Build Your Own LAN Cable Tester

Sure, you can buy a cable tester, but what fun is that? [Ashish] posted a nice looking cable tester that you can build with or without an onboard Arduino. If you don’t use an Arduino, the project uses a 555 chip to test the eight wires in an Ethernet cable. The readout is simple. When testing a conductor, one of 8 LEDs will light. If one doesn’t light, the cable is open. If more than one light up, there is a short. Mixed up pins will cause the LEDs to light out of sequence. You can see the device in the video below.

The 555 device is fine for the design and we were surprised that the project had provisions for using an Arduino as nothing more than a pulse generator. It could replace most of the circuit which is pretty simple. A decade counter converts the pulses into 8 pulses (a wiring change makes it reset on the 9th count). The rest of the circuit is nothing more than LEDs, resistors, and diodes.

This is a great example of how a few simple components can come together to do something significant. It would be tempting to use the Arduino to create the 8 or 9 output pulses and then measure them, but that would be a lot of I/O for a small Arduino. You could watch the return with analog inputs though, so that might form a further refinement of the circuit. There could even be an advantage of allowing a more detailed analysis on each pin.

Still, such flights of fancy aren’t nearly as simple, and this is above all a straightforward and pleasant project. Sure, it won’t replace a $12,000 cable tester, but it doesn’t have to. Another simple circuit (if you don’t count the scope) that is useful for cable testing is a time domain reflectometer.

35 thoughts on “Build Your Own LAN Cable Tester

  1. Ran into an issue this week on an installation of Cat6 cable. My contractor used a $50 tester which showed that the connections were good. Went to hook up the equipment which uses PoE and it would not power up the equipment. The Ubiquiti radios and their PoE injectors confirmed what my ByteBrothers RW tester saw…split pairs on the cabling. I had the contractor put new ends on the cable and it worked.

      1. A split pair doesn’t show up with a simple continuity tester like the above, it won’t work for split pairs. The only thing that works for split pairs is testers that are specifically meant for split pairs. If you need more information about this look up about cross talk. This is due high frequency transmission and rejection that is needed to allow Ethernet to work. Having part of the transmit wires of a pair twisted with one of the recieve wires will not be a reliable, or long distance connection. Also note that under 10ft some of testers with the split test feature might miss a split pair.

        1. A pocket cat is our daily driver and is $70ish , $100 if you want to have a toner probe. They are bulky, but the cheapest that don’t use proprietary batteries or fall apart. The above tester will certainly do for a homeowner, or the occasional user. As said before you do not want a split pair as it can possibly cause intermittent issues (A network pro’s worst nightmare). It’s a 5 minute – 1 hour fix for a home user, or a multi hour fix if its a whole building.

      2. I’ve seen plenty of $5 testers marked up to $40-50 that have no concept of split pairs. They’re pure continuity, and can even miss certain oddball continuity defects.

        You’ve got to get to about the $60 mark before split pair detection becomes a feature:

        (I can specifically recommend that one as I have the older version with the Test-Um branding, ten years and going strong. Probably on its third battery by now…)

  2. I have a Black Box TS031A That I got so long ago that I do not even remember where I got it that does this exact thing. I know that I did not pay more than $5 for it. I might have even gotten it for free. It comes apart into two pieces so that a cable can be tested that has its ends in different rooms. It is not the most sophisticated tester, but it finds 95% of the problems with RJ45 cables.

  3. You should label it a continuity tester which is all it is. It is a far cry from a comprehensive cable tester. I would have used something with a brain to drive it and had to tell me if it was a regular cable or a cross over as well as continuity.

      1. Some things that could be checked for with a nice modern microcontroller if someone wants to do the design are: 1) crosstalk into adjacent channels because the pairs aren’t done correctly. 2) out-of-spec attenuation at higher frequencies

      2. If I wanted to crank out an advanced design cable continuity tester it would only take a couple of hours. It is really not rocket science. I use a more pragmatic approach. I use my eyeballs to check for crossover cables and if a patch cord does not work, I try another one. If I find a bad one I cut the ends off it to keep it out of circulation. I don’t build a lot of things because I don’t really deem them necessary and also because you can more often than not get something pre built that is as good if not better for less. In this case screwing around making a board and an enclosure etc etc for something you can grab from China for under a 10 spot just seems silly to me.

  4. Continuity testing like this one is a function that just screams DIY.

    Another cool trick is to hack up an AUI to 10-base-T adapter to perform link detection.
    In the simplest case it’s two pins on the db-15 for power. It’s technically 12v but all I’ve tried run just fine off a 9v battery.
    Despite being 10mbit, most gigabit hardware is backwards compatible enough to link with it and indicate at least half the cable pairs are working well enough.

    I’d love some of the advanced tester features like echo distance reflection for measuring in-wall cabling, and something smart enough to do basic packet checksum verification, but shelf bought gear can get pretty stupid expensive quick.

  5. Has anyone looked into reading some low level Ethernet registers of a Raspberry Pi 3B+ to get the signal quality? Just need a Gigabit switch at the other end to send a signal. Then by adding some analog circuits that connect to the POE pins, it can even measure the DC resistance of each pair.

  6. Not possible for circuit to work as shown in video. Only a single LED can ever light at a time. If open it may not light. Incorrectly crossed wires in cabling will not modify lighting sequence.

  7. The argument for testing split pairs is moot because if you are terminating you have to do both ends wrong. If not you are not doing your job. It was another persons job and they have to bear the cost. A very basic version of this device can be built using just a pp3 and 2 network boxes and scavenge just 2 leds and a resistor out of a dead “anything you can get your hands on”. In Africa there are no electronics shops.
    Used something like this with zero active parts for 2 years and only got fancy on my work once or twice with a pro grade unit. Cheap devices can take more punishment and be used by unskilled people who are allowed to open up and learn from the innards. A double win.

  8. Problem of these continuity checkers is, that they can’t detect e.g. connector with bad contact (like only half crimped).
    Just last week I had problem, that PC disconnected often and linked only to 100Mbit on 1Gbit network.
    After re-crimping one of the RJ45 it was OK, because someone didn’t pushed the wires enough into the connector…

  9. The price of the proto board vs the price of a complete kit it’s not worth building your own other than for practice, China has full kits with varying RJ plugs aswell.

    $10 kit comes with everything you could want upto 1gbps, DIY Optic splcers ???

    1. Fiber optic splicers require micrometric screw positioning…a manual one could be fashioned from cheap Ebay micrometric screws, but if you’re already out buying from Ebay, you might just get a used splicer and just clean it…

    2. Optic splicers?

      Real big-boy fusion splicers from Fujikura will cost upwards of several thousand dollars to do right.

      I agree with the competitive cost with building vs. buying a tester. That said, it is useful to go through the thought process of HOW to build your own. I’ve been in enough situations where I don’t have easy access to overnight shipped equipment and have had to improvise tools from Arduinos or other hardware.

  10. It would be interesting to mix something like this with e.g. a KSZ9893R + ESP32. That way, you could first do a basic continuity test and then use the more advanced LinkMD features to get details on any fault that cannot be debugged with continuity alone. And with the ESP32, you can then check things like DHCP and control everything via WiFi/Bluetooth. It would be like a swiss army knife for 10/100/1000Base-T.

    1. N.B.: I’m suggesting a switch-IC, because this way you don’t need a GMII/RGMII capable microcontroller, to test up to 1000Base-T. Also, with this chip, ethernet diagnostics would be possible even on a simpler microcontroller without any ethernet connectivity using only SPI/I2C.

  11. Its funny and totally wrong that you put connectors on cabling in an office environment. Totally not the correct way to do structured cable. Punch down into a keystone jack..that what you do. Not putting ends on lousy cabling jobs. lol.

  12. If you write “555 4017 lan tester” in Google you’ll find this schematic on dozens of sites. Cheap Chinese “LAN testers” use it also. It can be useful for some basic LAN troubleshooting and is fun to make because it’s simple and usually works immediately after you make it. But for serious LAN diagnostics and troubleshooting you’d need better equipment, otherwise you’d be in surprise that all 8 LEDs lit in correct sequence but your LAN port is flapping or falling back to 10M.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.