Matt, Actually that is not necessarily true, and a lot depends on your materials, and much more importantly, your frequencies, as well as your specific applications.
I remember my first exposure to these kinds of "tweaking" was when I was at Hughes Space and Communications, in El Segundo CA, back in 1978, when I was a Designer working on the S-Band Receiver for the GOES / GMS Weather Satellite, where I was doing a lot of microstrip RF stuff, using rubylith at 10X to make 1/2 inch by 1 inch, 1 inch by 1 inch, and 1 inch by 1 and 1/2 inch (or even up to 2 inch), etched alumina substrates on kovar mounting plates, 99 percent of which I didn't understand at the time, but where I just did what the Microwave RF Engineers told me to do. One of the little "tweaks" that I "laid out" for them, that I didn't understand at the time, and which I can't remember enough about now to offer any intelligent comment on, but which I specifically asked them about at the time, was a series of broken little 50 ohm trace elements, which were almost square when they began, but of progressively slightly different smaller sizes, with gaps between them, all of which looked like a little dashed line going off to nowhere. (I actually think that I might even have one or two "blueprints" of some of those rubyliths that I did for them, possibly even with such a "stub" somewhere around here in my "unsolved paper puzzle", some of which was by hand, and some of which was done using a big "Gerber Scientific" X-Y Hand Crank "Dial It In To Less Than A Thousandth Of An Inch In Both The X and Y Axis Gizmo" on a very large Light Table.) Anyway, I seem to remember that I was told that they were little "tuning stubs", consisting of various specific inductance's and a variable amount of capacitance, all of which was achieved by placing a little "chip" of bare alumina substrate material, somtimes of different sizes, at certain locations on top of the broken up little trace elements of the "stub", where they would slide these little "chips" around, moving them back and forth on this broken up little piece of microstrip to nowhere, such that once things were "tweaked" just right, the little alumna "chips" were glued in place, ready for space flight. Almost sounds like this might have fallen into the realm of your last paragraph, but possibly with some differences. Once again, I can't really even remember enough about the circuits to even stick my neck out much further, other than to say that that's the way it was done at Hughes on GOES / GMS. Then again, there are products in certain realms where each "printed circuit" is a prototype, irrespective of the quantity, in that some times every single product that goes out the door has to have certain circuits individually compensated. "Tweaks" and "fixes" have been around for years in certain types of equipment, and are certainly legitimate "non-component" parts, or methods of varying the electrical characteristics of an otherwise "fixed characteristic component" part of a circuit in some industries. Now days they have even developed quite a bit in sophistication, such that you might "tweak" a circuit by automated equipment such as "laser trimming" a resistor in final test, even in high volume production. All depends on what your building and what your needs are, and where you're trying to shave corners, sometimes literally. JaMi ----- Original Message ----- From: "matt" <[EMAIL PROTECTED]> To: "Protel EDA Discussion List" <[EMAIL PROTECTED]> Sent: Friday, November 26, 2004 2:18 PM Subject: Re: [PEDA] Dielectric constant > These techniques might work reasonably well for a quick and dirty lab > prototype but they're definitely not for production. > If you really want to design some tuning elements into your product, you > should use appropriate methods. You can have variable impedance matching, > variable transmission lines, etc without paying for expensive parts. You may > need to place a short in a variable position , a zero ohm resistor for > example. A lot more repeatable than the eraser thingy . > You can really incorporate tuning transmission line based elements which > vary based on length and measure the boards during assembly, sort them in > several classes and then as a secondary operation place the variable short > in different positions corresponding to each class . > > Best is to design with circuits where pcb dielectric changes tend to cancel > , for example a place where a higher Dk increases a stray capacitance would > lower an oscillator resonant frequency but the same Dk increase would also > lower some tuned tank center frequency , eliminating the need to adjust one > for the other . Or even better yet, minimize the Dk effect on your circuit > from the design stage . Use a proper RF simulation package and see where Dk > matters and where it doesn't . > > best regards, > matt tudor > > -----Original Message----- > From: JaMi Smith <[EMAIL PROTECTED]> > To: Protel EDA Discussion List <[EMAIL PROTECTED]> > Date: Friday, November 26, 2004 4:49 PM > Subject: Re: [PEDA] Dielectric constant > > > >Brian, > > > >I know that you do not want to remove a protective plate such as gold, > which will > >prevent oxidation, such as what you will get with bare silver or copper, > which is > >liable to change over time due to the effects of the oxidation, but you can > do fine > >"tweaking" with solder plate, so long as you do not totally remove it, > which it > >sounds like you are doing. > > > >Another way to "tweak" a circuit would be to thin out the amount of "solder > mask" > >which will change the "dielectric constant" of a trace, and hence the > "velocity > >factor", although that would be a change of a different kind than the > "inductance" > >you are dealing with, and probably a little harder to get optimal results > with, and > >where you would probably use something like acetone or some other solvent > as opposed > >to an eraser. > > > >Then again, you could possibly actually get down to bare copper if you > really wanted > >to change inductance by thinning things out, and then give the area a very > thin > >protective spray to prevent oxidation, without affecting the net change in > >inductance too much, if at all. > > > >Respecting cleaning the "card edge connector fingers", which is usually the > first > >thing any Technician will look at: Digital Equipment (DEC) once put out a > tech > >bulletin back in the early 70's stating that anyone caught using an > "eraser" on card > >edge connectors would be fired on the spot, once they found out that that > was how > >their tech support people were "fixing" their malfunctioning equipment in > the field, > >but they additionally followed it up with much better designs on their > connectors at > >the same time, along with thicker plating on their "fingers", which lead to > yet > >another problem in some areas, due to the extra "resistance" of the thicker > gold > >plating. That's why the Military was also freaking out about the amount of > gold on > >connector contacts and fingers back in the early 70's, but for just the > opposite > >reason, the Military wouldn't stand for the extra resistance, or the gold > >embrittlement when it came to soldering. > > > >JaMi > > > >* * * * * * > > > >----- Original Message ----- > >From: "Brian Guralnick" <[EMAIL PROTECTED]> > >To: "Protel EDA Discussion List" <[EMAIL PROTECTED]> > >Sent: Friday, November 26, 2004 1:05 PM > >Subject: Re: [PEDA] Dielectric constant > > > > > >> Read my full reply/description. > >> > >> I do know that if you have gold, or silver plating, you will be stripping > a > >> few thousand atoms at a time from the vertical thickness, since, you see > >> them darkening up the eraser. Also, it seems to get rid of any oxidizing > >> layers. It's especially great at re-finishing gold-thinger edge card > PCBs. > >> > >> ________ > >> Brian G. > >> > >> > >> ----- Original Message ----- > >> From: "JaMi Smith" <[EMAIL PROTECTED]> > >> To: "Protel EDA Discussion List" <[EMAIL PROTECTED]> > >> Sent: Friday, November 26, 2004 3:51 PM > >> Subject: Re: [PEDA] Dielectric constant > >> > >> > >> > Brad, > >> > > >> > By using an eraser, you can do a very fine trim of the thickness of the > >> > trace, and > >> > hence a very fine "tweak" of the electrical characteristics, especially > if > >> > you have > >> > a solder plate / HASL. > >> > > >> > JaMi > >> > > >> > > >> > >> > >> > >> ____________________________________________________________ > >> 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] > ____________________________________________________________ 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]
