RE: [SI-LIST] : Trace impedance measurement for dual stripline

John Lin - TAO (LinJohn@digital.com)
Thu, 12 Feb 1998 10:28:17 +0800

Dear Bill,

Thanks for your valuable information.

I will try to have a copy of IPC-TP-949.

The coupon I designed is that

G-S-S-P 0 (via connect gnd and pwd planes)
L2
0---------------------------------------------------------------\
/--------------------------------------------------
----------- /
\--------------------------------------------------
-------------0 Loop-round

L3
0---------------------------------------------------------------\
/------------------------------------------------
------------- /
\------------------------------------------------
---------------0

top view

x x
0 0
0
0 L2 L3
L2 L3

Note: X means vias.

As you can see, the vias connecting power and ground planes are only
on one end of coupon.

L2 and L3 traces are parallel and located on the layer 2 and Layer 3
respectively.

The side view will be
----------G
--- L2
--- L3
----------P
It looks different from what you suggested.

Yours is like
-------------G
--- L2
---L3
--------------P

Will the difference make trace impedance waveform like ramp ,not a
straight line.
My TDR outputs a pulse with 35ns rising time.

You can see that the total trace length for L2/L3 is about 8 inch.

Please give any comments to my coupon.
Thanks for your help.

JOHNLIN
CAE Engineer of EDA Department
Digital Equipment Corp. Taiwan Branch
Email: Linjohn@mail.dec.com
TEL: 1-886-3-3900000 ext. 2152

----------
From: Bill Dempsey [SMTP:bdempsey@dnaent.com]
Sent: Thursday, February 12, 1998 1:51 AM
To: John Lin - TAO
Subject: RE: [SI-LIST] : Trace impedance measurement for dual
stripline

Hi John,

The IPC published a very good technical paper on
this exact subject. You can contact them for a copy
of the spec which is paper number IPC-TP-949.
IPC's phone number is (708) 677-2850.
The person who wrote the paper is Dal Currier who
works at Ambitech, (818) 882-9408.

Did you tie the G&P layers at both ends of the coupon?

Did you keep S1 far enough away from S2? S1,S2 are
parallel to each other but positioned so they don't couple.
Does your TDR have a <45ps rise time?
Your trace lengths are extremely short, 5.5 inches. This
translates into ~2ns trip delay. Do you have a fast enough
scope to see this?
You might want to double your trace length...it will depend
solely on your equipment.

This is how I would diagram your coupon:

-------------------------------------------------------------
| |
| G P G P |
| |
| G S1 P G P |
| o--------------------------------------------------- |
| | |
| o--------------------------------------------------- |
| |
| G S2 P G P |
| o--------------------------------------------------- |
| | |
| o--------------------------------------------------- |
| |
| G P G P |
| |
-------------------------------------------------------------
|<-------------------------- 14cm ---------------------->|

Then from a side view it would look something like....
----------------- G
- x S1
- x S2
----------------- P

where "x" is just showing the loop-around portion of S1 or S2.

Your coupon "should" work. I can't see your setup but it
sounds right. Call Dal at Ambitech for assistance; he's
really good.

Also, you might try posting your question at www.sigcon.com.

Regards,
B. Dempsey
bdempsey@dnaent.com

-----Original Message-----
From: John Lin - TAO [SMTP:LinJohn@digital.com]
Sent: Tuesday, February 10, 1998 8:00 PM
To: 'SI_LIST'
Cc: Hw Cheng; Jj Leu; Steve Ting
Subject: [SI-LIST] : Trace impedance measurement for dual stripline

Hi,

Can somebody suggest how to make a coupon for a trace impedance
measurement with TDR?

Currently I use PCB stackup to do trace impedance control. For
measurement purpose, I design a coupon on PCB waste area of a
design. The coupon is 1cm wide and 14cm long PCB board. On the coupon,
there are traces - one trace for one signal layer.

Now, I have a coupon for 4 layer - a dual stripline structure,
G-S-S-P.

My problem is how to connect G and P planes.

I doubt that connecting G and P planes with two vias, which I am used
to do for a TDR's probe, is not right. My last measurement for this
coupon showed the impedance waveform of traces on inner layers like a
ramp - gradually increasing from 69ohms to 87ohms. The expected value
is 90 ohms.

Can somebody explain why the waveform looks like a ramp instead of
straight line. Does it relate to the imperfect connection of G and P
planes.

Best Regards,

JOHNLIN
CAE Engineer of EDA Department
Digital Equipment Corp. Taiwan Branch
Email: Linjohn@mail.dec.com
TEL: 1-886-3-3900000 ext. 2152