You guys aren't serious talking of high Q structures on FR4 at 2.4GHz , are you? The only way to get a reasonable low loss circuit is to make it small , so that stray due to pcb is minimized. Even then it's not a high Q really. This FR-4 material is as lossy as it gets at 2.4, bordering the unuseable . I fought transmission line VCO's , directional couplers and other transmission line or microstrip stuff and the losses plus variation from board to board, due to laminate tolerances, plus the batch variance add up to just too much sometimes . You need to design the circuit for the board rather than the board for the circuit. It just adds up to too much compromise and expense , when oyu factor in all the extra testing, rework or lower yield, dispersion, extra firmware to deal with parameter vairation, etc . To deal with such large variations some IC's have variable attenuation stages or intermediate gain stages , firmware adjustable . That helps a bit .
best regards, matt tudor -----Original Message----- From: Brian Guralnick <[EMAIL PROTECTED]> To: Protel EDA Discussion List <[EMAIL PROTECTED]> Date: Friday, November 26, 2004 7:58 PM Subject: Re: [PEDA] Dielectric constant >> Do you notice any reduction in Q during trimming or is the effect >> insignificant in your circuits. > >On a purely flat traces without teeth, the Q would go down. In my case, the >teeth have already diminished the Q, sanding them down actually raises the Q >ever so slightly. The key here is to construct the right size teeth so that >the Q remains a constant. Obviously, this type of inductor does not contain >the highest Q possible for a PCB, but, my goal wasn't for the highest Q, >but, a constant one through the trimming process. > >> What do you do about temperature and aging? Are there other parts of the >> circuit that compensate (if necessary) once you have achieved a good >> single point calibration. FR-4, after all, has pretty big temp effects. >> (Not as bad as the sudden phase change at circa 19 deg C of Teflon though, >> I think). > >I would only recommend this type of inductor for room-temperature devices, >or, passive filter designs where heat isn't being focused around the >inductor. However, I am using 5 of them in an FM LMB tuner which does warm >up in the satellite receiver. Frequency ranges from 900 MHz to 2.5 GHz. >The LO frequency is 525 MHz above those. > >> Are there any aging effects? I assume you have to prevent oxidation after >> trimming - conformal coating? Does this affect the set point? > >I coat the area with a tuner stabilant. It's a UV curing epoxy which stops >all oxidization. I'm not sure of the long term aging, however, my current >circuit does have 3 tuning diodes on each precision coil & I've given the >design a +25% added clearance. > >________ >Brian G. > > >----- Original Message ----- >From: "Ian Wilson" <[EMAIL PROTECTED]> >To: "Protel EDA Discussion List" <[EMAIL PROTECTED]> >Sent: Friday, November 26, 2004 7:34 PM >Subject: Re: [PEDA] Dielectric constant > > >> Brian, >> >> In the past I was working on some High-Q structures on PCBs. Edge >> definition was a significant factor. I could get significantly higher Q >> by using a company that plated from a very thin copper layer that was >> plasma deposited on the substrate. The walls were much sharper than on a >> board done by a more traditional means. >> >> Do you notice any reduction in Q during trimming or is the effect >> insignificant in your circuits. >> >> Do the trimming fingers affect Q significantly? >> >> What do you do about temperature and aging? Are there other parts of the >> circuit that compensate (if necessary) once you have achieved a good >> single point calibration. FR-4, after all, has pretty big temp effects. >> (Not as bad as the sudden phase change at circa 19 deg C of Teflon though, >> I think). >> >> Are there any aging effects? I assume you have to prevent oxidation after >> trimming - conformal coating? Does this affect the set point? >> >> I used to do something vaguely similar on shorted stubs, by soldering a >> shim at the right point during test - shorting out more or less than the >> nominal length. Never looked as good as it could though, but it did allow >> accurate trimming in small volume applications. You way would probably be >> better and easier to control. >> >> Sort of reminds me of some papers I read years ago on "Dent Tuning". I >> initially thought it was some sort of technique developed by someone with >> surname of Dent (Arthur Dent maybe :-), until I started to read. It was >> real engineering - tuning large wave guides with a small hammer and >> banging at appropriate places along the line. >> >> Ian Wilson >> >> >> >> On 05:14 AM 27/11/2004, Brian Guralnick said: >>>Ok, Ok, Ok, >>> >>>Imagine one of my inductor generators now make these teeth which zipper >>>between the parallel layers. The points of the teeth have soldermask >>>openings. On the last wrapping of the inductor, there is an additional >>>ground strip, extra thick at 12 mil, with small teeth to match the >>>inductor ring's teeth, this one having solder mask on it's teeth as well >>>as the outer inductor ring's teeth, but, it has no solder mask over it's >>>outside. >>> >>>Very light "Pen eraser" erasing over the middle of a 2.5uh inductor will >>>cut around .1uh-.2uh over 2-3 strokes, my measurement limits is .1uh, >>>using a "Pencil eraser" cuts around 0.01uh per stroke. >>>(Calculated/estimated on a test oscillator circuit). Note that Pen >>>erasers are more abrasive versions of the pencil eraser. They tend to >>>thin out the exposed plating on the edges of the teeth without doing much >>>to the solder mask. I guess there is a lot less friction there. >>> >>>On the outer ground edge, it averages at 0.6pf load. Pencil erasing at >>>the amount of solder mask on the ground strip tunes this capacitor in >>>0.01pf decrerements. The pencil eraser, old fashioned pink type, seems to >>>grip at the solder mask where the pen one slides over it much easier. >>> >>>Depending on operating location, a drop of tuner stablant will secure the >>>values. >>> >>>________ >>>Brian G. >> >> >> >> ____________________________________________________________ >> You are subscribed to the PEDA discussion forum >> >> To Post messages: >> mailto:[EMAIL PROTECTED] >> >> Unsubscribe and Other Options: >> http://techservinc.com/mailman/listinfo/peda_techservinc.com >> >> Browse or Search Old Archives (2001-2004): >> http://www.mail-archive.com/[email protected] >> >> Browse or Search Current Archives (2004-Current): >> http://www.mail-archive.com/[email protected] >> > > > >____________________________________________________________ >You are subscribed to the PEDA discussion forum > >To Post messages: >mailto:[EMAIL PROTECTED] > >Unsubscribe and Other Options: >http://techservinc.com/mailman/listinfo/peda_techservinc.com > >Browse or Search Old Archives (2001-2004): >http://www.mail-archive.com/[email protected] > >Browse or Search Current Archives (2004-Current): >http://www.mail-archive.com/[email protected] > ____________________________________________________________ You are subscribed to the PEDA discussion forum To Post messages: mailto:[EMAIL PROTECTED] Unsubscribe and Other Options: http://techservinc.com/mailman/listinfo/peda_techservinc.com Browse or Search Old Archives (2001-2004): http://www.mail-archive.com/[email protected] Browse or Search Current Archives (2004-Current): http://www.mail-archive.com/[email protected]
