Talk:Microstrip

Inhomogeneity?

Really? No-one knows about heterogeneity? --203.94.174.142 (talk) 05:25, 27 September 2010 (UTC)

Resistivity Relevance

I dont think conductor resistivity has that much effect of the propagation down a microstrip. Not as much as its inductance for instance. Does anyone want to discuss this statement?? Light current 20:59, 24 August 2005 (UTC)

Agreed - but the article only says that the line has "specific electrical properties". If it's talking about the loss per unit length (rather than the characteristic impedance), then this does depend on the resistivity. Perhaps stating that something 'has properties', without saying what they are, is a bit too vague?catslash 14:03, 10 August 2006 (UTC)

Differential Traces

I appreciate very much having post the formula for microstrip. The PCB prototype shop where I work, now uses it in a tool that generates the layer-stack-up document that eventually goest to the manufacturing floor. If there is any interest we may furnish de VB6 source code version of the formula.

I request your advice for directions for where to find the formulas for Surface-Differential-Trace Impedance and Offset-Differential-Trace (internal)

Thanks again.
--Legarcia.sp 17:29, 8 November 2006 (UTC)

Sorry: I don't fully understand; are the differential traces next to each other on the same surface of the substrate (co-planar strips)? and with a ground plane (coupled microstrips)? or are they facing each other on opposite surfaces of the substrate? And the Offset-Differential-Trace (internal); is it like this [1]? --catslash 15:22, 9 November 2006 (UTC)

Bends

I disagree that a "large" portion of the signal would be reflected at a 90 degree bend.

"An abrupt 90° bend in a microstrip will cause a large portion of the signal on the strip to be reflected back towards it source, with only part of the signal transmitted on around the bend."

See the post here for an example of the effect of a bend:

http://circuitcalculator.com/wordpress/2007/04/06/effects-of-corners-in-pcb-traces/

OK - I will change it to "significant". I've posted my observations regarding the above link on that site. --catslash 18:12, 8 April 2007 (UTC)

I thought that the idea of reflections from 90 degree bends had been thoroughly debunked. Here is one of many references available on this...

http://www.pcbdesign007.com/pages/zone.cgi?a=55843

"In this presentation, an 8 mil wide 50 ohm microstrip is analyzed, with and without a 90-degree bend. Even at 25 GHz, the reflected signal is less than -0.25 dB, or less than 5% reflected back. This is a small, but easily simulated discontinuity."

I corrected a typo in the quote, -25 dB to the correct -0.25 dB which matches the 5% figure given.

If you say "significant", I would think this should be explained in more detail considering that it is clearly *not* significant in most situations. Gnuarm (talk) 19:35, 11 February 2010 (UTC)

It's true that the reflection from a single unmitred bend can be small. For a device that is supposed to transmit (as opposed to reject) a signal, the common requirement is that the return (reflection) be -30dB or better (I haven't got a reference for this), i.e. not more than 1/1000 of the power is reflected back. A return of -25dB is 0.3% not 5% (the amplitude of the reflection is 5.6% of the incident wave). A return of -0.25dB would mean that 94% of the incident power is reflected back (the amplitude of the reflected wave is 97% that of the incident wave). I think these digital PCB people are less fussy than the rest of us; they don't care unless a 0 gets changed into a 1 or vice-versa. --catslash (talk) 13:36, 12 February 2010 (UTC)

The article already mentions sever cutting of the mitered corner. I agree that a mitered corner will corner theoretically gives better performance, but in practice given a certain process tolerance, it is usually worse. Maybe this could be added to the article? — Preceding unsigned comment added by 86.81.172.19 (talk) 19:48, 29 January 2013 (UTC)

Yes that could be added to the article. However since mitres generally are used (see File:Antenne patch 2.4 GHz.JPG, File:Impanp.jpg, File:Stripline.jpg), a statement that they are not (or not always?) a good idea ought to cite a reliable source to back up the claim. --catslash (talk) 01:05, 30 January 2013 (UTC)

Oops, while editing I had a line in there mentioning FR4 multilayer PCB's, I apparently removed this line. In eg the multilayers I use, typically stripline linewidths are only say 12 mil wide (stripline between two .4 mm layers, with a .4 mm microstrip on top and a .4 mm DC distribution on the bottom). Then with a descent miter it can get as low as 4 mil, giving an etching tolerance of say 2 mils the effects can become significant (Ofcourse the lossiness of FR4 helps hiding some of the problems). I have to look for some proper reference, I remember seeing some IEEE-MTT paper showing the effects of etching tolerance having some nice plots. Also just from memory I believe Harlan Howe mentions this in his book Stripline circuit design, I will verify both the book and the paper. I do agree that mitered bends are used a lot especially on thicker substrates, I also use them very often. (I used to design circuits up to 120 GHz, lately only in the low couple of GHz range). — Preceding unsigned comment added by 86.81.172.19 (talk) 01:23, 1 February 2013 (UTC)

PCB Technology

Fewer words is surely more, but I reckon that this edit loses the key point. The point is that a device can be made in microstrip using fabrication techniques which are cheaper and more widely available than those needed to make the same device in waveguide plumbing. Also, waveguide/microstrip technology is not the same thing as a length of waveguide/microstrip, but is rather all the things that can be achieved using waveguide/microstrip - filters, couplers, transformers etc. --catslash (talk) 20:49, 25 September 2014 (UTC)

Perhaps the article needs to explain more clearly why the possibility of fabrication using PCB techniques is important? --catslash (talk) 21:00, 25 September 2014 (UTC)

Not sure what the objection to the word technology is - how about techniques or methods? As in ...can be made using the same methods used to make PCBs or can be made using PCB fabrication methods --catslash (talk) 23:01, 25 September 2014 (UTC)

I agree that the word "technology" is meaningful here. The important point is the manufacturing platform being used, not that it is printed circuit per se. One can mount a variety of things on pcb made in a variety of materials and technology. That is not the same as actually making it in pcb technology. SpinningSpark 23:47, 25 September 2014 (UTC)

Wtshymanski regularly goes off on one. His current hobby horse is to delete the word 'technology' wherever it appears in Wikipedia. 85.255.232.0 (talk) 12:35, 26 September 2014 (UTC)

And thank you, too. I object to the routine use of the word "technology" to pad out our overlong and vague articles. --Wtshymanski (talk) 15:59, 26 September 2014 (UTC)