Derating Formula Confusion

[bostonman]

I'm looking at an RF chip and it can handle +25dBm at +25 degrees C and it's derated linearly to +20dBm at +100 degrees C.

I'm trying to create a formula in Excel so it will show the dBm at all temperatures between 25 and 100, but for some reason confusing myself on creating the formula.

Anyone have input?

[IanB]

Use the linear interpolation formula:

(20 - 25) / (100 - 25) = (P - 25) / (T - 25)

T is the given temperature in °C and P is the unknown power in dBm. For example, if T were 75°C the power rating would be 21.7 dBm.

[bostonman]

This helps balance the equation.... and actually this is what I had before, but how should I create a formula to use in Excel?

[rstofer]

Create a column of temperatures, create another column of derated values using the formula above.  Then plot the columns.

https://www.ablebits.com/office-addins-blog/2018/08/29/make-line-graph-excel/#create-line-graph-excel

[bostonman]

I just need a single cell to show a value when it references a temperature typed into another cell.

[rstofer]

Put this formula in cell B1 then enter various temperatures in A1

=IF(A1<=25,25,IF(A1>100,0,(25-(5/75)*(A1-25))))

No guarantees, check various values before committing to the formula.

The formula delivers 25 for all temperatures less than or equal to 25 deg C.  It does derating for temperatures above 25 degrees out to 100 degrees and returns 0 for temperature above 100 degrees.

[dmills]

But that derates linearly in dB not linearly in power.
I would bet that you want to convert to mW first, apply the derating then go back to dBm.

I would also suggest that most RF parts are getting seriously non linear long before they hit the thermal limits (Exception might be an attenuator).

Regards, Dan.

[bostonman]

That formula worked, but the datasheet does state:

[/I'm looking at an RF chip and it can handle +25dBm at +25 degrees C and it's derated linearly to +20dBm at +100 degrees C.quote]

[rstofer]

That formula worked, but the datasheet does state:

[/I'm looking at an RF chip and it can handle +25dBm at +25 degrees C and it's derated linearly to +20dBm at +100 degrees C.quote]

The formula just extends the values below 25 deg C and makes the assumption that the magic smoke escapes above 100 deg C.  Something needs to be put in the formula or it needs to flat reject values < 25 and > 100.  It is easily changed to do that.  In fact, for > 100 deg C it simply returns 0.  It could do the same for < 25.

You could put the formula =25-(5/75)*(A1-25) in B1 and promise to not enter values < 25 or > 100 deg C.  That is not a very robust solution.  In the end, the formula should do something sensible for all real numbers.  I guess we could limit it to realistic numbers.  Temperature below absolute 0 doesn't make sense nor does it make sense above, say, 1200 degrees C where Silicon melts.

I have no idea what the datasheet says because I didn't find a link.  Sometimes the datasheet will have a graph of the derating or Safe Operating Area (SOA).

[bostonman]

The chip is: MLD-1640

[Ian.M]

I'd be somewhat inclined to interpret the  MLD-1640 datasheet as derate absolute power linearly, as that's what's the norm for nearly every other semiconductor device that uses linear derating.   If they want you to derate the dBm number linearly they should have explicitly said so.  However derating linearly by dBm would derate more aggressively at lower temperatures (still in the range 25 deg C to 100 deg C), so if in doubt, and in the absence of input from their application engineer, it would be more conservative to do that.

[rstofer]

I'd be somewhat inclined to interpret the  MLD-1640 datasheet as derate absolute power linearly, as that's what's the norm for nearly every other semiconductor device that uses linear derating.   If they want you to derate the dBm number linearly they should have explicitly said so.  However derating linearly by dBm would derate more aggressively at lower temperatures (still in the range 25 deg C to 100 deg C), so if in doubt, and in the absence of input from their application engineer, it would be more conservative to do that.

It seems to me the datasheet said exactly that.  They gave a dBm value for 25 deg C and a dBm value for 100 deg C and told us to derate linearly between these endpoints.  It sure seems to me that they gave the endpoints of a straight line in terms of dBm.

Of course, I could be very wrong...

[IanB]

I don't think dBm vs power makes much difference over such a small range. 20 dBm is 100 mW and 25 dBm is 316 mW. Half way between 100 and 316 is 208 mW, which is 23 dBm. So interpolating on the dBm scale to 22.5 dBm is conservative and therefore safe.