diyAudio

diyAudio (https://www.diyaudio.com/forums/index.php)
-   Everything Else (https://www.diyaudio.com/forums/everything-else/)
-   -   Zip cord for speaker test (https://www.diyaudio.com/forums/everything-else/371099-zip-cord-speaker-test.html)

Pano 18th April 2021 03:02 PM

I can run some IMD tests tomorrow. I'll try 3 different speakers.

jneutron 18th April 2021 03:14 PM

While I use transmission line theory to calculate the settling time in the tens of microseconds domain (that falls apart after 10's of microseconds), it could just as easily been evaluated using a simple LCR model.

For your speaker, the 78 hz and 2.1Khz peaks will have zero settling time if you use a 25 to 30 ohm RFZ cable. The 5 to 10 ohm midrange band will also have very little settling time as well, the ratio maxing out at 5 to 6:1, represented on my graph by the 20 to 16 ohm impedance which settles to 90% in 1.2 to 1.6 uSec.

If the RFZ of the split zips is 500 ohms, then the peaks are 20:1, settling to 90% at 5 uSec, but the midrange valley is 100:1, settling quite off my chart, the 50:1 hits 90% at about 15uSec, 100:1 I'm estimating at about 20 uSec.. So the midrange settling time is much longer.

Raising the cable RFZ should then cause a temporal shift of the midrange band in the 20 uSec range (granted, to 90%), a delay in the frequencies we are the most sensitive to ITD. The shift at 78hz of 5 usec is meaningless to human perception, we will focus instead on the amplitude and see that content as centered and unmoving..at 2Khz, 5 usec is probably below human discernment as well, but there is a dithering grain of salt there so I cannot say that with full confidence.

My test is designed to look for that delay with respect to the centered image content around the 78 hz peak, as it will shift far less and we are not sensitive to ITD in that band. The 2khz content will also be more time stable, so even if it is within the ITD sensitive range, the shift is far less.

jn

To get closer sim values, use the two RLC models taken from the two cable set RFZ's, and plug them into a sim with that speaker at 78hz, 400 hz, and look for the relative shifts in time.

edit: for a cable with RFZ at 25 to 30 ohms, the total spread in settling time (to 90%) across all frequencies falls below 2 uSec. I would consider this to be below human thresholds. And by not going any lower in RFZ, there is little risk of the phase margin crossing zero when the speaker unloads before unity gain BW of an amplifier. So the cable should not require a zobel.
edit2: for your speaker, the midrange band has the highest RFZ to impedance ratio, so it is expected that that band will be more sensitive to IMD changes with the higher RFZ cable. I would recommend running the IMD tests in the midrange area to see if IMD changes level with the high cable RFZ.
Edit 3: And I want to thank you for taking this effort, regardless of outcome...

Pano 18th April 2021 07:48 PM

3 Attachment(s)
You're welcome! Just please let me know exactly what you need as this is new territory for me. I want this test to clearly show what you are looking for.

Here are plots for 3 other speakers that can easily be tested. A small fullrange in a sealed box, a cheap 2-way in bass reflex and a vintage Magnepan MG-IIB. If any of these look like good test mules, please let me know.

Pano 18th April 2021 07:53 PM

REW includes several standard dual tone tests for IMD, and also custom frequencies and ratios. We can test just about anywhere you want. Even multitone. I would suggest not going above a sample rate of 96K, as my soundcard isn't as clean above that as below.

Richard Ellis 18th April 2021 08:12 PM

I foresee some cherished 'beliefs" utterly dashed here, but not without lots of arguments as to proceedures & resultant "interpretations" of the acquired data.
Too bad Transmission Line Theory doesn't apply in this domain...akin to comparing Light-speed to the speed-of-sound. Good luck all.





--------------------------------------------------------------------------Rick....

jneutron 18th April 2021 08:39 PM

Rick,
You are not the first to misunderstand, and you won't be the last.

You obviously missed (or glossed past) where I stated a T line analysis falls apart when you go past 10 or 20 microseconds.

It has also been proven in the past that a properly done multiple RLC element analysis approaches my t-line results as the number of elements increase.

Perhaps you should read Cyril Bateman's papers. I especially enjoyed when he built a reflection bridge that showed audio reflections in a cable connected to a speaker. I hadn't thought that possible because the wavelengths are so so long.

Edit: Rick, the experiment is being setup in order to either prove or disprove a hypothesis. Nothing more, nothing less. Beliefs "dashed" has nothing to do with it. It is about the possibility of advancing knowledge.

jn

jneutron 18th April 2021 08:43 PM

Quote:

Originally Posted by Pano (https://www.diyaudio.com/forums/everything-else/371099-zip-cord-speaker-test-post6622335.html#post6622335)
You're welcome! Just please let me know exactly what you need as this is new territory for me. I want this test to clearly show what you are looking for.

Here are plots for 3 other speakers that can easily be tested. A small fullrange in a sealed box, a cheap 2-way in bass reflex and a vintage Magnepan MG-IIB. If any of these look like good test mules, please let me know.

Your first plot, the WT3, is by far the best candidate as it has the low impedance midrange with the highs on either side. A speaker that has very constant impedance through the entire audio band will be useless for this experiment, as nothing is varying.

The crux is to find a set of frequencies where the amplifier helps to suppress IMD caused by the speaker, and the high RFZ cable gets in the way of the amplifier doing that. I've no idea what frequencies, but the test signals should be centered in the midrange valley.

jn

jneutron 18th April 2021 09:22 PM

Pano, was thinking about it. Perhaps a test signal near but above the 78 hz resonance (maybe 80 or 90 hz), and a second somewhere in the impedance valley. The excursion of the LF will be trying to change the eddy losses of the higher frequency due to pole and front plate conductivity, so that might produce some interesting phenomena that the high RFZ cable is not as well equipped to keep under control.

jn

Pano 18th April 2021 09:28 PM

OK. :up: It's easy enough to poke around at various frequencies. I'll try a few and post them.

Speedskater 19th April 2021 08:26 PM

That cable looks like one Enid Lumley (RIP) did decades ago.


All times are GMT. The time now is 01:10 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2021 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 16.67%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2021 DragonByte Technologies Ltd.
Copyright ©1999-2021 diyAudio

Wiki