MassiveProng wrote:

On 3 Mar 2007 23:08:30 -0800, "David L. Jones"
Gave us:


On Mar 4, 12:38 pm, MassiveProng
wrote:

On 2 Mar 2007 15:09:30 -0800, "David L. Jones"
Gave us:


Which is why you do it for each range and then spot check it to see
that there is no funny business. Perfectly valid technique for home
calibration of a scope vertical scale.

Dave

It doesn't matter how many "places" you "spot check" it, you are not
going to get the accuracy of your comparison standard on the device
you intend to set with it. What you do is take the basic INaccuracy
of the device needing to be set, and add to it the basic INaccuracy of
the standard to which you are setting it. You CANNOT get any closer
than that. So, a 0.5% meter, and a 0.5% scope cannot be used together
to make the scope that accurate. You need a *finer* standard than the
accuracy level you wish to achieve.

You need to understand that as a basic fact, chucko.

LMAO!
If I use 0.5% accurate meter to adjust a something, then the accuracy
of that adjusted device at that point in time at that adjusted value
*becomes* 0.5%.


Absolutely incorrect!

If you do that, the MINIMUM error is 0.5%. It is ALWAYS greater
than that value by that value plus the error of the device you think
you set.

How can you not understand that basic fact?


The device that was adjusted only gets it's accuracy
figure of 0.5% *after* the adjustment.


Absolutely INCORRECT!

The error of a device is NOT tied to how it got set or what it got
set with, dip****, it is tied to precision of the circuits the device
are based upon.


The 0.5% of the device does NOT
get added to the 0.5% of the meter in this particular case!


Wanna bet?
Further more, if one did this procedure using thousands of meters to
"calibrate" thousands of other meters, the net resulting error is *NOT*
the sum; it is the square root of the sum of the squares!
But taking only *one* reference ("standard") and using it to
"calibrate" only one device, the result is technically indeterminate but
may be bounded by the sum of the (instrument) errors - and could be
*worse* (anybody hear of "cockpit errors"?).