Ethernet cable testers are dime a dozen, but none of them are as elegant and multimeter-friendly as this tester from our Hackaday.io regular, [Bharbour]. An Ethernet cable has 8 wires, and the 9 volts of easily available batteries come awfully close to that – which is why the board has a voltage divider! On the ‘sender’ end, you just plug this board onto the connector, powered by a 9 volt battery. On the “receiver” end, you take your multimeter out and measure the testpoints – TP7 should be at seven volts, TP3 at three volts, and so on.
As a result, you can easily check any of the individual wires, as opposed to many testers which only test pair-by-pair. This also helps you detect crossover and miswired cables – while firmly keeping you in the realm of real-life pin numbers! This tester is well thought-out when it comes to being easily reproducible – the PCB files are available in the “Files” section, and since the “receiver” and “sender” PCBs are identical, you only need to do a single “three PCBs” order from OSHPark in order to build your own!
Bharbour has a rich library of projects, and we encourage you to check them out! If you ever want to get yourself up to speed on Ethernet basics, we’ve talked about its entire history – and we’ve even explained PoE! After some intensive learning time, perhaps you can try your hand at crimping the shortest Ethernet cable ever.
Is this cheaper or easier to use than an offshelf product?
From my experience, I’ll put it this way – this kind of adapter lets you do way more than the cheap testers can, and it’s cheaper than the testers which let you do the things that this one can!
I’ve checked the description of the $7 and $10 testers, also the description of the project. It seems they all do open and short circuit testing?
I’ve used two types of cheap off-the-shelf testers – the “scroll LED through 8 individual wires” one and the “four LEDs that indicate pairs” one. Former wouldn’t notice if wires were reversed or shifted by one, latter IIRC wouldn’t notice reversed or swapped pairs. Basically, this tester will protect you from basic wiring errors quite well. Plus, if you ever do passive POE for your own stuff, you can check that too ;-P
Wouldn’t it work better with a Simpson 260 meter? At least the display is more impressive.
I’d be afraid to take a Simpson 260 up on a ladder! Out of curiosity, I checked do see if the voltage divider was OK with the loading from an analog meter and it is fine. I have been using the 10MOhm digital meters for so long, I forgot how to do the input impedance calculation. My little Triplett 310 hand held analog meter shows 240KOhms input impedance on a 12VDC scale which makes sense for 20KOhms/Volt now that I think about it.
You don’t need to be afraid up on the ladder with the Simpson. Everyone below you needs to be afraid!
Ethernet is not only about DC continuity, it’s a whole lot about AC characteristics, Otherwise, you wouldn’t need different speed classes, and that needs a reflectometer for accuracy: https://en.wikipedia.org/wiki/Time-domain_reflectometer
I have seen those on the cheap too, even on Hackaday: https://hackaday.com/2020/12/14/dead-simple-time-domain-reflectometry-with-just-a-battery-and-an-oscilloscope/
I made something similar and much, much less elegant to try to trace cat 5e wiring in a home we purchased. If you have a known set of runs, or a simple cable you’re testing, this will be great!
However, from experience, if you’re dealing with an unknown network, it’s easy to waste a great deal of time when spending $50-$100 on the right meter will save days worth of running between rooms and/or floors. Simple voltage or resistance based tests won’t show things like wiring that’s spliced into loops or has unterminated connections behind walls that are reflecting signals and disrupting packets.
Sorry if this is a stupid question, but would this work with Cat6 cables? It says it’s for Cat5 cables?
Yes, CAT5 and CAT6 have the same pins. The only difference is that CAT6 has special isolation and the twisted pairs are more twisted. Its also sticky haha
Yup, it works for cat6 too. Cat6 is just a subset of ethernet cable with better electrical specs.
I don’t see any reason that it wouldn’t work. AFIK, the main difference between Cat5 and Cat6 is the signal integrity
Any 8 conductor cable with a RJ45 plug on bud! CAT5, 5e, 6, 6a, etc.
No stupid questions, only learning opportunities!
Yes, but it won’t be able to distinguish CAT5 and CAT6 like a more expensive meter might be able to do.
Nine volts is almost ten fingers on two hands?!
Yeah I don’t get what the 9 volts has to do with 8 wires. A volt could easily have been defined as 2.5x or 1/6th of what it is and there’d still be 8 wires.
9 volts can easily be buck converted down, volt-by-volt, to supply the multimeter with a stepped voltage numbering system that reflects pin numbers. It’s that simple.
I’ve never done surface mount soldering before. Are those resistors a reasonable job for a first-timer or would this be a good excuse to get one of those adafruit miniature hot plates?
The 0603 resistors can be soldered by hand. A soldering iron with a very fine point and pair of fine pointed tweezers are needed to hold the part in place while soldering. The hard part for me is seeing them. One of those magnifier hoods helps a lot. One of the lamps with the magnifier lens in the middle will also work.
In my opinion, surface mount resistors (unless they are super tiny ones like 0402 or smaller) are easier than through hole parts. Just put a small spot of solder on one of the two pads, then hold down the resistor with tweezers while you reheat the solder. Once done, you can solder the other side. I always like to flow a little fresh solder into the first connection as a last step to make sure the first connection is not “cold”.
That hot plate would be cool to have, but I’ve not needed one yet…
I also believe that solder paste is generally better for smd components. Put a small blob on, put the component on, heat up and it just works…
In my opinion, this could lead one to think the cable is good when it is not. I had an IT tech years ago make several cables and most of them failed. They said they prefer to buy them because hand made cables are not reliable. I looked more into it, they never used the correct pairs, they just said as long as the colors are the same on each side then it doesn’t matter. Split pairs all over the place! Well, this tester would pass that cable.
This tester is only for a wire map, this does no where near a fluke tester is capable of that costs around $20k and is used by low voltage electricians everyday, it’s cool if your getting into it but no where near practical. Fext and next is so much more important than wire map.
I had to test out my under the carpet CAT5 extension cable the other night. It is a RJ45 connector soldered to a piece of ribbon cable (that is flat to go under the carpet) that exits to a CAT5 and spliced to a short CAT5 with RJ45. I had some poor connections in the cable that is around tens of ohms vs 1 ohm, so only getting 100Mbps connections vs 1Gbps.
This only check the signal order, not quality of electrical connection i.e. good crimping, oxidized cheap connector, or poor cable quality (e.g. broken, oxidized).
BTW one trick I use to make extension cables: Connect the wires to the right pins on one side making sure it is correct, plug both ends together. I can now use the multimeter to check which wires goes to the other connector instead of relying on wiring diagram.
At first sight I thought clever idea!
Then I thought about using an analog input on a microcontroller to automate the test.
Then I thought, if you use separate digital pins per-input neither the voltage divider nor the analog pin are needed.
Then I realized that just re-invents the typical ethernet cable tester.
Having a tester like this in your toolbox can help you set up cable connections smoothly and efficiently. Thank you for sharing this post.