Path: ...!eternal-september.org!feeder3.eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail From: Jeroen Belleman Newsgroups: sci.electronics.design Subject: Re: Dressing RG6 Date: Sat, 18 May 2024 22:49:32 +0200 Organization: A noiseless patient Spider Lines: 137 Message-ID: References:

<20240514b@crcomp.net> <66h74j1vfmbjvvl98jk1k017pimtinv2l5@4ax.com>

<20240514d@crcomp.net>

MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 8bit Injection-Date: Sat, 18 May 2024 22:47:34 +0200 (CEST) Injection-Info: dont-email.me; posting-host="9806cee71c9826ddf37f18fe6497ab82"; logging-data="3114696"; mail-complaints-to="abuse@eternal-september.org"; posting-account="U2FsdGVkX1/KxpEuzCbnojUD+GKeObiw" User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101 Thunderbird/102.13.0 Cancel-Lock: sha1:GondGrBuKE9n/bPX2nbuZD0aZcY= Content-Language: en-US In-Reply-To: Bytes: 6805 On 5/18/24 17:17, Phil Hobbs wrote: > Jeroen Belleman wrote: >> On 5/16/24 17:41, Phil Hobbs wrote: >>> On 2024-05-15 17:25, Jeroen Belleman wrote: >>>> On 5/15/24 16:27, John Larkin wrote: >>>>> On Wed, 15 May 2024 11:03:22 +0200, Jeroen Belleman >>>>> wrote: >>>>> >>>>>> On 5/15/24 01:33, Don wrote: >>>>>>> Jeroen Belleman wrote: >>>>>>>> Phil Hobbs wrote: >>>>>>>>> John Larkin wrote: >>>>>>>>>> Don wrote: >>>>>>> >>>>>>> >>>>>>> >>>>>>>>>>> The parasitic capacitance created between coax and its metal >>>>>>>>>>> armor can >>>>>>>>>>> open a Pandora's box of potential problems. >>>>>>>>>> >>>>>>>>>> Capacitance between the coax outer and the copper pipe? Proper coax >>>>>>>>>> shouldn't have any external field. >>>>>>>>> >>>>>>>>> If the whole system is really coaxial, that’s true. Leaky >>>>>>>>> shields, ground >>>>>>>>> loops, and so on, will modify that. >>>>>>>>> >>>>>>>>> Depending on the application, you may or may not care. >>>>>>>>> If the whole system is really coaxial, that’s true. Leaky >>>>>>>>> shields, ground >>>>>>>>> loops, and so on, will modify that. >>>>>>>>> >>>>>>>>> Depending on the application, you may or may not care. >>>>>>>> >>>>>>>> I've been putting coax inside copper tubes or braids to measure >>>>>>>> and/or reduce the transfer impedance (leakage). I did that to >>>>>>>> measure small signals in a particle accelerator, which typically >>>>>>>> has kicker magnets and RF cavities with kA currents and kV >>>>>>>> voltages nearby. >>>>>>>> >>>>>>>> A colleague developed a special low transfer impedance coax >>>>>>>> cable for this sort of application. It had two screens with >>>>>>>> intermediate magnetic shielding. It was unpleasant to work >>>>>>>> with, because part of the magnetic shielding was a steel >>>>>>>> spiral foil tape that was razor sharp. But it worked really >>>>>>>> well. >>>>>>> >>>>>>> Empirical observation always trumps theory for me. Did you ground [1] >>>>>>> the copper tubes or braids? >>>>>> >>>>>> Both ends were connected to the connector shields. The point of >>>>>> the exercise was to reduce transfer impedance, which at low >>>>>> frequency (<1MHz) is simply proportional to screen resistance. >>>>>> >>>>>> Jeroen Belleman >>>>> >>>>> Two parallel coaxes can make an attenuator. >>>>> >>>>> What was the coupled frequency response like? >>>>> >>>> Ah sorry, this message didn't seem to get sent... >>>> >>>> At low frequency, the transfer ratio was simply the ratio >>>> of screen resistance over characteristic impedance. At medium >>>> frequencies, a few octaves roughly around 1MHz, there was a dip, >>>> and above that a steady rise of about 10dB/decade. >>>> >>>> Not all cables behaved the same. RG58 is poorly screened and >>>> doesn't have the dip. UT141 had a very deep dip. >>>> >>>> Details at >>>> . >>>> >>>> Jeroen Belleman >>> >>> Very interesting results, Jeroen. Thanks for posting them. >>> >>> Is the MF resonance due to the inductive and capacitive coupling >>> cancelling each other? (They're 180 degrees out of phase, of course.) >>> >>> The frequency is way too low to be a transmission line effect in a 1-m >>> length. >>> >>> Cheers >>> >>> Phil Hobbs >>> >> >> The original data came from an HP3577 and I recorded only the >> magnitude. Since this looks like a resonance, that's also what >> I'd expect. >> >> I can't easily go back and look again. I did this in 2009, and >> I'm now retired. At the time, I was trying to make a choice for >> cables connecting beam trajectory pick-ups in the CERN PSB to >> their pre-amplifiers. >> >> I suppose -but did not verify- that the dip is a resonance of >> the outer inductance with a parasitic capacitance of my setup, >> with the screen resistance as the damping element. I can't quite >> make it fit that model though. The screen resistance doesn't >> differ enough between, for example, UT141 and RG58 to explain a >> deep resonance for the former, and its total absence for the >> latter. >> >> Jeroen Belleman >> > > Plus you had some pretty frou-frou RG58 there, with foil and two braids. > > The normal stuff is one tinned-copper braid with about 80% coverage. You > can probably make a directional coupler with a pair of patch cords and some > heat shrink. (I should try that.) > > Cheers > > Phil Hobbs This was Draka's version of RG58, I think it was made to CERN's specs, fire retardant and rad-hard. From an RF standpoint, it wasn't so fancy and it didn't fit very well in standard RG58- compatible connectors. Speaking of directional couplers, I made one with bits of UT85. It was inspired by a 1979 publication by Udo Barabas in IEEE Transactions on Microwave Theory and Techniques. It worked surprisingly well, with a measured transmission bandwidth of 40kHz to 9GHz, although the directivity deteriorated beyond a GHz or so. It was a rough demo of the concept. I'm sure it's possible to do much better. It was my suggestion for a GHz bandwidth beam position signal processing front-end in the LHC. I wrote it up in . Jeroen Belleman