The meaning of Impedance Mismatch

Jun 15, 2004 code patterns suggest edit

You’ve probably heard the term Impedance Mismatch thrown around when discussing object relational mapping. I’m sure it comes up every morning at the water cooler. Maybe you’ve even thrown it around yoursef a few times. Do you know what the term means?

Object relational mapping refers to the process of mapping your relational data model to your object model. Object Spaces is a highly publicized framework for doing just that. The mismatch I refer to is a result of the differences in structure between a normalized relational database and a typical object oriented class hierarchy. One might say Databases are from Mars and Objects are from Venus. Databases do not map naturally to object models. It’s alot like trying to push the north poles of two magnets together.

Interestingly enough, the term “Impedance”, now bandied about in software engineering circles, is borrowed from electronics. I’m going to do a disservice to electrical engineers all over the world by offering a very simple explanation. (My aplogies to you EEs out there).

Impedance is the measure of the amount that some object impedes (or obstructs) the flow of a current. Impedance might refer to resistance, reactance, or some complex combination of the two.

Perhaps an example is in order to illustrate impedance mismatching:

Imagine you have a low current flashlight that normally uses AAA batteries. Don’t try this at home, but suppose you could attach your car battery to the flashlight. The low current flashlight will pitifully output a fraction of the light energy that the high current battery is capable of producing. Likewise, if you attached the AAA batteries to Batman’s spotlight, you’ll also get low output. However, match the AAA batteries to the flashlight and they will run with maximum efficiency.

So taking this discussion back to software engineering, if you imagine the flow of data to be analogous to a current, then the impedance of a relational data model is not matched with the impedance of an object hierarchy. Therefore, the data will not flow with maximum efficiency, a result of the impedance mismatch.

Found a typo or mistake in the post? suggest edit

Comments

14 responses

Bill Gates • June 15th, 2004
Funny, I have not heard the term Impedance Mismatch thrown around when discussing object relational mapping. By the way, it is not polite to talk about "Impedance" in polite company.
Ian Griffiths • June 15th, 2004
A DC situation like this isn't really where you normally talk about 'impedance mismatches'.

The situations where 'impedance mismatches' are usually used are in situations with changing signals. For example, if you have a digital signal, with the voltage changing suddenly from 0V to 3V, if the wire carrying that signal has a sudden change in impedance, you tend to get reflections bouncing back down the wire.

So if there's an impedance mismatch, signals get messed up as they cross the mismatched boundary. I think this situation is somewhat closer in spirit to the way that 'impedance mismatch' is used in O/R mapping examples.
Haacked • June 15th, 2004
Good point. I should come up with a better example. With DC, we're more likely to talk about Resistance since the Reactance part of the equation will be 0 or close to 0.
Radu Grigore • June 15th, 2004
In fact impedance is very similar to refraction index, the difference being that they are practicaly relevant in completely different frequency domains. An example of mismatch is: you look thru a piece of glass (let's say you are admiring a new notebook in a store) and you see, vaguely, your face. This happens because air and glass have different refractive indexes so part of the light bounces back on the separation surface between air and glass. Exactly the same thing happens to signals when they cross from one type of cable (let's sey 50ohms) to another type of cable (let's say 75 ohms).

Using "impedance mismatch" for the troubles you get when you try to fit a relational DB with an OO application is merely an analogy (while the two phenomenons presented above are actually the _same_ at a basic level).

regards,

the EE guy
ai • March 2nd, 2005
to the EE guy

how do you solve the problem of impedance mismatch as in from a 50 ohm source to a 75 ohm load
The guy that pays for the dela • December 27th, 2007
In my experience this is a term "Architects" like to use when there is a problem with development that causes delays, also know as cost overruns.
ffoulks • April 21st, 2011
As has been pointed out, we are dealing here with the figurative use of a technical term. (I realize that engineers have difficulty with the notion of figurative language. An example: When Shakespeare writes "I [...] trouble deaf heaven with my bootless cries ...", he is using the notion of unshodness figuratively to represent abjection. When a figurative usage becomes common enough, the term acquires a new literal meaning.)
In the very broadest sense, impedance is what happens (figuratively) between two entities when they lack some expected structural similarity. When someone says, "Well it worked on my machine, why doesn't work on the network?", we have an impedance problem. In software development, within the time gap between the release of a new database version and the release of the application version that implements it, there is an impedance. The key here is "expected". In ordinary circumstances, it is reasonable to expect the application to mirror the database. Pointing out the impedance is a polite way of dealing with some idiot who doesn't understand the difference between the database and the application.
In the case of O/R impedance: When we use an automated tool to generate a UI and its CRUD from a database, we are pretending that the UI *is* the reality the database is supposed to represent. We have ignored the actions of the real objects, and when reality sets in in the form of business rules, we end up with an ad hoc mess. We haven’t really saved time by skipping the object level.
Ricardo • July 25th, 2012
I've seen the term "Impedance Mismatch" several times in papers about software or information systems. For instance:
"Information systems have traditionally suffered from an
impedance mismatch. Their operational environment is
understood in terms of actors, responsibilities, objectives,
tasks and resources, while the information system itself is
conceived as a collection of (software) modules, entities
(e.g., objects, agents), data structures and interfaces. This
mismatch is one of the main factors for the poor quality of
information systems, also the frequent failure of system
development projects." [A Requirements-Driven Development Methodology, J. Castro,2001).
Excellent explanation for the software development context. Thank you.
intekhab alam • October 28th, 2012
impedance mismathing means input impedance of an electrical load never equal to the fixed output impedance of the signal source to which it is ultimately connected
Gabriel Ramirez • April 15th, 2013
Great! i liked the explanation
SomeDudeOnline • December 3rd, 2014
I believe you're thinking of the word "Impotence"
nitizkumar • August 15th, 2016
Great article, love the way you broke it down, I wrote something similar but little more technical if some one is interested at http://www.scaffoldthis.com...
NN • February 9th, 2017
I know this comment is 13 years old but ... "frequency domains"? That terminology is twisted from its real meaning, which doesn't apply here anyway (or were you thinking of www.hertz.com?); just as "impedance mismatch" was a ridiculous analogy when Grady Booch used it in the first place. The phenomenon that he was trying to describe was simple enough to understand; the obscure analogy made it more confusing, not less
stoptryingtosoundsmort • August 10th, 2017
I don't like this usage of "impedance mismatch", to me it makes a simple concept more complicated. Everyone here trying to explain why its appropriate is focusing on the word impedance, but really I think "mismatch" is what the software people really mean. There is a mismatch between the OOP model and the relational model.... stop trying to make things sound more sophisticated than they are...