Measuring capacitors and inductivities not related to speakers

redmountain · Nov 21, 2024

Is this possible,
and how exactly ?
Sense resistor values ?
YT videos ?
Thank youand best regards

Klaus

John Mulcahy · Nov 21, 2024

Sure, same as any other impedance measurement. REW can fit a component model to the measurements after you have made them.

luigi57 · Nov 27, 2024

Hi John,
I'm using your fantastic software to test audio transformer for MC cartridge; my task is to measure all transformer parameters to simulate it.The transformer I want to test is an old Audio Tecnica AT660T with a step-up ratio ranging from 1:34 to 1:125: do you suggest to measure impedance from the secondary side in order to have a step-down transformer?
Why low-frequency is limited to 2Hz? Is there any possibility to extend to 0.2Hz?
Many thanks for your suggestions.
Luigi

John Mulcahy · Nov 27, 2024

You can change the graph start frequency in the View preferences. I don't have any advice on transformer measurement, sorry.

luigi57 · Nov 28, 2024

Thank you John.
But, also the sweep frequency can be lower than 2 Hz?

dcibel · Nov 28, 2024

Sweep can start at 0Hz, but IR window needs to be >500ms to capture frequencies below 2Hz.To capture frequencies to 0.2Hz requires a window of 5 seconds.

luigi57 · Nov 28, 2024

Thank you Decibel.
Do you mean that to capture frequencies to 0.2Hz only the right IR Window must be >5 seconds for sine sweep, or even the left IR Window must be changed?

dcibel · Nov 29, 2024

0.2Hz wavelength is 5 seconds long, so total window needs to be at least 5 seconds, simple as that. How much left and right depends on where the reference marker is located. If ref at IR start, the 5s right, 0 left as bare minimum

dcibel · Nov 29, 2024

Keep in mind that most audio interfaces have DC blocking caps on the input, so you may not find very high signal level at 0.2Hz, SNR could present some problems in accuracy/reliability of results. As well, both channels of audio input need to be well matched, which will primarily be affected by component tolerances in that DC blocking high pass circuit. Simple loopback test of each input channel will confirm on your equipment.

luigi57 · Nov 29, 2024

My sound card has -3dB points at 1,4Hz and 80KHz and the interface to sound card has -3dB points at 0,16Hz and 250KHz. My interest is on measuring impedance and amplifier, not speakers response. SNR, for sure, can limit low-level results.

trobbins · Nov 29, 2024

-3dB at 1.4Hz for a loopback is a good start for using the low frequency capability of REW.

My EMU 0404 has a DAC/headphone output with -3dB at 1.5Hz (up to 94kHz), but the ADC/input amp coupling is far more limiting with -3dB at 2.8Hz at 44.1/48kHz sampling, rising to 5Hz at 192kHz sampling (even though the AK5385 HPF -3dB spec is down at 1Hz for 44/48kHz sampling). With respect to using a compensation curve to flatten the loopback response, the compensation is -11dB and 140 deg phase at 2Hz, so some heavy lifting from the compensation when measuring below 2Hz. The bulk of that low frequency roll-off is from 47uF 63V coupling caps, which can be swapped out for higher capacitance parts (and even at lower voltage if phantom power is disabled), although it could be practical but quite risky to bypass them for dc response if the soundcard was dedicated to a particular testing setup.

redmountain · Dec 7, 2024

I set up impedance measurement
Sense resistor is 66 Ohms
Will this work for up to 4 Hy coils / 68 Ohm resistance ?
Cap up to 3 µ F read ok

Klaus

John Mulcahy · Dec 8, 2024

Should be fine.

trobbins · Dec 9, 2024

Hi Klaus, it looks like you may be measuring iron or ferrite cored inductors? If so then their inductance may be significantly affected applied signal voltage level, which may then need to be part of the measurement reporting. Does your application also pass dc through the inductor?

redmountain · Dec 11, 2024

Yes
> https://ferrite.de/en/products/p-cores/p18-slash-11/
I developped with these some 40 years ago
octave LC filter EQs with +-20dB and FET circuit...
And now maybe a revival...This thing sounds terrific :-))
I am Measuring with a Mac headphone out @ about 5dB headroom FS
Should I compare @ other levels ?
I struggled with the 2 half ferrites by applying different force on
holding them together...
...Measuring with REW the inductance

Thanks

Klaus

trobbins · Dec 11, 2024

Ta - so a P18 pot core - and yes very common 40 years ago

.

Signal voltage shouldn't then be a concern wrt the core material. The measured inductance may have a large tolerance between samples if the coreset is ungapped. Do you only have ungapped coresets, or some gapped sets? Are you using the default core clamping clips? Do the coresets include centre leg adjustment?

redmountain · Dec 12, 2024

https://cdn.radiantmagic.com/product_images/86/24/P-Kerne_250px.jpg?1621679557
two halves, ungapped AFAIK
I will talk to Ferroxcube supplier in Germany...

Thanks a lot
Best

Klaus

redmountain · Dec 12, 2024

@ John
REWv 5.31.3
In dark mode, component view - info isn't visible
( letters are dark )
In light mode I can see the dark letters
Best

Klaus

redmountain · Dec 12, 2024

Opening a file with REWv 5.31.3 with multiple measurements gives funny phase pics
on some measurements...
Not so in REWv5.20.13

John Mulcahy · Dec 12, 2024

Please attach mdat files rather than images.

redmountain said:
In dark mode, component view - info isn't visible

Thanks, fixed for the next V5.40 beta build.

redmountain · Dec 13, 2024

redmountain said:
Opening a file with REWv 5.31.3 with multiple measurements gives funny phase pics
on some measurements...
Not so in REWv5.20.13

Today, opening these files everything displays normal'
???

redmountain · Dec 13, 2024

Hmmhh...
The impedance measurements are inconsistent here...
Component model available on some measurements and not at all on others
Restarting REW gives another behaviour...
I check with SPL measurements and these are ok
It's impedance that isn't consistent,
at least with my measured ferrite coil mentioned above
??
Don't know how to troubleshot...

Klaus

redmountain · Dec 13, 2024

Screenshot and mdat...

redmountain · Dec 13, 2024

10 minutes later...

John Mulcahy · Dec 13, 2024

Your measurements become extremely noisy at high frequencies, so much so that the model fit can fail. You can help that by using a lower end frequency for the measurement and by applying some smoothing. You could also start the measurements at a lower frequency, e.g. 10 Hz, to provide some additional data.

Measuring capacitors and inductivities not related to speakers

Member

REW Author

Registered

REW Author

Registered

Member

Registered

Member

Member

Registered

Member

Member

REW Author

Member

Member

Member

Member

Member

Member

Attachments

REW Author

Member

Member

Member

Attachments

Member

Attachments

REW Author

Popular tags