High Frequency Flex minutes -032316

Minutes of March 23, 2016 Meeting

(11:00 AM Central)

 



Project Leader:  Glenn Oliver, DuPont

 

Project Facilitator:    John Davignon, HDP

 

Present on the call:

 


Glenn Oliver, DuPont

Vinh Ha, Ciena

Chun Chang, Ciena

John Davignon, HDP

Brian Butler, Introbotix

Joe Smetana, Nokia-US

Rick Hinkle, Panasonic

 


Next Meeting:  April 6, 2016 @ 11 AM CT

 

Next Meeting Agenda:

 

·         Review the design and panel layout from Ciena.

Action items:

 

·         Farhad will check the design impedances and line widths. (target 50 Ohms) –COMPLETED

·         Glenn will send the CAD data with connector designs to Eric/Vinh, if Eric can get the board files of the connector they can do the FSS modeling and see if a modified anti pad design can help. -COMPLETED

·         Joe S will try to find someone who can do an EMI radiated spectrum (ground planes are also shield planes).

·         Jeff to let the team know if MicroConnex is interesting in building any test product. No decision yet.  Johnd sent email to MicroConnex March 28

Discussions

 

Glenn reviewed his clarifications to the stack up:

 

His design consist of two 12” x 18” sub-panels constructed to go into a full 18” x 24” panel, but can be fabricated as single 12” x 18” panels. For easy of discussion we will call them part 1 & part 2

 

Part 1 is a microstrip test vehicle

Part 1 (microstrip) only utilizes the top layers L1 & L4 – this could be a two layer panel if each part is fabricated separately.

 

Part 2 is the stripline version of the TV

Part 2 (stripline) uses all 4 layers.

 

 

Farhad was able to complete the CST modeling (3D solver), he found some ringing (red numbers in data) that he felt uncomfortable. Glenn could only do rectangles in the Polar software.  This is why he does not have values for the circles in his Polar data.

Vinh said that he has seen a lot of ringing at high frequencies, Glenn said that he expects to find that. The ringing will start at difference frequencies for each cross hatch.  This is what we want to see, we need to understand the design limitations for each crosshatch and frequency interactions.

 

The Molex connector part was discussed, this is the SMA connector Glenn will use.  Vinh probably has the artwork in his files. This discussion was continued after the meeting and has been concluded.

 

Some discussion on why we would not try to get as close to 50 Ohm impedance as we can with the line widths. Glenn said that he does not believe the impedance data so why bother to change the line widths.   

Vinh asked if we would let the fabricator adjust the line width. We will have to finalize this discussion at the next meeting.

 

We do not need microvias since it is so thin, we can use through drilling.

 

Vinh asked if we needed a calibration structure. Glenn said that if we were looking into the 40 Ghz range we would need them. This would be a good thing for the future work, next phase.  Vinh wants to add a calibration model coupon to the edge. He would like to get a panel to take these measurements.  Vinh will look at it and perhaps add some features to make the calibration.  There might be a little room left from the design to squeeze a calibration coupon on the panel. Vinh is OK with sharing this calibration data with us.

 

Brian asked what specific measurements we want to see. We would like to collect Impedance, Loss, Dk.

Impedance waveform from the TDR

S-Parameters 

SPP: Alpha & db/cm length impedance

Effective Dk

Frequency of interest is up to 20 Ghz. Brian B would like to go to 50 Ghz to see some of the differences in crosshatch designs. The team was OK with Brian trying to get some data at that frequency.

 

NOTE: DC resistance is not insignificant with Flex

NOTE: There are 3 different line lengths on the panel for both test measurements, the 100 mm length can be used for both the SPP and the SMA testing.

 

  Our Goal: To understand the effect on Impedance/Loss and Performance due to the crosshatch mesh.

 

Glenn's Hypothesis: These estimates (impedance) are not believable or accurate. Glenn expects them not to match up to the real results. Also there is a limit with both frequency and each cross hatch design at which it will start ringing.  We are aiming to provide some guidance to the designer to tell the tradeoffs between flexibility and performance.

 

The Project Matrix as of March 9:

 

2D and 3D modeling:  M-Flex has this capability and will support the modeling effort.  Possible help from Ciena on modeling.

 

S Parameters: Glenn said that he can do the S parameters testing. Glenn has a solderless connector pattern to use.  

 

SPP:  Brian Butler, Introbotix has volunteered to help evaluate the SPP coupons. Glenn has gotten with Brian and they have a design ready.

 

Analysis: Glenn/Joe/Farhad can do the analysis.

 

DFM: Glenn said that TTM can do the DFM for us and can help build the Gerber data once we supply the data.

  

CAM help with the layout and modify coupon designs: Vinh Ha, Ciena has volunteered to help with any coupon modifications.  Glenn Oliver has volunteered to make sure that this interaction gets completed and the correct coupons are placed on the board. 

 

Crosshatch DESIGN: Glenn has designed a crosshatch matrix using circles and diamonds to vary the % copper/open area on the ground plane (approximately, 25%, 50%, 75% and 100%).  See March 9th Call presentation for openings.

 

Potential Matrix: designed by Glenn Oliver

Design consist of two 12” x 18” sub-panels constructed to go into a full 18” x 24” panel, but can be fabricated as single 12” x 18” panels. For easy of discussion we will call them part 1 & part 2

The coupons have three line lengths. 250 mm/ 100 mm/ 30 mm.

 

Part 1 is a microstrip test vehicle. It only utilizes the top layers L1 & L4 – this could be a two layer panel if each part is fabricated separately.

Part 2 is the stripline version of the TV and uses all 4 layers.

 

·         S parameter launch: The 250 & 100 mm length will be used for these measurements.

·         SPP (Introbotix) The 100 & 30 mm lengths will be used for these measurements –Received specific footprint from Introbotix

·         TDR coupons: Loss and Dk and impedance for various cases / designs -

·         Cross Hatch ground planes:  In progress of being designed, see above

Fabricator:  Team plans on building 5 panels of each image/design for this phase per fabricator. Each panel is 12” x 18” and can be built as an 18” X 24” or as a 12” X 18”.  TTM is still the only verified fabricator.  Waiting on MicroConnex decision to build or not.

 

Next Meeting:  April 6, 2016 @ 11AM CT          

 

                                                        EOM

Attachments