Propagation Times in Asymmetric MicroStrip envirnoment

[mkbs]

Propagation Times

Hi to all,
I have a question about Propagation Times:
I have seen on many articles that the Prop.Time in StripLine is
PT=SQ”Er”/11.8 ns/in
and in MicroStrip environment is:
PT=Fraction*(PT in StripLine)
But, if I have an Asymmetric MicroStrip environment
what I have to do ? What is the Equation ?
(E.g. Top-GND-In1-In2-VCC-Bottom)

Many thanks
MBKS

 

[pisoiu]

Re: Propagation Times

Hi,
Roughly, you can consider a propagation time of 6,7 ps/mm in a FR4 based PCB. Various transmission line configurations will barely change this value, and you must work at extemely high frequencies in order for this parameter to count. Almost all calculations formula for propagation time are approximate, and if you want to know this parameter exactely, you have to measure it. There are differences between FR4 substrates, depending on manufacturer, and Er does vary with frequency, so calculations will not give you precise results.

/pisoiu

 

[mkbs]

Re: Propagation Times

Hi pisoiu,
you consider this propagation time(Pt) (6,7 ps/mm) in all environment ?
4 example if I route a trace throught "Top - In1 - Top" or "Bot - In2 - Bot"
I have to consider this Pt or have I to consider the Pt in every single Layer ?
Tanks

 

[pisoiu]

Re: Propagation Times

Yes, I consider equal propagation times in both cases, even if the propagation time through an inner trace is a little bit higher than on outer ones. The difference is not semnificative. But this issue is dependant on application. For example, if I have to route a high speed ECL differential clock signal from one point to another, the propagation time itself is less important (it is important relative to other signals or clocks), it is more important to minimise the skew between those two traces, and this is done by keeping both traces lenghts equal. Often (and also it is recommended), both differential traces are going on the same layer, so there are no inner/outer propagation differences from this point of view. But as I said in the previous topic, if you really need exact results, you'll have to measure it.

/pisoiu

 

[EEHardware]

Propagation Times

Stripline: Tprop=SQR(Er')/11,Er'=0.475Er+0.67
Microstrip:Tprop=SQRT(Er)/11.8
FYR.

 

[sigint]

Re: Propagation Times

Prop delay in FR4 material is about 150 Ps/in for microstrip and about 170 Ps/in for stripline with a dielectric constant of about 4.5 if you have a microstrip trace 2 inches long it would be 300 Ps, if the trace then transitions to an inner layer and travels 2 more inches the additional prop delay is 340 Ps, total prop delay = 300+340 or 640 Ps.

Hope this helps! Remember the only factor that affects pure prop delay, (disregarding added capacitance), is the dielectric constant.

SiGiNT

pisoiu is absolutely correct - almost any impedance calculator will also tell you prop delay, but due to the variations in materials it will only be approximate - for instance pre-preg comes in several different varieties with different weaves and thickness of glass fibers - this all will affect the delay.