of those chips, and the motherboard is dead,
so the chip is now an orphan. But what was pretty funny in
my mind, is when the motherboard was new, one of the first
features I tested was VCT... and the Ethernet connector had
dirt on one pin, and one pair measured open at zero feet :-)
So it did not pass the test, on the very first usage.
Just inserting and removing the connector five times,
cleaned off the manufacturing residue. Then it passed.
But at least it does show, that it is possible to implement
Time Domain Reflectometry inside a NIC, for testing that
a wire is broke somewhere. It can show you, for example,
that a wire is broken inside a 100 foot long Ethernet cable.
The distance to the trouble, is only a rough estimate.
It's not as good as a real TDR instrument with 35pS
rise time tunnel diode stimulus.
After giving up on Ethernet, Marvell today does seem to be
interested in NICs again. But perhaps it will be some
time before we see a product with the Marvell name on
the box. (A couple years ago, there were boxes with
Marvell printed on them, you open the box, and RealTek
was inside...) Whether they will have VCT, is another matter.
Paul
--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)