A very interesting read... however, I have a good friend... one of the best RF Engineers I've ever met. Former HP to Aglient guy now working for another big name company doing serious rf and lightwave work. He took some time to out to run through various rf connectors on the companies very high-end test equipment. Something he does at/for his job all the time.
A very quick summary of the end results... PL-259 Connectors are not really the UHF horror story people might assume. A sidebar... I didn't see any mention of Epoxy dilectric type PL-259 connectors. If you search the yahoo rfamplifiers group a year or two back... you can find his post in one of the more interesting threads. I don't have the time, nor do I care to debate the subject... but I do agree with my friends testing. Sure there are much better UHF connectors available... but using a good quality PL-250 in the lower UHF Range is really no big deal. cheers, skipp [paste text] > allan crites <[EMAIL PROTECTED]> wrote: > Jeff DePolo, Gary Schafer, Nate Duehr, and Burt Lang, > It is certainly interesting to read the widely diversified lack of authoritative information on the subject of "constant impedance" and other descriptions of the impedance characteristics of connectors used for RF, namely PL259, BNC, TNC, type N, SMA, SMB, and the ubiquitous RCA (shortened) phono connector which has or have been presented in these forgoing discussions. > There are however in my observation, several individuals who by their experience or knowledge, have contributed factual and important descriptions of the impedance characteristics of RF connectors. They are Jeff DePolo, Gary Schafer, Nate Duehr, and Burt Lang. Burt was right on when he said that a PL-259 was a 35 Ohm connector (because of the Bakelite insulator dielectric) and that with air as a dielectric it results in a 50 Ohm characteristic impedance. The only saving grace for "UHF" connectors is in the use of Teflon dielectric in place of the Bakelite, and use in lower frequency applications. > As an engineer employed in the Engineering Dept.of Amphenol-Borg Electronics Corporation in Broadview, IL (Amphenol) during the 1950's and 1960's I was personally involved in the evaluation of many of the RF connector products manufactured by Amphenol and others. > Along with my engineering supervisor appointed by the company, Norbert J. Sladek, I was the co-author of many engineering reports which contained the measurements of RF characteristics of which I have performed. > The most note-able here of which is "Engineering Report on the VSWR of Series 83 UHF Connectors for the Frequency Range 20-500 Mc/s" by N.J. Sladek and A.B. Crites, dtd September 14, 1955. > In it there are many graphs and tables of the measured impedance characteristics and RF signal losses of the PL 259 (Amphenol # 83-1SP) mated to other Series 83 connectors. For those desiring a copy I would suggest requesting a copy of the report done for E.P.A. 340-25-1913 from Amphenol now located in Danbury, Connecticut. It should be noted here that the 83-1SP and mating connectors all used either a Bakelite or Teflon dielectric. (There are "UHF" connectors in the marketplace which use white Nylon or other substances and are lower cost and unsuitable for high frequency use.) > At the conclusion of the project Mr. Sladek made the following rule of thumb regarding the use of "UHF" connectors. "Don't use UHF connectors in the UHF band". > There were also 25 other reports generated during this time, on projects that were related to RF connector development at Amphenol, in which I participated. > The term "constant impedance" used to describe a parameter of an RF connector is to my way of thinking a misnomer as there are no "constant impedances" in the construction of an RF connector, only nominal 50 ohm sections or structures, line sections known as discontinuity capacitances, and higher impedance inductive compensations. All these discontinuities are the result of gaps in the mating of the center conductor and outer conductor, the step up (undercuts in) or down in the dielectric (so called insulator supporting the center conductor) and/or the center conductor, and the difference in center conductor and outer conductor diameters of the junction of the coaxial cable interface to the connector dielectric (sometimes referred to as the bead). > Another factor in the design of RF connectors is the reflection from the faces of the dielectric which is the transition which departs from or to another dielectric section ( air to Teflon and back, Teflon to Polyethylene, etc.) These reflections are more important however, in the GHz frequencies. > For those who are interested, a discussion of the methods used to achieve a low reflection RF connector for use at higher radio frequencies can be found in " Development of 50 Ohm Low Reflection, Captivated Contact TNC Connectors" by A.B. Crites, R.P. Honn, and N.J. Sladek, dtd February 19, 1958. E.P.A. 340-25-2163. > Also see " The Interrelationship Between Dielectrics and The Electrical Properties of Radio Frequency Coaxial Connectors" by N. Sladek, dtd June 1, 1959, for USAF Contract AF33(616)-6243 for the fundamental considerations in the design of RF connectors. > I agree with the comments of the BNC connector being inferior to the type N, and I prefer the use of the type N connector where ever and when ever possible for the reasons given. > The SMB connector was developed by Amphenol for use on RG-58/U and other similar size Teflon coaxial cables as a lower cost, quick connect / disconnect connector. > The SMA was developed by another connector manufacturer as a screw on version. The Motorola, 800 MHz, 100W., Quantar Transmitters uses these SMA connectors. > One connector not mentioned is the type C which was developed as an alternative to the type N, having a Teflon dielectric which encompassed the mated center conductor contacts, leaving no air section, having improved power handling, high voltage breakdown, and VSWR, along with an improved quick connect / disconnect feature. > One last item. I own an HP 8714A N/A and I'm sure I could make measurements of any parameters relating to the impedance and insertion loss of RF connectors. We did not have the equipment in 1955 to do the wondrous measurements available today. > > Allan Crites WA9ZZU > > > > > Gary Schafer <[EMAIL PROTECTED]> wrote: > Hi Allan, > > Coax line impedance is determined mainly by the size of the center conductor and its spacing to the outer shield of the cable and somewhat by the dielectric in the cable. The same is true for a connector. Any time you change the size of that center conductor or the spacing to the shell the impedance changes slightly. In a PL type connector the ratio of center conductor to shield changes from what the ratio of the cable to shield is so that gives an impedance change. > > A connector with a constant impedance may have different sizes of center conductors (center pin) than what the cable size is but that change in size maintains the same ratio of pin size to shield as the ratio of the cable center conductor to shield. That keeps the impedance the same through the connector as what the cable is. > This is why a connector for a 75 ohm cable will be slightly different in size than one for 50 ohms. Note that 75 ohm cable has a smaller diameter center conductor than a 50 ohm cable that has the same outer shield size. > > 73 > Gary K4FMX > > > --------------------------------- > > From: [email protected] [mailto:[EMAIL PROTECTED] On Behalf Of allan crites > Sent: Thursday, March 22, 2007 10:57 PM > To: [email protected] > Subject: RE: [Repeater-Builder] LMR feedline revisited and revised! > > > I sure would like to hear what you all mean by " constant impedance ". > > Allan Crites, WA9ZZU > > Gary Schafer <[EMAIL PROTECTED]> wrote: > > > > > -----Original Message----- > > From: [email protected] [mailto:Repeater- > > [EMAIL PROTECTED] On Behalf Of Ken Arck > > Sent: Thursday, March 22, 2007 7:42 PM > > To: [email protected] > > Subject: Re: [Repeater-Builder] LMR feedline revisited and revised! > > > > At 04:33 PM 3/22/2007, you wrote: > > > > > > > > > > > > > > >doesn't make it right though considering the potentual losses. > > > > <---Psssst...don't tell anyone but a PL259 ain't as horrible (loss > > wise at least) at VHF & UHF as "folklore" would have you believe. The > > real issue is not of loss but rather that of a PL259 not being a > > constant impedance connector. This is where BNC's and N's shine. > > > > Ken > > > > Ken is exactly right! I would venture to say that there are probably few to > none on this list that have any equipment that could measure the difference > in loss between a PL259 and a good N connector. That loss thing is an old > myth. Now as Ken said they are not a constant impedance and you will get a > bump in the impedance with one that can give a mismatch in a fixed tuned > circuit and you can have what is called "mismatch loss" which results from > the circuit getting detuned slightly because of the impedance difference. > There can also be swr losses caused by the impedance mismatch but a direct > loss from the PL259 at VHF and UHF is almost non existent. At least not > measurable. > > If you really want to get picky use only N connectors and not even bnc. A > bnc is not a constant impedance connector either although somewhat better > than a PL259. Yes I know you can plug an N connector into a bnc but the bnc > still is not a constant impedance device. > > Case in point: The Motorola 2600 service monitor uses an N connector in > order to meet the flatness spec over the entire range even though most > people want a bnc for the rf connector on it. So they supply an N to bnc > adaptor with each unit for those that insist on bnc connectors. > > 73 > Gary K4FMX >
