Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronic 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

How to model Pad, Bond-wire and channel behavior in design

circuitking

Full Member level 5
Joined
Jan 8, 2018
Messages
255
Helped
1
Reputation
2
Reaction score
1
Trophy points
18
Activity points
2,214
Hello,

I have an ASIC that is sending data to FPGA.

Between ASIC and FPGA, I would have LVDS (including Pre-emphasis), Pad, bond-wire, and channel. do all kinds of FPGA's LVDS pins have input impedance as 100 ohm?

Now, I want to estimate the behavior of the Pad, bond-wire, and channel and include them in the LVDS design.

Anyone has inputs on this?. Thanks
 

dick_freebird

Advanced Member level 5
Joined
Mar 4, 2008
Messages
7,140
Helped
2,092
Reputation
4,188
Reaction score
1,939
Trophy points
1,393
Location
USA
Activity points
57,288
1nH/mm of bond wire. Coupling, you'd need
a good EM simulation package. You can get
bond wire resistance from wire vendors or
third party assembly houses' web pages,
need to know wire diameter and material
(or whether wire-bonded at all; bumps /
pillars are a whole 'nother thing).

Bond pad you could estimate as areal C of
Met1 over field, dimensions from layout,
thicknesses from PDK modeling / device
construction docs (though what's provided
or even allowed to be seen, varies by foundry).

LVDS pins will have 100 ohms between them
but you can't assume this is all. There will be
ESD protection, which presents C and at some
point, conduction to the rails.

You might cheat and look for other LVDS ASIC
pads' IBIS models etc.
 

dick_freebird

Advanced Member level 5
Joined
Mar 4, 2008
Messages
7,140
Helped
2,092
Reputation
4,188
Reaction score
1,939
Trophy points
1,393
Location
USA
Activity points
57,288
Pads will not be over active area (unless a specific
enablement of circuit-under-pad is made in the
PDK). You ought to see the full un-cut field oxide
thickness between lowest metal of the pad-stack,
and the silicon surface.

There is probably, somewhere in the PDK RCX "stuff",
area and periphery capacitance numbers for Met1
not-over-active.
 

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Top