TDR for measuring cables/connectors - what is it worth to radio hams?

[tggzzz]

Time Domain Reflectometers (TDRs) are a useful tool for testing cables and connectors, but just how useful are they for radio amateurs (hams)? The reason behind my question is that I have a spare Tektronix 1502, and I want to decide whether to sell it at a hamfest or on fleabay or somewhere else.

It is a small (30x13x40cm) portable oscilloscope with a 200mV ~50ps risetime step generator. It displays impedance as a function of distance and can:

• measure impedance variations in connectors/filters/antennas/PCBs
• locate short/open circuits and damage in cables
• locate intermittent faults in cables and connectors
• locate connectors in cables

but it cannot tolerate any external voltage in the cable.

The traditionally simple first example shows measuring cable length and impedance: a 2m 50ohm cable, a connector, a 1.2m 75ohm cable, an open circuit "termination".

The second example shows a 7.5cm 50ohm line terminated in a short circuit. This is inside the Tek 1502, being the section of the stripline sampler between the BNC connector on the left and the sampling diode (SD). The short circuit is inside the BNC connector; if there is static electricity inside a cable, it will be discharged while connecting the cable, before damaging the diodes. The scale can be seen from the IC with 0.1" pitch leads. The tunnel diode (TD) step generator is on the far right.

The third example is two 50ohm cables with an SMA connector. The expanded X and Y scales indicate that the connector is "visible" as a 2.5cm ~55ohm discontinuity.

The fourth example is two SMA T connectors in the middle of a 50ohm cable. Each connector is clearly visible as a separate "capacitative dip" in the reflected voltage, in between wiggles due to the connector/cable connection. Discontinuities 3.2cm apart are easily resolved, and it looks like it could resolve discontinuities 2cm apart.

The fifth example is two traces from a subtly broken cable. Moving the cable/connector made easily visible and significant changes.

My unit is functional with unimportant (to me) limitations.

• there is case damage which I repaired with methylene chloride solvent, but it would be unwise to use it in a wet environment, e.g. the Patriot missile defence systems mentioned in a service manual
• the NiCd battery pack was defective and the undervoltage/overvoltage protection prevented the 1502 turning on. I'm only interested in mains-powered operation, I use a homebrew "battery emulator". It should be possible to replace the NiCds with NiMH cells
  
• the chart recorder works, but I don't know where you could get a fresh supply of thermal paper. The recorder draws too much current to be used without the battery pack, but can be used if a 12V PSU is connected to the battery terminals with a flying lead and 4mm banana sockets
• for "hardcopy" I replace the chart recorder with a homebrew "XY output" and a DSO or XY pen plotter

So if you are a radio ham, what , if anything, would you use such a unit for?

[w2aew]

As a ham, a TDR is quite useful - especially for determining length of hunks of coax, looking for breaks/shorts in coax, determining feedline impedance, etc.  Those are nice little TDR units - I wouldn't mind having one of my own...

[tggzzz]

I can confirm that using >=2500mAH AA NiMH cells instead of the NiCDs does appear to work. The cells I use can have a trickle charge of 75-125mA, so the 150mA from the 1502 implies they shouldn't be trickle charged for too long. Alternatively, replace the 0.3ohm resistor that defines that current with a 0.47ohm resistor.

Having read the TekScope magazine introducing the 1502, it isn't surprising that the CRT survived when the case was cracked. The CRT is mechanically isolated from the chassis so it will withstand 26 12" drops! The same article states the resolution is 0.6"/1.5cm.

[tggzzz]

Curiously and pleasingly, I've had a request to use the TDR to help diagnose a problem with a coax plus dipole antenna installation. There's only access to one end since the installation is inside a glider's fuselage.

Unsurprisingly perhaps, a TekScopes article lists "aircraft maintenance" as being a good use-case for the TDR. And, of course, one of the service manuals has "Patriot missile defence system" on the frontispiece.

[tggzzz]

Looks like we will have some idea soon: there's one for sale (no affilation etc) at the moment: http://www.ebay.co.uk/itm/Tektronix-1502-Time-Domain-Reflectometer-TDR-/282171555247 Surprisingly, and unlike mine, the batteries are functional to some degree.

Memo to self: stop replying to your own posts )

[Electro Fan]

Just replying to your Memo.  Plenty of folks out here happy to learn from your experience and thinking and always happy to see a fellow EEVer find some new gear, especially vintage Tektronix.  Good luck with the 1502.   Plus, if w2aew would like one and doesn't already have one it probably belongs on the list. 

[Electro Fan]

Not nearly as cool as the Tek TDR but could save a few $

https://www.amazon.com/dp/B00A20KO18/ref=dra_a_cs_mr_hn_xx_P3159_100?tag=dradisplay-20&ascsubtag=d4dc59c6e047ec8220edf90ac216d032_S

[tggzzz]

Not nearly as cool as the Tek TDR but could save a few $

https://www.amazon.com/dp/B00A20KO18/ref=dra_a_cs_mr_hn_xx_P3159_100?tag=dradisplay-20&ascsubtag=d4dc59c6e047ec8220edf90ac216d032_S

There are other ways to save a few dollars, if you are prepared to accept some compromises: https://entertaininghacks.wordpress.com/2015/07/27/poor-mans-homebrew-tdr-with-4cm-resolution-part-1/

I really must get around to finishing "part 2"!

[tggzzz]

Curiously and pleasingly, I've had a request to use the TDR to help diagnose a problem with a coax plus dipole antenna installation. There's only access to one end since the installation is inside a glider's fuselage.

Finally got around to doing that. Conclusion: it works (unsurprisingly), and we were able to detect small, ahem, infelicities.

The only reference we had for what a "good" dipole "should" look like was a desktop antenna; that was sufficient, but I would have preferred being able to compare with a "gold standard" glider installation.

We were able to detect the coax's impedance (75ohms), that there were no problems along its length, and none in the antenna itself.

We did spot an "upwards blip" (i.e. inductive rather than capacitative) at the near end of the coax, at the 50ohm-75ohm transition. We were able to trace this to a wire joining the shield to the connector. Waggling the coax next to the connector made clear changes on the display.

So the glider's owner is rather happy, since there's no need to crawl inside the fuselage - which, near the tail, is only as thick as your thigh

[Bicurico]

Look here: http://vma-satellite.blogspot.pt/2016/08/reflectometer-measurements-part-2.html

Any comments welcome - it was just my initial tests.

Regards,
Vitor

[tggzzz]

Look here: http://vma-satellite.blogspot.pt/2016/08/reflectometer-measurements-part-2.html

Any comments welcome - it was just my initial tests.

I'm afraid I don't quite understand what the illustrations show, in particular, the "frequency span" vs cable length. Given that it looks like you have only three cable lengths, I don't see how you can have such a smooth curve. Traditionally people show the actual data points in an "x-y scatter" graph; that allows the reader to see what is known and unknown, and doesn't rely on the vague behaviour of a spreadsheet's interpolation algorithm.

I believe your "Bandwidth 1MHz" should be "Frequency difference 1MHz". "Bandwidth" frequently has a secondary descriptor in dB; thus you say "the filter's 3dB bandwidth is 100MHz and 20dB bandwidth is 135MHz".

You may find my other, incomplete, measurements helpful https://entertaininghacks.wordpress.com/tag/sdr/

[Bicurico]

I measure the width of the "spikes". It is related to the length of the unterminated cable (or location of cable fault). I only tested with three different length (which was the amount of cables I had available with known length). I agree that "bandwidth" is not the correct term.

The curve is obviously a smoothed XY scatter graph and yes, it is kind of poor to draw it with just three pairs of data (hence why I included the table). Still, better than nothing.

I have a professional TV field meter (Rover Instruments HD Touch) with Rover Instruments CNG 70 Noise Source and their Reflectometer App. I can use it for comparisons.

Unfortunatly I did not have spare time to continue my "investigations" in this field.

Regards,
Vitor


Regards,
Vitor

[tggzzz]

I measure the width of the "spikes". It is related to the length of the unterminated cable (or location of cable fault). I only tested with three different length (which was the amount of cables I had available with known length). I agree that "bandwidth" is not the correct term.

I'd just call it "Delta-f",  "\$\Delta\$f", or "Frequency Difference".

The curve is obviously a smoothed XY scatter graph and yes, it is kind of poor to draw it with just three pairs of data (hence why I included the table). Still, better than nothing.

I have a professional TV field meter (Rover Instruments HD Touch) with Rover Instruments CNG 70 Noise Source and their Reflectometer App. I can use it for comparisons.

That's useful; I had nothing, so had to compare mine with something predictable.

If you want to do serious measurements like this, then you will find that imperfections in the cable's connector will mask what's in the cable. One technique around that is "CW Burst", a combination of a TDR and a swept-frequency reflectometer.

Unfortunatly I did not have spare time to continue my "investigations" in this field.

Yeah Life is short.

[KD0CAC John]

I'm a ham and have a Tektronics 1503C , got it cheap enough to not worry about not having the battery opt. , but would like to add it ?
I do some tower work , putting up & taking down , also scrap out a lot of coax [ taking off the outside insulation to get braid ] and find all kinds off variables , signs water , lousy coax - the braid is so loose that it must leak out / and in .
Setting up correctly so that you can pin-point a flaw helps a lot .
I also have a Sark 110 that has a TDR function , but haven't use for the TDR function yet .
If I were looking for another TDR , I would want the 1503 with the digital screen , I would be concerned about the tube issues .
And then the battery opt. the B .
They can be found cheap , if your not in a hurry , I watch eBay for about a 1 yr. , got mine for around $200 with the LAND kit & all manuals .   

[tggzzz]

Thanks for the comments about how you use a TDR. It is pleasant to hear they can be useful for hams.

I got mine a lot cheaper than $200, I'm pleased to say. No, I refuse to consider the petrol and time spent replacing caps

A working 1502 with original battery sold for £140 last week on fleabay, and "spares or repairs" can be found for similar prices!

I would be concerned about the tube issues .

What tube issues are you thinking of?

If properly reassembled (which mine certainly isn't!) then they are rainproof while operating, and can be submerged to 3ft when not operating.

Apparently the CRT is mechanically isolated from the chassis to protect against impact (IITC 26 12in dro0ps onto concrete). That's plausible since when I got one of mine a 3"*0.25inch piece fell out of the case, and there were numerous hairline cracks. I assumed it would be a write-off, but no: it worked perfectly (and methylene chloride did a very good job or repairing the case).

[Muttley Snickers]

Slightly off topic warning, over the years I have owned a number of TDRs and still currently have a Tektronix TS-100 and a 1503C (needs repair) and I picked up a Nano-tronix cable length meter off Ebay a while back and have been meaning to do a review and show some teardown pictures but simply got sidetracked with other projects, the unit was listed as recently purchased or newly acquired but after receiving the unit and finding an issue I immediately contacted the Australian distributor who were extremely helpful and were able to confirm that the unit was in fact purchased a couple of years earlier and that it should be able to measure most cable types but they did not have the same cable types at hand to compare with and verify.

After correspondence with the seller who was an electrician and a good fellow the indications were that they did not know how to use the meter properly and it never worked for them but after thorough testing I found that it did sort of work but was unable to test spooled cable yet boxed multicore, coaxial or freely run cables would show up with surprising accuracy when either metered out or compared with our other existing TDRs which can readily measure spooled cables and display other fault conditions, I spent a day or so getting the meter to read consistently and if I pulled out 20 meters from a 100 meter roll it would only see the free 20 meters but was unreliable and I found that by adding a 200 ohm resistor in series to spooled cables the reading would be within an acceptable range depending on the particular cable being tested, I put this down to the pulse possibly being induced across the cable windings with spooled cables being in very close proximity and generally tightly wound.

Most of the cables we use are factory meter marked such as Cat5e, Cat6e, RG58, RG6, mains, security cable and so on but some cables such as figure 8 are not meter marked and are provided on spools of either 100, 250 or 500 meters so determining how much is left on a spool prior to a cable run can be a bit hit and miss or you would need to run it out first which can be a pain particularly in large manufacturing plants and occupied areas. Anyway, they are also sold as a Megger TDR-900 and the more recent models also incorporated an oscillator for cable tracing. One benefit of these types of meters along with our TDRs over our other cable fault locators is that the end of the cable does not need to be strapped or shorted out to take a reading.

On most of our cable testers and TDRs the NVP for a particular cable type must be known and entered but after years of using them you soon know these values off hand or you can simply refer to a cable manufacturers data sheet, I created a small laminated data sheet specifically for our most commonly used and encountered cables. We have also found that the onboard cable libraries included with most of these testers are heavily based around the telephony industry and associated cables so the manual entry of other cable specifications can be a chore on occasions, the Tektronix TS-100 is a fine example whereby custom cable entries will be lost once the battery is disconnected and these units do draw current when not in use so unplugging the battery after each use is routine, anyway just a few examples of tested cables below and I may follow up with some more pictures and perhaps a couple of waveforms and although I don’t have a very flash scope you should get an indication anyway. 

Finally, if anyone else owns either of these cable length meters could they please verify whether they are able to measure any of the noted failed cables listed below whilst still on the reel with either the Nano-tronix TDR-010, Megger TDR-900 or similar meters, obviously these are not failed cables by any means but the particular meters inability to measure them, many thanks.             

Cable Type/ Format/ Result

Clipsal CAT 5E Cable/ Box/ Passed
Clipsal CAT 6E Cable/ Box/ Passed
Clipsal RG6 75 Ohm Coaxial/ Box/ Passed
Garland RG58 50 Ohm Coaxial/ Box/ Passed
Electra 4 Core 7/0.20/ Box/ Passed
Electra 4 Core 14/0.20/ Box/ Passed
Electra Figure 8 14/0.20/ Reel/ Failed
Electra Figure 8 24/0.20/ Reel/ Failed
Narva Figure 8 23/0.32/ Reel/ Failed 
Prysmian 1.3mm TPS Lighting/ Reel/ Failed
Prysmian 2.5mm TPS Mains/ Reel/ Failed

Note: Passed indicates accurate reading, Failed indicates unable to measure.

[AG6AV]

I use my Tek 1502, and 7S12 to test coax cables I make. I do a SOL (sort, open load) test.  Recently, I purchased a vector network analyzer, and now I use that (Agilent 8753ES) to test coax cables up to 6 GHz.

[tggzzz]

Recently, I purchased a vector network analyzer, and now I use that (Agilent 8753ES) to test coax cables up to 6 GHz.

Out of curiosity, what range does that give?

The question is somewhat moot, since I sold my spare 1502 today. The buyer seemed pleased, I'm pleased, what's not to like

[KE5FX]

Recently, I purchased a vector network analyzer, and now I use that (Agilent 8753ES) to test coax cables up to 6 GHz.

Out of curiosity, what range does that give?

The question is somewhat moot, since I sold my spare 1502 today. The buyer seemed pleased, I'm pleased, what's not to like

The longest cable that can be tested is related to the span (1 MHz - 6 GHz) divided by the # of points in the sweep, which can go up to 1601 points on the 8753C.  At 3.75 MHz/point, the time domain range for a typical cable with Vf=0.66 would be about 0.66 * c / 3.75 MHz = 53 meters.  Then you need to divide by 2 if you're making a reflection measurement like distance-to-fault.  Call it about 25 meters or 80 feet.

The resolution depends on various factors but is usually about equal to the reciprocal of the span width.  6 GHz would give you the ability to distinguish features down to about an inch apart, again allowing for the velocity factor of the cable.

[texaspyro]

• the chart recorder works, but I don't know where you could get a fresh supply of thermal paper.
  
  

Long ago, I built a little PCB that fakes the Tek chart recorder timing track signals and allows you to use dirt cheap (< $2/roll) EKG paper instead of that un-obtainum $50/roll if-you-can-find-it stuff.   I recently dug out the boards that I built at the time and will be getting them up and running.

They are not a perfect solution... there is no timing feedback so they have a pot for adjusting the timing.  Also, they can't detect end-of-paper so you have to watch the chart for the tell-tale red stripe that indicates you are near the end of the roll.

https://www.eevblog.com/forum/repair/tektronix-1502-tdr-degraded-risetime/msg1491484/#msg1491484