to create and rate content, and to follow, bookmark, and share content with other members.
Question asked by
on Jun 27, 2007
on Aug 1, 2007 by RANDY_RHEA
Show 0 Likes
We over mold all of our circuit boards. I tried to add a substrate to simulate the passive circuit including the properties of the over mold. I am having difficulty adding the substrate. Is there a tutorial available?
No one else has this question
Mark as assumed answered
This content has been marked as final.
Show 2 comments
(Required, will not be published)
Jul 28, 2007 8:02 AM
I don't think we understand the question. Can you ask this another way? You have a passive circuit and an overmold and you want to.... ?
I'm not certain what the effects of an overmold would be... which is part of the problem.
Show 0 Likes
Aug 1, 2007 5:35 AM
I'll try to help. You weren't very specific, but I'll make a few assumptions and offer some ideas.
If you are doing an EMPOWER electromagnetic and circuit theory co-simulation of your circuit, the following remarks may not be appropriate.
If your circuits are purely distributed filters, couplers, etc., with no or limited passive components, then you may easily add an additional dielectric layer to your EMPOWER simulation. You will only need to know the dielectric properties and thickness of the overmold. Layers are covered in the EMPOWER section of the manual and in examples.
I'll assume you are just doing a circuit theory simulation.
You'll notice for a circuit simulation you don't need to enter substrate parameters for lumped components. That's because the built in models consider only the resistance and reactance of components with unloaded Q for some of the parts. The PWB metal and substrate will have the largest parasitic impact on circuit response. Next in importance is the capacitance of components to the groundplane through the PWB substrate. Third down the list of important effects is the overmold. It will introduce only a slight amount of shunt capacitance. There isn't much point in worrying about the overmold until the metal pattern and component/substrate characteristics are modeled.
The metal pattern is best handled using a co-simulation with EMPOWER. It automatically removes the components, does an EM simulation of the metal (capacitance and inductance) and adds the components back. GENESYS was the first product in the industry to ever do this. The effect of the metal can be profound on circuits above 100 MHz and even lower in frequency for bandpass filters and other high Q circuits. With current computer technology, simulation gets very time consuming for large circuits with lots of metal and components. However, first stuying smaller subcircuits, gaining experience, and working up to moderate sized circuits can be effective.
The component/substrate interface can be modeled using Modelithics substrate dependant component models
. It adds cost to the tools and complexity to the simulation environment, but it improves accuracy when that is required. Even these models don't consider overmolding.
An alternative is to create your own component models from measured data on your specific substrate. This could even include overmolding. GENESYS has some really great tools for creating user models. Developing the models is timeconsuming, but you need to do it only once for each substrate you use, and you'll learn a great deal about how parasitics impact your circuit. That may be more valuable than the models themselves.
I know nothing about your specific circuit specifications, so all I can say is that overmolding may have absolutely no degrading effect to possibly totally destroying the design. As a sanity check, I would build and measure a circuit, then add the overmolding, remeasure, and then decide how much effort needs to be put into overmold modeling.
Clear skies and high Q
Show 0 Likes
Retrieving data ...
34980A 34925A Analog Bus Armature Relays Programming
34925A MUX bank switching avoidance during 60 channel scan
HP 8591A problems
HP 8590A RES BW NOISE FAIL 100000 on Confidence test - Troubleshooting?
HP 8591A spectrum analyzer (little problem)