Odd Coax Cable Behaviour

[Steve Kavanagh]

I have a roll of old coax I've used for making assorted cables.
Recently, I've had a couple of short cables (used for DC control
applications) that show resistances on the order of 100kohms between
centre conductor and shield, low enough to upset the operation of the
equipment with which they were used.

Taking a careful look at this I notice this old cable is marked "ITT
SURPRENANT DIV RG-174/U L.N.". I take it that the "L.N." must mean
I've stumbled across the sort of low noise coax used with
accelerometers and the like. It would seem that the process of
soldering connectors to the cable, or this plus the application of a DC
voltage to the cable for some time, is causing the dielectric to become
conductive on occasion. Other cables I have made up from the same
RG-174/U are fine, however. Cutting the connector off the cable
appears to restore a >20 Mohm resistance, too.

Is anyone familiar with this sort of cable (this one has a uniformly
black dielectric) ? Any ideas as to what might be going on and how I
could avoid the low-resistance problem ? (other than buying some new,
normal cable....I have lots of the old stuff I would like to use up)

Steve

 

[Tim Wescott]

I have a roll of old coax I've used for making assorted cables.
Recently, I've had a couple of short cables (used for DC control
applications) that show resistances on the order of 100kohms between
centre conductor and shield, low enough to upset the operation of the
equipment with which they were used.

Taking a careful look at this I notice this old cable is marked "ITT
SURPRENANT DIV RG-174/U L.N.". I take it that the "L.N." must mean
I've stumbled across the sort of low noise coax used with
accelerometers and the like. It would seem that the process of
soldering connectors to the cable, or this plus the application of a DC
voltage to the cable for some time, is causing the dielectric to become
conductive on occasion. Other cables I have made up from the same
RG-174/U are fine, however. Cutting the connector off the cable
appears to restore a >20 Mohm resistance, too.

Is anyone familiar with this sort of cable (this one has a uniformly
black dielectric) ? Any ideas as to what might be going on and how I
could avoid the low-resistance problem ? (other than buying some new,
normal cable....I have lots of the old stuff I would like to use up)

Steve

Flux + humidity + time = conductivity, sometimes. This is why you clean
boards ever so carefully before use. It could be the soldering itself,
not the dielectric at all.

 

[Steve Kavanagh]

Flux + humidity + time = conductivity, sometimes. This is why you clean
boards ever so carefully before use. It could be the soldering itself,
not the dielectric at all.

True...but in one case the resistance was there within a few hours (or
even minutes...I don't recall).

Steve

 

[John Woodgate]

I read in sci.electronics.design that Tim Wescott <[email protected]>

Yes: the dielectric is coated with graphite. Check carefully; usually
there is a very thin surface film that carries the carbon, and you can
cut that back at the termination so you don't get the leakage
resistance.

 

[Steve Kavanagh]

Thanks, John. Indeed, upon closer inspection the dielectric has a
black, conductive layer over a very small diameter layer of ordinary
clear polyethylene. With care and attention it can be cut back
independently. I've done so on one cable that was giving me
trouble...we'll see if it stays OK.

I suppose it won't matter much on low frequency signal cables or low
impedance DC stuff, though I imagine the high frequency attenuation may
be pretty high and there might be impedance issues at RF.

Regards,
Steve

 

[Steve Kavanagh]

Thanks, John. Indeed, upon closer inspection the dielectric has a
black, conductive layer over a very small diameter layer of ordinary
clear polyethylene. With care and attention it can be cut back
independently. I've done so on one cable that was giving me
trouble...we'll see if it stays OK.

I suppose it won't matter much on low frequency signal cables or low
impedance DC stuff, though I imagine the high frequency attenuation may
be pretty high and there might be impedance issues at RF.

Regards,
Steve

 

[Joe McElvenney]

Hi,

Is anyone familiar with this sort of cable (this one has a uniformly
black dielectric) ?

There is a type with a black carbon-loaded inner sheath
designed to reduce noises due to cable movement. It possibly
absorbs charges due to local changes in cable capacitance (the
parametric amp principle) but there I'm only guessing. Maybe the
black stuff has migrated or perhaps it was there all the time and
your previous projects were not so choosey.


Cheers - Joe

 

[John Woodgate]

I read in sci.electronics.design that Steve Kavanagh

Thanks, John. Indeed, upon closer inspection the dielectric has a
black, conductive layer over a very small diameter layer of ordinary
clear polyethylene. With care and attention it can be cut back
independently. I've done so on one cable that was giving me
trouble...we'll see if it stays OK.

It should do. I have some that I acquired many years ago and it's still
OK. Not the same spec as yours, but similar.

I guess IT&T bought the original company, because it wasn't ITT France
that offered it to me.

I suppose it won't matter much on low frequency signal cables or low
impedance DC stuff, though I imagine the high frequency attenuation may
be pretty high and there might be impedance issues at RF.

No, it's OK at RF. The carbon layer is just an addition to the shield. I
think this type of cable was originally developed in France, and I was
offered some more than 30 years ago as 'Cable Ecran Haute-Efficace' -
'EHE cable' [Cable with high-efficiency shield].

The versions with no copper shield but just a drain wire aren't too
good; the drain wire creates an asymmetry in the structure that causes
the efficiency of the shield to diminish.

 

[Adrian Tuddenham]

I have a roll of old coax I've used for making assorted cables.
Recently, I've had a couple of short cables (used for DC control
applications) that show resistances on the order of 100kohms between
centre conductor and shield, low enough to upset the operation of the
equipment with which they were used.

Taking a careful look at this I notice this old cable is marked "ITT
SURPRENANT DIV RG-174/U L.N.". I take it that the "L.N." must mean
I've stumbled across the sort of low noise coax used with
accelerometers and the like. It would seem that the process of
soldering connectors to the cable, or this plus the application of a DC
voltage to the cable for some time, is causing the dielectric to become
conductive on occasion. Other cables I have made up from the same
RG-174/U are fine, however. Cutting the connector off the cable
appears to restore a >20 Mohm resistance, too.

Is anyone familiar with this sort of cable (this one has a uniformly
black dielectric) ? Any ideas as to what might be going on and how I
could avoid the low-resistance problem ? (other than buying some new,
normal cable....I have lots of the old stuff I would like to use up)

I think you have got cable with a partially-conductive black coating on
the outside of the dielectric. This is in contact with the screen and
it is there to prevent charges building up on the outside of the
dielectric and causing electrical noise when the cable is moved.

The coating must be carefully stripped back a short distance at the plug
to prevent a partial short circuit between the screen and the centre
conductor. If you haven't been doing this, it would explain your low
resistance readings.

Are you sure this is really co-ax? Most of the older low-noise cables
of this type were audio cables with an unspecified impedance and may
well give trouble in digital or HF circuits due to being unsuited to the
terminating impedances.

 

[Robert Baer]

I have a roll of old coax I've used for making assorted cables.
Recently, I've had a couple of short cables (used for DC control
applications) that show resistances on the order of 100kohms between
centre conductor and shield, low enough to upset the operation of the
equipment with which they were used.

Taking a careful look at this I notice this old cable is marked "ITT
SURPRENANT DIV RG-174/U L.N.". I take it that the "L.N." must mean
I've stumbled across the sort of low noise coax used with
accelerometers and the like. It would seem that the process of
soldering connectors to the cable, or this plus the application of a DC
voltage to the cable for some time, is causing the dielectric to become
conductive on occasion. Other cables I have made up from the same
RG-174/U are fine, however. Cutting the connector off the cable
appears to restore a >20 Mohm resistance, too.

Is anyone familiar with this sort of cable (this one has a uniformly
black dielectric) ? Any ideas as to what might be going on and how I
could avoid the low-resistance problem ? (other than buying some new,
normal cable....I have lots of the old stuff I would like to use up)

Steve

It has been over 20 years, but the "LN" which stands for "low noise"
(as mentioned), means that the insulator is impergnated with some
carbon, to drain of piezoelectric, triboelectric and other induced
voltages from bending and/or crushing.
You either live with the characteristics, or use standard coax when
the LN could cause problems.

 

[Phil Hobbs]

It has been over 20 years, but the "LN" which stands for "low noise"
(as mentioned), means that the insulator is impergnated with some
carbon, to drain of piezoelectric, triboelectric and other induced
voltages from bending and/or crushing.

Triboelectric it is--as in rubbing a balloon on your hair and sticking
it to the ceiling. Cable flexing and vibration causes changes in
capacitance and triboelectric charging. If you have a significant bias
voltage between shield and centre conductor, the V*delta C noise is
usually much worse than triboelectricity, but you can get rid of it by
getting rid of the bias voltage. Triboelectric noise occurs even when
the dc voltage across the insulation is zero, so it needs a special
cable rather than a circuit mod. Putting the amplifier at the
transducer works even better, usually.

Cheers,

Phil Hobbs