Quote from: tinman on October 18, 2015, 01:04:12 AM
What i cant find is the option where the scope displays the numerical value of that math trace/wave form.

Confirmed problem--no way to select Math as a source to measure against.  I think your Atten has the same guts as my Siglent.

Quote from: Dog-One on October 18, 2015, 01:49:58 AM
Confirmed problem--no way to select Math as a source to measure against.  I think your Atten has the same guts as my Siglent.

That is correct. After much research,i see many saying that the math trace cannot be used as a measuring source-and yes,it is the same as the siglent. The only way it can be done is to use the software that came with the scope,and do it via your computer

In this case,we will not need it,as we are now dealing with nice clean sine waves.

Second test.

Dissipated power of CVR on primary side now left out.

Enjoy

https://www.youtube.com/watch?v=RBhmjuuJehE

I watched test #2 and test #3.  I can see that you have a USB port so perhaps there is hope to reflash your firmware.

Test #3 was very good and well documented, and the 105% efficiency calculation amazes me because yet again, measurement uncertainties favour what looks like over unity.  It seems to happen all of the time, it's uncanny.  There doesn't seem to be a 50/50 split, it's more like a 98%-2% split favouring over unity when it comes to measurement uncertainties.

If I was in your shoes I would be looking to tigten up my measurements because the extra 5% is almost certainly well within your error tolerance.

Let's have a look at the secondary measurement for fun.  You make individual measurements of two resistors, so that's two sources of error tolerance.  Then you make two voltage measurements, which is another two sources of error tolerance.  Then you do two multiplications and one addition, which can compound the sources of error tolerance.  You end up with a pretty high stack of pancakes of error tolerance.

Without changing anything, you can make a single resistance measurement and a single voltage measurement.  Then do the multiplication and you get an output power calculation with only two sources of error tolerance multiplied together.

Let's look at that multiplication in detail:  (x + delta_x) * (y + delta_y) = x*y + x*delta_y + y*delta_x + delta_x*delta_y

Everything highlighted in bold is part of the error tolerance.  The first two terms are significant and the third term is small.  And that's just for one multiplication.  You can see how the pancake of error tolerances can quickly become a big stack.

So please consider doing the same measurements but work on strategies for making them "tighter."  If you do that don't be surprised to see the "extra" 5% melt away.

Quote from: MileHigh on October 18, 2015, 03:58:06 AM
I watched test #2 and test #3.  I can see that you have a USB port so perhaps there is hope to reflash your firmware.

Test #3 was very good and well documented, and the 105% efficiency calculation amazes me because yet again, measurement uncertainties favour what looks like over unity.  It seems to happen all of the time, it's uncanny.  There doesn't seem to be a 50/50 split, it's more like a 98%-2% split favouring over unity when it comes to measurement uncertainties.

If I was in your shoes I would be looking to tigten up my measurements because the extra 5% is almost certainly well within your error tolerance.

Let's have a look at the secondary measurement for fun.  You make individual measurements of two resistors, so that's two sources of error tolerance.  Then you make two voltage measurements, which is another two sources of error tolerance.  Then you do two multiplications and one addition, which can compound the sources of error tolerance.  You end up with a pretty high stack of pancakes of error tolerance.

Without changing anything, you can make a single resistance measurement and a single voltage measurement.  Then do the multiplication and you get an output power calculation with only two sources of error tolerance multiplied together.

Let's look at that multiplication in detail:  (x + delta_x) * (y + delta_y) = x*y + x*delta_y + y*delta_x + delta_x*delta_y

Everything highlighted in bold is part of the error tolerance.  The first two terms are significant and the third term is small.  And that's just for one multiplication.  You can see how the pancake of error tolerances can quickly become a big stack.

So please consider doing the same measurements but work on strategies for making them "tighter."  If you do that don't be surprised to see the "extra" 5% melt away.

The thing i find interesting is this. When extra energy seems to be shown,it is always assumed that there is a measurement error that go's against the posted result's-no mater how much care we take with them. I have never heard any like your self MH say that the measurement error could be the other way,and you have more extra energy that you measured.

Anyway,i did as you said,now that we have very smooth wave forms over resistive loads--we used the DMM to make the measurements.

Here is one thing to think about in this test. If !if! the DMM is making an error,then that error will be the same on both the input and output,as the voltage(and thus the wave form) across R1 and R2 are very close to the same,and the voltage(and wave form) across the primary and 98.4 ohm load resistor are much the same-as seen on the scope before the test.

As we have never seen a transformer configuration(and thus measurements to boot)like this before,do you think that there is a chance that maybe!just maybe! it might be worth while pursuing further. Maybe it might be time to rethink as to what field really is inducing current flow through the secondary in a transformer. Maybe the backyard boys have it right,and the books of yesty year are incorrect.

https://www.youtube.com/watch?v=jn5Hx_zO2pA