Is it possible to make a ribbon cable tester with only passives?

[Mp3]

I need  (want) to make a ribbon cable tester for 16 pin IDC ribbon cables....

they are super  expensive and the only affordable one i found is for 40 pin cables. That is fine but not what i need.
https://riddledtv.com/arcade/images/40pin_Cable_Tester_800.jpg

Would it be possible to make with only some LEDs, a few resistors and some pin headers to connect the ribbon cables to? and some buttons or switches to activate it?


I saw this too but it looks a lot more complicated.
https://www.gadgetronicx.com/idc-ribbon-cable-tester-circuit/

[Rerouter]

If you just need to make sure no connections are open, then yes you can do it with passives,

If you want to make sure there are no shorts / reversed connections, then you would need something like a decade counter and a 555 clock, so it gives each pin a positive in sequence,

[AVGresponding]

You don't need anything more than resistors, diodes, LEDs, and switches.

Start by drawing out a truth table, that will help you design the diode/switch logic.

Bear in mind, the higher the number of conductors, the more complicated your switches are going to need to be!

[soldar]

Or just put a push-button switch on each line and test each line manually.

[Shock]

Send data over 8 wires with a loop back connector using the other 8 wires.

Edit:

Or what about 2 wires looped back 7 times.

[Mp3]

Edit:

Or what about 2 wires looped back 7 times.

How would one accomplish this? I feel like a 555 can handle this, but i'm not quite sure where to start.

It seems like it could be very elegant.

[soldar]

Send data over 8 wires with a loop back connector using the other 8 wires.

Edit:

Or what about 2 wires looped back 7 times.

How would that detect a short between two adjacent wires?

[janoc]

Send data over 8 wires with a loop back connector using the other 8 wires.

Edit:

Or what about 2 wires looped back 7 times.

How would that detect a short between two adjacent wires?

Or wires crossed twice - the second swap would cancel out the first one and the cable would be still bad while you are none the wiser ... Probably not a likely problem with IDC connectors, but the OP didn't say what kind of faults they want to test for.

[soldar]

i'm sure it could be done with a simple IC, but i only know microcontrollers!

Where can I find "passive" ICs and microcontrollers? I never heard of them.

[soldar]

One switch per line. One actuator closes the even numbered switches, another actuator closes the odd numbered switches.

[Muttley Snickers]

I was thinking that perhaps a basic LAN Cable Tester would suffice to quickly check for opens, shorts and crossed lines between each connector. The only drawback is that you would need to conduct two tests on each ribbon cable as most of these testers are limited to only four pairs or eight wires.

[Mp3]

I was thinking that perhaps a basic LAN Cable Tester would suffice to quickly check for opens, shorts and crossed lines between each connector. The only drawback is that you would need to conduct two tests on each ribbon cable as most of these testers are limited to only four pairs or eight wires.

Interesting idea....!

I could make a board which connects each end of the ribbon wire to some switches which would switch between the low 8 and high 8 pins on each 16 pin ribbon connector?

It's a little less DIY than making the whole tester from scratch, but might be more reliable for finding shorts.

[Rerouter]

Or if you want to go back to micros e.g. a atmega328, have 2 shift registers to feed the pins on one side, and wire to the micro inputs on the other side, you can check for every possible short open or cross,

First test shorts / crossed wires, enter a 1 to the shift register and clock it along, each clock the micro should see only the expected pin go high
Then test poor connections, clear the shift register so it is low on all pins, turn on the pin pull ups for each micro input, and read them, if any are not low, then you have a poor connection.

[Mp3]

I think i have a slightly better grasp on this.

I think i will base it on this: https://cdn.instructables.com/FYK/WOL6/JLMKK0XP/FYKWOL6JLMKK0XP.LARGE.jpg

And then i guess all i need to do is figure out the best way to test all 16 conductors. Since this circuit is for an ethernet cable, it has no problem stopping at 9 conductors.

I'm imagining i would need a ON-ON, SPDT switch to somehow switch between the top & bottom 8...

[Rerouter]

If you want to go the micro route, here is a schematic I just through together (MISO would act as the shift register latch, assuming your just going to bit-bang this)

Edit: you would add something like 220 ohm resistors to each pin on 1 plug so a short doesn't overload the shift register

[Mp3]

Wow man.... thank you! that's amazing, though i barely understand it  I will spend a while studying and get to understand what's going on!!

Would that linking of two 74HC595 work in a non-micro based circuit as well? I assume it would. What if i just used one 74HC595 and used a 8-pole switch, or a SPDT ON-ON connected to an analog multiplexer IC to switch between the two sets of 8 conductors so i could use just one 74HC595?

(I'm not against your design at all, just trying to understand the possibilities and how it works.)

[soldar]

I cannot think of anything simpler than this:

[Yansi]

If you want to go the micro route, here is a schematic I just through together (MISO would act as the shift register latch, assuming your just going to bit-bang this)

Edit: you would add something like 220 ohm resistors to each pin on 1 plug so a short doesn't overload the shift register

What do you need the crystal for and where is all the decoupling?

[Rerouter]

You likely do not need the crystal, that was just so it would work with the usual arduino bootloader, e.g. pop the chip off an UNO board and pop it in his circuit. This was an assumption to make it more available to beginners

Decoupling like the output resistors for the shift registers got missed because I literally threw it together in a few minutes.

[janoc]

I cannot think of anything simpler than this:
(Attachment Link)

That won't detect swapped wires (maybe not a bit problem with IDC connector) nor a short between adjacent wires.

[soldar]

That won't detect swapped wires (maybe not a bit problem with IDC connector)

Not a problem with ribbon cables which is what we are discussing.

Correction: Yes, it would as the wrong light would turn on.

nor a short between adjacent wires.

I believe you are mistaken. Think about it. One push-button switch tests the continuity of the even-numbered wires and the other one the odd numbered wires. If there is a short between adjacent wires then any switch would light both lights.

Correction: You are right. The lights would need to go at the other end, connected to gnd. That would solve it.

[Shock]

I like the way you took my idea of the loopback and made the worst possible implementation of it by connecting adjacent traces.

You can measure voltage drop with a mV meter to characterize the length and connection quality. But for simple signal integrity A PC can be used to test data over serial/parallel/ethernet loopback with existing software. I don't see a point in making a dumb tester that really tells you nothing about the cables performance.

[hamster_nz]

Rather than an switch for each wire, just use three buttons?

First buttons test wires 1,4,7,10...
Second buttons tests wires 2,5,8,11,....
Third buttons tests wires 3,6,9...

It would pick up any open wires, shorts to a neighbor, and shorts to neighbor's neighbor.

It could also pick up a backwards error, as the LED 1, 2 & 3 will not light in the same order as the buttons are pushed.

You could most likely get away with two buttons to test odd and even wires, but having seen the inside of an IDC connector I think that a short between might be possible with a bent 'barb' that pierces the cable (or whatever they are called).


EDIT: A link to another thread on the topic https://www.eevblog.com/forum/projects/ribbon-cable-idc-connector-failure-modes/

[soldar]

I like the way you took my idea of the loopback and made the worst possible implementation of it by connecting adjacent traces.

Are you talking to me? And, if so, can you explain what is wrong with my idea? You know, rather than just saying it is the worst.

[Moshly]

You could use a multimeter and have a big series loop with a different value resistors at each node, & just measure the total resistance.

[latigid on]

Solid-state cable tester?



Pretty simple really, counters and drivers with good cables indicating a diagonal line. Bad cables have dark LEDs, LEDs lit off-axis (short circuit), offsets, or inverted from the other corners.

https://www.midiphy.com/en/shop-details/177/84/midiphy-wirescanner-pcb-midiphy-wirescanner-idc-micromatch-ribbon-cable-diagnostics

[cdev]

Yes, if you also have some electricity, you can use passive LCD (thermal sensing) material.

Cables with continuity will heat up, those without wont.

I need  (want) to make a ribbon cable tester for 16 pin IDC ribbon cables....

they are super  expensive and the only affordable one i found is for 40 pin cables. That is fine but not what i need.
https://riddledtv.com/arcade/images/40pin_Cable_Tester_800.jpg

Would it be possible to make with only some LEDs, a few resistors and some pin headers to connect the ribbon cables to? and some buttons or switches to activate it?


I saw this too but it looks a lot more complicated.
https://www.gadgetronicx.com/idc-ribbon-cable-tester-circuit/

[vk6zgo]

You could use a multimeter and have a big series loop with a different value resistors at each node, & just measure the total resistance.

But, but, that isn't complex enough for this forum!

[perieanuo]

hi,
simpliest, just send 5V voltage on some wire  and at the other end do a global short, back on input side put leds+resistors. if by hazard the 5V wire is bad reverse the cable.
simple is good.
of course if you want testing the current capability, add on each wire some charge
like others are saying, with an arduino you can make a smarter test in 11-2 hours using the gpib's and some buffer transistors or scrapped uln/udn chip (it takes half an hour to make the sw and maybe a hour to solder the board between arduino and cable).I suggest you test continuity AND current capability on each track.
I used to manufacture in one company a machine with 10-way ribbon cable, never had an issue with them (the testing was done after the ribbon was assembled inside and we tested the machine for minimum 8h, so it was no need to check the ribbon before). never had a ribbon dead and the machine was a laser alignement system for radio-therapy. the ribbon was of course chinese fab (ordered usually by 1000 pcs). I guess you're concerned with no reason or your app is military-one !?
regards,pierre