[sdiy] OPA604 anyone?

[René Schmitz]

Ian Fritz wrote:

>> I take both Self's and claims on datasheets with a healthy portion of 
>> skepticism. I don't have the means to measure on this scales so I 
>> can't say for sure who is right or wrong here.
> 
> 
> Always healthy to be skeptical.  But given the performance and 
> reputation of the two parties, I'm heavily prejudiced towards BB.

I've studied the datasheet more closely, and I think I can explain the
"discrepancy" of a factor 10.

Lets assume we're building a 3.5x amp, and try to get an estimation for
THD at an amplitude of 7.75Vrms.

So lets take the graph for THD vs. Output Voltage in the OPA604
datasheet.

 From that we see that at 1kHz the THD is 0.0003% for 3.5Vrms(~10Vpp)
and 0.0007% for 7.75Vrms(~20Vpp). Or we can say this amplitude
difference does account for a factor of about 3.5 in THD. (doubly log 
paper means a power law :-)

Then noting from the figure of THD vs. F with parameter that the gain
where Burr Brown measured 0.0003% is G=1.

Judging from this graph we can estimate that the distortion for G=3.5
would lie about midway between the lines for 1 and 10 (this is double
log scale, and 3.16^2 = 10 Rounding 3.5 to 3.16 :-)) this would mean
roughly 0.001.

Now taking that factor 3.5 for the different amplitudes into account we
would end up at about 0.0035%.

Which is *surprise* quite near what Self has measured. IMHO there is no
discrepancy other than the figures for optimal performance (BB has
carefully chosen the amplitude where they spec'd THD to be close to the
minimum...) and those in an actual circuit.

Cheers,
  René


-- 
uzs159 at uni-bonn.de
http://www.uni-bonn.de/~uzs159