RF (radio frequency) tests performed on my system.

[anthonyTD]

I agree, it must have an impedance/load somewhere, otherwise it would disapate nothing, all very interesting, keep scratching your head, watch you dont get splinters though.

I put my RF hat on last night and had a think about how the grounding box cables are made.
The construction will effectively create a balun that can reflect an impedance connected to the end of the cable.

The length of the cable (balun) will determine where impedance nulls and peaks will occur across the frequency range. This is the theory so the test is to find out if the grounding box actually has an impedance.
This impedance is relative to free space (RF/antenna engineers will understand this). The impedance of free space is 377 ohms, though this could be thought of as a construct as it is calculated from the theoretical capacitance and inductance of free space, or a vacuum if you will.

What is really needed is a Network Analyser to measure the reflected impedance but I don't have one.
The next best thing is to use the calibration signal of the Spectrum Analyser which is a 50MHz square wave, so displays frequencies at 50MHz intervals across the screen.
I quickly made an adaptor so the grounding box cable could be connected directly to the Spectrum Analyser.



The spectrum was monitored with and without just the lead connected, no difference was found. When the grounding box was connected to the other end of the cable a difference was observed, but only to the bottom 4 frequencies, namely 50/100/150 and 200MHz.

The differences were small but repeatable so the Analyser was set to display to 200MHz and at its maximum vertical resolution of 3dB per graticule line.

Display with/without just the cable connected:



Note the respective levels of the four spikes.


Display with cable and grounding box connected:



Note the 50 and 150MHz responses have increased and the 100 and 200MHz responses have decreased.


OK what does this tell us? It doesn't tell us what the impedance is as that is beyond the scope of the test.

It does confirm that the grounding box has an impedance at RF.

[Gaz]

Are your scopes able to focus/magnify in on the audio band and surrounding area? Say from 5hz to 30khz?

Wow, are you a bat?

Most humans can only hear 20hz to 20khz

[Edward]

Wow, are you a bat?

Most humans can only hear 20hz to 20khz

Not at all Gary. I can hear from about 20hz to about 16khz. The point I was making, not terribly well I guess, is the possibility of stuff just outside the human audio band possibly impacting the human audio band (as supertweeters are often claimed to do). Also Alan's instruments were 'seeing' the full range up to 1ghz, not in or near the human audio band.

But I guess your recognised all this?

[Macca]

Seems to me that if we are going to consider that frequencies beyond the limits of human hearing can be <ahem> 'perceived' (despite the lack of any evidence to support this) then I don't see how some arbitrary cut off at any point in the frequency range all the way up into gigahertz can be consistent with this theory.

BTW The reason we can hear supertweeters is because they have output within the accepted range of human hearing. No mystery there.

[jandl100]

BTW The reason we can hear supertweeters is because they have output within the accepted range of human hearing. No mystery there.

Yup, listen to most 'supertweeters' on their own and you can easily hear them squeaking away. My guess is that they usually start at 8-10khz.
Bit of a cheeky con imho.

[walpurgis]

Suertweeters that have clearly audible output are being misused. They should roll in above the output from the main tweeter, usually somewhere above 20kHz.

The exception is use with the Celestion HF1300 which output cuts off abruptly at around 13.5 kHz. A supertweeter used above these will have output audible to some, when employed with the correct 12/18 db/oct intro slope.

[jandl100]

Sure, they should be properly integrated into the speaker design with a bespoke crossover - but I bet the vast majortity are simply used as add-ons plonked on top of the main speakers leaving them to squeak away and sound 'impressive' because of the accentuated upper treble.

[anthonyTD]

Suertweeters that have clearly audible output are being misused.

[struth]

Sure, they should be properly integrated into the speaker design with a bespoke crossover - but I bet the vast majortity are simply used as add-ons plonked on top of the main speakers leaving them to squeak away and sound 'impressive' because of the accentuated upper treble.

depends on the supertweeter. some only work over a certain freq, and others need a cap of correct value in line

[walpurgis]

Most tweeters seem to roll off at the top by around 12db to 18db/oct, so intro slope of the supertweeter needs to match this. Both need to be down a few db in output at the response overlap.