On Apr 6, 2:27 pm, "Ralph Mowery" wrote:
"Ed Nielsen" wrote in message

. ..

Hex-crimp fittings have 6 points where the connector is pinched into the
cable creating small impedance mismatches. Impedance mismatch creates
reflection. If the reflection(s) is(are) severe enough, whatever is at
that(those) frequency (frequencies) may not work. Plus, their return loss
is a fair amount lower than that of compression connectors (~18dB as
compared to 30dB).

I've replaced hex-crimp fittings on DirecTV systems because of hex-crimp
fittings which were causing some channels to not work.

Funny how it works where a bad connector will blank out just a couple of
cable chanels. First time this hapened to me about 20 years ago the cable
guy came out and said it was a bad connection. Almost laughed at him,but he
replaced the connector at the outside of the house and it cleared right up.


Very good info. The interesting thing is since my original post I
went and bought a 500 ft roll of Carol Brand RG-6 QS from Home Depot,
and made some cables with the crimp-on connectors I already had. When
I swapped these new RG-6 QS cables with ones I made a while back with
regular RG-6 with the same crimp-on connectors, I got surprisingly
horrible results. A couple of analog channels don't come in at all
(ch 28 and 56), where they used to come in with a fairly good picture.
Yet other channels on lower frequencies, such as VHF appear to be the
same. I inspected the connectors and they appear to be OK, but I am
guessing there must be an issue with these connectors and the RG-6 QS,
where it's causing the issues that were described above with loss. Is
this possible?? It seems to be affecting various UHF channels (ch.
28, 56, 62).

Thank you very much for the help.

--
Chris

In message .com,
writes
On Apr 6, 2:27 pm, "Ralph Mowery" wrote:
"Ed Nielsen" wrote in message

. ..

Hex-crimp fittings have 6 points where the connector is pinched into the
cable creating small impedance mismatches. Impedance mismatch creates
reflection. If the reflection(s) is(are) severe enough, whatever is at
that(those) frequency (frequencies) may not work. Plus, their return loss
is a fair amount lower than that of compression connectors (~18dB as
compared to 30dB).

I've replaced hex-crimp fittings on DirecTV systems because of hex-crimp
fittings which were causing some channels to not work.

Funny how it works where a bad connector will blank out just a couple of
cable chanels. First time this hapened to me about 20 years ago the cable
guy came out and said it was a bad connection. Almost laughed at him,but he
replaced the connector at the outside of the house and it cleared right up.


Very good info. The interesting thing is since my original post I
went and bought a 500 ft roll of Carol Brand RG-6 QS from Home Depot,
and made some cables with the crimp-on connectors I already had. When
I swapped these new RG-6 QS cables with ones I made a while back with
regular RG-6 with the same crimp-on connectors, I got surprisingly
horrible results. A couple of analog channels don't come in at all
(ch 28 and 56), where they used to come in with a fairly good picture.
Yet other channels on lower frequencies, such as VHF appear to be the
same. I inspected the connectors and they appear to be OK, but I am
guessing there must be an issue with these connectors and the RG-6 QS,
where it's causing the issues that were described above with loss. Is
this possible?? It seems to be affecting various UHF channels (ch.
28, 56, 62).

Thank you very much for the help.

--
Chris


I've worked in the cable TV industry for nearly 44 years, and have NEVER
found that a certain type of male F-connector is, in itself, responsible
for 'horrible results' in terms of RF throughput. Provided that (a) the
shell makes good contact with a few strands of the braid, (b) you
haven't managed to wind a strand of the braid around the inner
conductor, and (c) you haven't left a thin film of the dielectric on the
surface of the inner conductor (yes, I've done all of these!), the
connector WILL appear to work normally. All that suffers is the
screening effectiveness and mechanical reliability.

Regarding whether some male connectors create a significant mismatch due
to crushing of the dielectric, even if this does happen, is not going to
be significant in normal use (up to at least 2GHz) until the outer
conductor is almost touching the inner. You can easily prove this for
yourself by looking at the RF throughput of a piece of coax, while
progressively crushing it with a large pair of pliers.

The thing which causes the most noticeable impedance mismatch is
normally the female part of the connection. Connectors with a
'duck-bill' inner (two parallel fingers) tend to be more capacitive than
those with a 'tulip' (three or four fingers). They are also likely to
make a less reliable ohmic contact.

Don't get me wrong. I do not advocate sloppy practices when fitting
connectors. In the UK, the compression and 'snap-and-seal' types are the
de facto standard. However, I would not hesitate to use crimp or
screw-on types for home use, provided you know how to fit them properly.
And I do.

Ian.
--

At a trade show several years ago, a vendor demonstrated the effects
stapling cables could have on signal transmission. He took a VCR, a ch. 3
modulator and a piece of drop cable and attached the cable to a piece of
wood utilizing a regular staple gun that you buy at any home improvement
store. Used the gun as most people would, and inspection of the cable
showed it to be fine (undamaged). By the seventeenth staple, ch. 3 was
completely gone..

It doesn't take major crushing to create mismatch, and as to whether it is
significant sort of depends on what happens to be trying to get through at
the particular point where the mismatch occurs.


At http://www.cencom94.com/gpage.html9.html, there is a picture of a sweep
trace of some cable with hex-crimp connectors (gotta scroll down a little).
--

CIAO!

Ed N.


"Ian Jackson" wrote in message
...
Regarding whether some male connectors create a significant mismatch due
to crushing of the dielectric, even if this does happen, is not going to
be significant in normal use (up to at least 2GHz) until the outer
conductor is almost touching the inner. You can easily prove this for
yourself by looking at the RF throughput of a piece of coax, while
progressively crushing it with a large pair of pliers.

Ian.
--

Ed Nielsen wrote:
At a trade show several years ago, a vendor demonstrated the effects
stapling cables could have on signal transmission. He took a VCR, a ch.
3 modulator and a piece of drop cable and attached the cable to a piece
of wood utilizing a regular staple gun that you buy at any home
improvement store. Used the gun as most people would, and inspection of
the cable showed it to be fine (undamaged). By the seventeenth staple,
ch. 3 was completely gone..

It doesn't take major crushing to create mismatch, and as to whether it
is significant sort of depends on what happens to be trying to get
through at the particular point where the mismatch occurs.

A periodically repeating mismatch, such as the one produced by the
staple demonstration, can cause extreme effects as the demonstration
showed. This is a very much worse case than a single mismatch.

Roy Lewallen, W7EL

In message , Roy Lewallen
writes
Ed Nielsen wrote:
At a trade show several years ago, a vendor demonstrated the effects
stapling cables could have on signal transmission. He took a VCR, a
ch. 3 modulator and a piece of drop cable and attached the cable to a
piece of wood utilizing a regular staple gun that you buy at any home
improvement store. Used the gun as most people would, and inspection
of the cable showed it to be fine (undamaged). By the seventeenth
staple, ch. 3 was completely gone..
It doesn't take major crushing to create mismatch, and as to whether
it is significant sort of depends on what happens to be trying to get
through at the particular point where the mismatch occurs.

A periodically repeating mismatch, such as the one produced by the
staple demonstration, can cause extreme effects as the demonstration
showed. This is a very much worse case than a single mismatch.

Roy Lewallen, W7EL

Stapling - or any other small mismatch repeated at regular intervals -
can indeed produce a severe structural mismatch (with the associated
suckout) at frequencies where the intervals are one wavelength, and
multiples thereof. However, the presence of a one-off connector where
the match is distinctly questionable will usually go completely
un-noticed (except to the most discerning of engineers). I still
maintain that any problem will be because there is no proper continuity
through the connector, or possibly an inner-to-outer short (maybe
partial). I must admit, I haven't used an RF TDR in earnest for over 20
years. I have found that, if it buzzes out OK at DC, it will generally
be OK at RF.
Ian.
--

With all due respect, that's a rather frightening position to maintain.
I've had to replace both .750 and .500 because of dings. Complaints were
that certain channels were out. A couple of years ago I had to replace a 4
foot piece of RG 6 inside a wall (splitter to outlet) that had the
attenuation of a 100 foot cable. 950MHz to 1450MHz worked (though not near
as well as it should have), but the rest of the bandwidth up to 2200MHz
didn't work at all. Electricians had greatly exceeded the minimum bend
radius when they made up the outlet.

DC does not necessarily mean that RF will work, nor does RF necessarily mean
that DC will work.

--

CIAO!

Ed N.


"Ian Jackson" wrote in message
...
years. I have found that, if it buzzes out OK at DC, it will generally be
OK at RF.
Ian.
--

In message , Ed Nielsen
writes

"Ian Jackson" wrote in
message ...
years. I have found that, if it buzzes out OK at DC, it will
generally be
OK at RF.
Ian.

With all due respect, that's a rather frightening position to maintain.
I've had to replace both .750 and .500 because of dings. Complaints
were that certain channels were out. A couple of years ago I had to
replace a 4 foot piece of RG 6 inside a wall (splitter to outlet) that
had the attenuation of a 100 foot cable. 950MHz to 1450MHz worked
(though not near as well as it should have), but the rest of the
bandwidth up to 2200MHz didn't work at all. Electricians had greatly
exceeded the minimum bend radius when they made up the outlet.

DC does not necessarily mean that RF will work, nor does RF necessarily
mean that DC will work.


Ed, I'm sure what you say is true. However, one-off moderate mismatches
(or even repeated moderate mismatches which are buffered by
attenuation), such as might be caused by the use of poor quality
connectors, should not cause a failure of service unless the service
would have been marginal if the connectors were good.

As for bends in cable, if you can, tie a loose knot in a piece of coax,
run a wideband sweep through it, and observe the output. Now pull the
knot tight. Let me know when the output starts to be affected.

Let me say again, I'm not advocating sloppy practices and poor
workmanship. Manufacturers' specs and industry standards should always
be adhered to. This avoids endless truck-rolls (God - I hate that
Americanism!) to the 1% of customers where some bright spark has cut one
corner too many. But, especially in the amateur world, we should maybe
not worry too much about using things which industry has rejected, often
for reasons which have absolutely no impact on what we are trying to do.
So I'm NOT throwing out my stash of crimp and screw-on F-connectors!

Ian.
--

"Ed Nielsen" writes:

It doesn't take major crushing to create mismatch, and as to whether it is
significant sort of depends on what happens to be trying to get through at
the particular point where the mismatch occurs.

At http://www.cencom94.com/gpage.html9.html, there is a picture of a sweep
trace of some cable with hex-crimp connectors (gotta scroll down a little).

That particular sweep seems to be from a highrise building which had hex
crimp connectors every 25 feet all the way up the building. A single
hex crimp, or a bunch of hex crimps spaced irregularly, would have far
less effect. So whether this matters partly depends on whether you're
wiring an apartment building or just a house.

Dave