maximum range of impedance measurement

JLM1948

Member
Thread Starter
Joined
Jan 12, 2019
Messages
124
Hi John,
I have designed a small add-on that is supposed to allow measuring high-impedance components such as inductors and LC tanks, which is a quite common request from DIYers.
It's basically a driver stage allowing to inject up to 10mA AC and as low as 1 uA, and a high-Z buffer (100 Megohms) for the DUT measurement.
It seems to work OK, since I get correct measurements of a few transformers I had lying around. The only problem I have is that the Y-axis is limited to 1k. I've tried to change it but it is stubborn.:)
I could probably find a workaround by fooling the reference, but I think it would be much simpler if I could change the scale up to 10 Megohms.
 

JLM1948

Member
Thread Starter
Joined
Jan 12, 2019
Messages
124
Hi John,
I have reviewed the impedance calibration procedure.
The initial cal is good, with less than 1dB difference between channels.
Then I make the first measurement with zero ohm specified for the sense resistor. I get the following screen:
open cal avg4 1.jpg


Then I try to make te next measurement (short_circuit), but I don't have access to it because the button is red.
measurement panel 1.jpg


So I'm stuck...It seems the calibration file has not been completed, but I don't know why.
 

JLM1948

Member
Thread Starter
Joined
Jan 12, 2019
Messages
124
Hi John,
I have tried with the new procedure, with initial conditions here:
initial test conditions 1.jpg

and the levels here
levels during initial test 1.jpg

and the results, which I can't make sense of.
initial test conditions result (open) 1.jpg


And I'm still stuck here, because the open circuit test is in red, which I understand means it has not been validated.
Is there an action required to validate it?

BTW: apparently the input impedance cannot be specified at more than 1Meg. I tried entering 100Meg, which is the actual resistance of the input buffer, to no avail.
I understand that the leads and parasitic capacitance may fool teh calculation, if it is based only on amplitude, ignoring phase.
 
Last edited:

JLM1948

Member
Thread Starter
Joined
Jan 12, 2019
Messages
124
meantime I've been thinking about it. With such a high load, the voltage drop in 1K has to be very small, so noise eand rounding errors make the result uncertain.
I think I should revise the method, with the load grounded and the DUT in series between the source and the load, so I would have a graph of current. Is there a way to invert for displaying the impedance?
 

John Mulcahy

REW Author
Joined
Apr 3, 2017
Messages
7,212
Did you read the help I linked? The calibrations are carried out by clicking on each of the buttons in turn. REW can measure impedance accurately over a wide range using the new calibration procedure, which compensates for magnitude and phase across the frequency range and accommodates the complex impedances of the input and the test leads. Please try following the processes and then see what benefits additional circuitry might bring. With the standard setup a source with low output impedance will reduce measurement noise.
 

trobbins

Member
VIP Supporter
Joined
Nov 26, 2017
Messages
121
Location
Australia
Perhaps re-check that you can do a simple direct loop measurement from your output to each input, in turn - to confirm you have the input channels identified and working and with levels that have sufficient headroom and dynamic range. Perhaps then connect your sense resistor and wiring, but again repeat those two simple direct loop measurements, to provide confidence that your setup wiring is still ok.
 

John Mulcahy

REW Author
Joined
Apr 3, 2017
Messages
7,212
I did.
As I wrote earlier, I can't because the first button stays red, and I don't have access to the others. I can restart the first button, that's all.
What message is shown afterwards, if any? Please attach the result of that calibration measurement as an mdat file.
 

JLM1948

Member
Thread Starter
Joined
Jan 12, 2019
Messages
124
Hi John,
Took me some time to build a clean jig according to the Help and triple check everything.
Here is the result of the open test.
simple jig open meas.jpg

I don't know if it's correct or not, since I don't know what to expect from this open test. Is it supposed to determine the actual input impedance value?
After that, I know that the normal procedure is to run the short-circuit test, but I can't access it, the open test tag is in red and I cannot access the short circuit test. Of course I can run teh oepen test an infinite number of times.
Again, I suspect that REW does not consider the first test as valid.
 

John Mulcahy

REW Author
Joined
Apr 3, 2017
Messages
7,212
The help says what to expect, your result is abnormal:

The completed measurement shows the level of the measurement channel (usually right) compared to the reference channel, where a reading of 100 Ohms corresponds to 100%, 99 Ohms would be 99% etc. If the difference between the 2 channels is too large (more than 2 dB) the calibration is abandoned as it is likely there is a connection error or the input channels have different gain settings, re-check the connections and channel gains and try again.
 

JLM1948

Member
Thread Starter
Joined
Jan 12, 2019
Messages
124
simple jig levels during initial test.jpg


The 30k reading would mean that the diiference in level is big, isn't it? I use a 1k sense resistor.
 
Last edited:

JLM1948

Member
Thread Starter
Joined
Jan 12, 2019
Messages
124
Here are the loopbacks.
loopback L a.jpg


loopback R a.jpg


The reason loopbacks are at -27dBfs is because the soundcard's input limiter sets in at -25.
i had noticed that and it didn't pose any significant problem for most measurements since I have an interface that provides buffering, gain and balancing.
Now I suspect that some non-linearity happens below -25, which could very well false the impedance measurement.

I'll use my other set-up, but it will take some time to build adapters, since it's separate ADC and DAC with XLR connections.
 

John Mulcahy

REW Author
Joined
Apr 3, 2017
Messages
7,212
An input limiter at -25 dBFS? That would be very unusual. The Komplete's mix knob is turned all the way to Host?
 

John Mulcahy

REW Author
Joined
Apr 3, 2017
Messages
7,212
Something doesn't seem to add up. When you click the Open circuit cal button you should see this dialog:

54776


Clicking OK should then run the measurement and show a result titled "Impedance open cct", but your results look like the result of pressing the Start button rather than the Open circuit cal button.
 

JLM1948

Member
Thread Starter
Joined
Jan 12, 2019
Messages
124
An input limiter at -25 dBFS? That would be very unusual. The Komplete's mix knob is turned all the way to Host?
Yes. I found that the hard way, when I used it to graph the frequenvcy response of a x-over and the ratio of a liliter.
 

JLM1948

Member
Thread Starter
Joined
Jan 12, 2019
Messages
124
Something doesn't seem to add up. When you click the Open circuit cal button you should see this dialog:

View attachment 54776

Clicking OK should then run the measurement and show a result titled "Impedance open cct", but your results look like the result of pressing the Start button rather than the Open circuit cal button.
Correct.
I just re-did the whole procedure with another soundcard (AKAI HR) with 1Meg input Z, and everything went according to plan, except that the measurements are significantly off. A 47nF cap that measures 45.4 with my LCR meter is identified as 60+nF. measurements on the secondary of a mic input xfmr are identified as a 1.18uF capacitor.
I'll post the screens later.
 
Top Bottom