Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

Flip chip bounding package impedance

Status
Not open for further replies.
S

shanmei

Advanced Member level 1
Joined
Jul 26, 2006
Messages
430
Helped
8
Reputation
16
Reaction score
8
Trophy points
1,298
Location
USA
Activity points
4,496
The normal flip chip bounding character impedance is 50ohm.

Is there any way to reduce the character impedance?

It seems that the wider the bounding the less character impedance is. Is there any material that has less area to achieve the low character impedance of the bounding? Thanks.
 

Many applications -desire- a 50-ohm trace impedance but
you can't say this is "normal" (as in, intrinsic to the package
and its bumps / balls / columns). You have to design the
package traces and features to achieve that goal, and a
goal of a different Zo is as "engineerable". But somebody
will have to be that engineer, and probably needs a good
depth in EM finite element modeling of packages.
 
The usual way to achieve really low characteristic impedance, e.g. for RF power devices, is to have multiple bond wired in parallel.
 
I am wondering whether there's confusion between the
shunt impedance as a transmission-line for signals, and
the series pin impedance as you'd like it to be for power
and ground. OP?
 

dick_freebird said:
I am wondering whether there's confusion between the
shunt impedance as a transmission-line for signals, and
the series pin impedance as you'd like it to be for power
and ground. OP?
Click to expand...

I want to change the character impedance.
 

shanmei said:
I want to change the character impedance.
Click to expand...

What is your transmission medium you are attempting to alter? Stripline, ustrip, via, CPW, etc. Numerous variables impact impedance of a transmission line. Trace width, Dk of laminate, Height above and below reference planes (assuming stripline). Any ECAD tool should have a TLINE editor function, of which, you can enter your specific transmission line parameters to calculate the effective impedance.

Typically there will be a minimum Return Loss value specified (let's say 15 dB for arguments sake) this could be verified with an S-parameter extraction with a ECAD tool such as HFSS or PowerSI.
 
FvM said:
The usual way to achieve really low characteristic impedance, e.g. for RF power devices, is to have multiple bond wired in parallel.
Click to expand...

The OP appears to be referring to FlipChip package, of which, there are no bond wires.
 
  • Like
Reactions: FvM

    FvM

    FvM

    Points: 2
    Helpful Answer Positive Rating
The OP appears to be referring to FlipChip package, of which, there are no bond wires.
Click to expand...
Right. In this case translate "multiple bond wires" to "multiple balls".
 

FvM said:
Right. In this case translate "multiple bond wires" to "multiple balls".
Click to expand...

Multiple balls and or wires would not effect the transmission line impedance he is referring to, per se. Yes, you could have a tightly coupled differential pair with VSS/GND guard bands, but this is not the impedance he is referring to. He is referring to single-ended (one transmission line) 50 ohm impedance. In your example, increasing number of balls, bumps, and or bond wires would be ideal for reducing L and improving Power Integrity. As for each N you receive a 1/N reduction in L. The would be helpful for reducing voltage ripple from di/dt, and or reducing DC IR drop.
 
Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top