Home Brew Cable Tester that Really Works

[FrankB]

We all have needed a cable tested from time to time. Is it broken or intermittent, and if it is, which end is at fault?

Well, just take an old chassis, cake pan or whatever and install chassis mount plugs and sockets on it for your commonly used cables.

I would suggest PL 259/So 239, BNC, RCA, USB, 4 & 6 wire telephone, 1/4" Mini, & sub mini; mono & stereo, Even AC cord conn. etc.

Now you could use a battery & LED to indicate continuity. BRRRRRP! (Wrong!!
Yes, they do make some spendy commercial units that use them.
No good!)

The reason is: NO CURRENT DRAW. Yup, you can have low current connectivity, but when there is a current draw, the connection fails. (Been there, Have both the tee shirt and cap).

I suggest using a salvaged filament xformer. 5 or 6.3 volts is nice. (Be sure to fuse it, too.) And use some pilot lite sockets for say any 6.3 V or even up to 8 V pilot lamp that draws 100-200 MA. (#47 or #1847 bulbs are really common & cheap.)

To use it is simple: Plug your cable into it to test. Turn it on, and look at the lights. Wiggle the cable at the connectors
If the lights stay on solid, its good. If they blink when you wiggle the cable or move the ends, the cable is bad.

If the fuse blows, the cable or connector is shorted.
(The only down side to this simple tester is that it does not tell you if you have crossed wires in a multiple wire cable. However one can do that also with a multi-position rotary switch and a little design work.)

Now you will only need as many lights as the cables you want to test have wires & grounds in them. One can also add switches and shift the cable wires around to reduce the number of lamps needed. I have made them with up to 25 lights.

The nice thing about this project is you can likely find everything you need to build it from a well stocked junk box, spares, or a hamfair.

With a little additional time and work, you can add a pilot lite socket on top and a switch to check 6.3 V pilot lamps with also.

[TonyDuell]

A couple of thoughts :

1) You could use LED+series resistor as the indicator if you wired a suitable resistor in parallel with that combination, so as to draw sufficient current through the cable

2) I don't like the idea of deliberately blowing a fuse on a shorted cable (which is a very common failure. How about wiring a bulb in series with the transformer secondary as well as one at the end of the cable. For a good cable, both bulbs end up in series and glow. For a shorted cable only the one at the transformer end glows (and it will limit the current, of course).

[lesmw0sec]

I would recommend a good old fashioned door bell and battery. You can keep your eyes on the connections and just listen. Not a good idea though if any semiconductors are on one end though, given the back EMF!

Les.

[Craig Sawyers]

A cable tester from 1974 at CERN. Discrete TTL and a few transistors to cope with up to 244 pin connector assemblies.

Darn - the pdf exceeds 1.9MB, which is the list limit for this type of file.

Best laid plans etc.

Craig

[ms660]

I've always tracked down cable faults with a multimeter.

Lawrence.

[karesz*]

Hi Craig,
Can you transcode it as DJVU please? _or make 2 pockets of them...
Karl

[Biggles]

We have the same problem on telephone lines at work. They check out ok on a loop test, but can't supply the current required to maintain the loop to operate the phone due to a high resistance which won't show up with a meter. I designed a bit of gear to get round this which injects 20mA round the loop and measures the current at the same time. This does a "real" check on the pair. The unit also supplies 50 volts to a pair and measures the current drawn which should be zero on a good line. This checks for line insulation problems (ring trip etc) which again don't always show up with a simple resistance check with a multimeter. The last check looks for a standing voltage on the disconnected pair, which should be zero. If it shows a voltage this could be down to a contact fault. Saves an awful lot of my time when trying to find a good spare pair on a cable and saves lugging a lot of test gear usually across a muddy field.
Alan.

[Oldcodger]

IN my day, as a TTA, we were told that the BEST way to find high current /resistant faults was to use a high current meter.(ms660- multimeter shows up ones that are not respondent to high current use That way you draw current through a fault and the point of volt drop can be found. Modern day HIGH impedance meters DO NOT find this.
Biggles- we did this sort of test on remote UAX in the 70's using Tester SA 7083 , CONNECTED TO A TEST FRAME. Battery test- shows up presence of contacts between pairs. Balistic test- shows up both legs of line equal. Earth test- looks for earth on each leg of line.
I'd suggest that a bit more than 50v is needed for an insulation test ,as is reverse voltage tests to look for rectified loop at higher voltage ( as in when line is rung)
But nothing compares to faulting on the inner Hebrides- on a manual board with magneto ringing. Maggie ( phone operator) says that Old Donald's phone won't ring. Next question- is handle easy or hard to turn. Easy and chances there's a wire down. Hard and it's off to the foreshore to pick up a stick-to throw at the tangled wires.

Just some reflections, but then I've been in the game for 50 years, seen phones from Magneto ones and ones that need a local battery to Analogue + Digital systems and Voicemail.

But on the more archaic multi channel systems, the lower frequency systems wil again give a clue to the fault type. 12 CHANNEL carrier ( YE OLDE colonial systems often had more than one system operating on sets of overhead. ). LF system working at low gain, with HF system failing= almost certainly a wire down.

[Biggles]

I should really have built a ring current generator for a full test but wanting to keep portability up and complexity down it had to do with 50 volts. Actually this is generated by connecting two small 5V-24V converter modules (RS) in series. The whole idea was to avoid humping more gear around than was absolutely necessary. I'm sure you will know where I am coming from!
Alan.

[ms660]

Originally Posted by Oldcodger

(ms660- multimeter shows up ones that are not respondent to high current use That way you draw current through a fault and the point of volt drop can be found. Modern day HIGH impedance meters DO NOT find this.

Not sure if I'm with this but I would do cable voltage measurements with the cable connected to the load it's supposed to be supplying.

Lawrence.

[FrankB]

You can't always test with the original load on the cable. If you disconnect the load, you can determine if its the cable/wiring or the load. Especially when you are dealing with many variables.

Now here we have to use this sort of rig, as some of the antennas I placed are 150' (+) up in the top of a fir tree. Add that with lightening strikes, pre-amps that fail from lightening or the weather, (Here in Washington State we have an odd thing we call rain. We are located on the Southern edge of the Olympic Rain Forest. Its not unusual to get 1.5+ " of rain in a day here.
Not to mention climbing up & down the tree is physically taxing, and dangerous. What is fun is when the limb you are standing on AND the limb you grab break at the same time. (I have that tee shirt & hat) Spurs and safety harness are a necessity, along with a wire core safety rope.
Sometimes we have hooked the coax up to a car battery, and used a headlamp from a car up at the top to check for voltage & current. Add in failed connectors, splices, grounding blocks, squirrels (aka Tree Rats), and the like it makes for a challenging job.
Also replacing a long run of cable in a tree is SO much fun..especially in the wind & rain.(My friend & old co-worker, an old Navy man, actually got seasick at the top of a tree from the wind! He was up about 150' or more. Humorous afterwards, yes, but very dangerous.)
Thankfully, I am now retired, and leave the climbing to the youngsters.