Don Vischulis wrote:
> 
> Dear Ralph & cbyxbee
> 
> The concern for the temperature rise above ambient is always a concern
> for reliable operation of a circuit.  My experience indicates that the
> voltage (I*R) drop through the circuit is quite often the limiting
> factor.  Either the voltage drop will prevent proper operation of an
> electronic device or will produce noise in the circuit.
> 
> From the circuit parameters given, a calculation of the resistance per
> inch of the circuit is possible.  I don't recall how to calculate it,
> but an electrical engineer or someone involved in hybrid circuits should
> be able to produce the calculation.
> 
> Don Vischulis
> [log in to unmask]
> 

Snip-snipped all the other stuff.

Don, you are absolutely correct.

The existing "thermal rise above ambient" charts are somewhat useless for "serious" 
printed circuit (electrical) design.  They are only good for regulatory 
requirements (from a NRTL - like UL) for "ensuring or insuring" the printed board (and 
it's assembly) don't get too hot, burn the insulation off the conductors, and burn the 
house down.

For most "signal" conductive patterns, the conductor size was determined more by the 
"minimum" design width of the conductor requirements due to "manufacturing 
producibility" limitions rather than conductor current density and the resulting "IR" 
voltage drop.  In the case of "higher" current power/ground distribution, the 
"resistance" of the conductor is more of a driver.  Personally, I recommend the use of 
"thermal rise above ambient" charts as a wall chart for reference only.  The "ohms per 
square" method to determine the conductor resistance and calculate the IR voltage drops 
and conductor width/length to approximate the inductance for the required electrical 
signal integrity requirements (and yes, power/ground distribution has electrical signal 
integrity requirements).  Then look at the wall chart to see if it meets the "thermal" 
requirements -- it always does because the conductor current density (I sqr R loss) to 
meet "serious" electrical requirements is always going to be less than for conductor 
resistance-heater (based on the thermal rise above ambient) to help warm up the ambient.

The problem with this approach is that the design engineer must know and establish a 
"noise" budget that includes such things as power/ground loss, switching, ground bounce, 
under/over shoot, thermal, and all the other voltage/current variations early on in the 
design.  (For years, I refer to this concept with the catchey phrase "printed board by 
design and not by default".)

In past postings (they should be in the technet archive), I've commented (some of 
our fellow technetters would say rather lengthly) on current carrying capacity and sheet 
resistance (ohms per square).  Yes, using sheet resistance (like the hybrid and 
integrated circuit people do) is easy to approximate conductor resistance.  For example, 
35 micrometer thick copper has a sheet (easy to remember rule-of-thumb) resistance of 
approximately 600 microhms/sq.  A conductor 50 mm long and 2.5mm wide would be 20 
squares long, times 600 microhms/sq. would give you a conductor resistance of 12 mohms. 
 For a 10 deg C above ambient thermal rise, D-275 would allow about 4.5 A of current 
flow.  Electrically the voltage drop would be 54 mV (4.5 A X 12 mohms) which in some 
applications would not be acceptable.  The voltage drop is proportional to conductor 
length (for the same condutor width and current density), the thermal rise is a 
constant, regardless of length.  This does not include the "non-dc frequency effects" of 
inductance, capacitance and impedance - but then these are in the IPC-D-31x high 
speed/frequency design/performance documents.

-- 
Ralph Hersey,

Ralph Hersey & Associates
PHN/FAX 510.454.9805
e-mail:  [log in to unmask]

***************************************************************************
* TechNet mail list is provided as a service by IPC using SmartList v3.05 *
***************************************************************************
* To unsubscribe from this list at any time, send a message to:           *
* [log in to unmask] with <subject: unsubscribe> and no text.        *
***************************************************************************
* If you are having a problem with the IPC TechNet forum please contact   *
* Dmitriy Sklyar at 847-509-9700 ext. 311 or email at [log in to unmask]      *
***************************************************************************