Yes. You can calculate the effect by considering impedances in parallel.RNMarsh said:
Most amp topologies will give higher distortion into a lower impedance load, so yes it may be audible. Any benefit does not come from this, but from reducing RF pickup in the cable.
He appears to have used a 50R bridge (see bottom of page 6) to measure a non-50R cable (Supra, 203pF/m so not 50R, see top of page 2, and page 13). Hence the results are meaningless. A 'reflection bridge' does not magically distinguish reflections from forward power; instead it measures voltage and current in the cable and calculates what the forward and reverse signals would be to give that voltage and current assuming a particular cable impedance.jneutron said:
On page 6 he says
This is only true at the very lowest audio frequencies. Across most of the audio band R and C dominate. In both cases wave propagation does not occur. When R and G dominate you get a distributed potential divider. When R and C dominate you get something like 1-D diffusion or heat flow. In this case I am not sure that the idea of 'reflection' has any meaning, as reflection is a wave phenomenon.
The dominance of R and C are shown by his own figures (bottom of page 13) although it is unclear how he measured these. At 1kHz G is 0.298uS, but wC is 6.2uS; R is 0.0926 while wL is 0.011 - C and R dominate. This table includes a column labelled Z0, which must be wrong because it just gives one number in the region where the result will be complex and so requires two numbers. Perhaps he meant |Z0|? Failure to distinguish between Z0 and |Z0| often trips up transmission line newbies.
On page 7 he says
This assumes that both the termination and cable impedances are resistive, which is extremely unlikely in the audio frequency region. More Z0 vs. |Z0| confusion?
I am going to guess that he is someone whose equipment capability exceeds his ability to draw correct conclusions from that equipment.
Last edited:
How did CB measure 10kHz reflections with the bridge that starts at 100kHz??
8721A Directional Bridge, 100 kHz to 100 MHz [Obsolete] | Keysight (formerly Agilent’s Electronic Measurement)
P.S.: This starts to remind me his capacitor articles series. Audiophiles then started to argue how bad are the capacitors.
8721A Directional Bridge, 100 kHz to 100 MHz [Obsolete] | Keysight (formerly Agilent’s Electronic Measurement)
P.S.: This starts to remind me his capacitor articles series. Audiophiles then started to argue how bad are the capacitors.
Last edited:
Yes, I did not realise until after I posted that the article was by CB. He did some useful work on capacitors, although sadly his results are often misunderstood - they actually show how good most capacitors are. On speaker cables he may have been at least partly out of his depth.
Still a bit of fun to look at this and a LOT more interesting than discussing the merits of glorping things 🙂
Edit: and agree CB did prove capacitors are a lot better than many people are programmed to believe.
Edit: and agree CB did prove capacitors are a lot better than many people are programmed to believe.
Yes but many people did not take into account he measured distortion directly across the capacitor terminals and not on the load resistor of some usual value. So it might have said something for speaker crossover capacitors but not for the amplifier coupling capacitors.
DF and PMA..very valid observations, both on 50 ohm as well as lower frequency limit.
The answer is, he made his own bridge. That was the appendix 2 thing I mention that is for some reason, not available on any of the links. it was made to work between 1Khz and 1Mhz. I have that in both electronic and hard copy form, but I'm waiting for the person who sent it to me to chime up as well as let me know if I am allowed to share that content.
As he passed away, I am absolutely in the dark as to whether I can provide the material under the copyright fair use doctrine.
It is frustrating to me.
And DF, your "guess" is just that, a guess.
cheers, jn
The answer is, he made his own bridge. That was the appendix 2 thing I mention that is for some reason, not available on any of the links. it was made to work between 1Khz and 1Mhz. I have that in both electronic and hard copy form, but I'm waiting for the person who sent it to me to chime up as well as let me know if I am allowed to share that content.
As he passed away, I am absolutely in the dark as to whether I can provide the material under the copyright fair use doctrine.
It is frustrating to me.
And DF, your "guess" is just that, a guess.
cheers, jn
Originally Posted by vacuphile View Post
... Loudspeakers that measure identical on the most relevant parameters will sound identical....
... Loudspeakers that measure identical on the most relevant parameters will sound identical....
I did not read his science fiction scenario. Fun is the fact that I said exactly the opposite of what I imagine he imagines?
in his part two article, he demonstrates reflections of a 10 Khz signal in a 4.9 meter long cable. So, 5 times longer wavelength, 80% of the length..
Even in his part 1 article, figure 8 on page 7 (which I previously pointed out and you ignored), a 10Khz wave showing reflection in a 4.9 meter cable.
What is of interest in figure 8, is that he shows the delay. That delay is the multiple reflection pass composite, and it will exactly match the lumped LCR element analysis.
Or, will you ignore that as well?
jn
I do not know, but his article says
So I started to search for HP8721A specifications.
Further, his oscilloscope sreen shows no time and no voltage division, no us/div no V/div. To me, such plots are always suspicious.
So I started to search for HP8721A specifications.
Further, his oscilloscope sreen shows no time and no voltage division, no us/div no V/div. To me, such plots are always suspicious.
Last edited:
Richard, could-you provide the references of your drivers, if they are accessible individually on the back of the M2, for I can calculate the two RLC and RC compensations networks for each speaker, if you are interested in trying-it ?
I'm sure you will enjoy the result, apart, may-be, for the RLC of the bass driver, that you could prefer without, if you are lacking some bass.
I'm not a specialist of microwaves (;-) so, I had not read all here about this controversy about R terminals efficiency.
I just have a single and simple question: What could be the utility of a 100 Ohm termination resistance, if the speaker that end the cable is 6 Ohms up to very high frequencies (RC) and the source made, as usual like a LR in //( 15 Ohms, 1µH) , or near 0 Ohm (real life) ?
Last edited:
So is for me. I'm not buying any of these dubious/strange experimental result until I see a theoretical justification/model that supports the findings, the interpretations, and their relevance. This is not quantum mechanics, so it should be a pretty easy exercise to create a mathematical model.
As I previously said, a 1000m fundamental wavelength signal and its harmonics, TEM propagating on a 1cm unmatched TL, does create reflections. Using the classic TL model that every EE was taught, one could easily calculate the amplitudes of the reflected wave, but is this relevant in any way, shape or form?
Ah, gents..I have been using the incorrect terminology for Cyril's appendix 2 writeup. Sorry about that, my bad..🙁
He mentions the use of the 8721A, but laments that it is not "easily switchable" for other impedances.
He mentions a vector network analyzer by Steve Dunbar in Electronic Design May 29, 2000, but that it absorbs at least 6 dB...
He then goes on at length describing a two transformer directional coupler. He also used a two gang switch to select the measurement impedance, with choices of 15, 25, 30, 39, 50, 62, 75, and 100 ohms.
Sorry for the confusion..
Jn
He mentions the use of the 8721A, but laments that it is not "easily switchable" for other impedances.
He mentions a vector network analyzer by Steve Dunbar in Electronic Design May 29, 2000, but that it absorbs at least 6 dB...
He then goes on at length describing a two transformer directional coupler. He also used a two gang switch to select the measurement impedance, with choices of 15, 25, 30, 39, 50, 62, 75, and 100 ohms.
Sorry for the confusion..
Jn
A picture of a green phosphor screen, and you are going to gig him for no time and voltage division? Really? Getting a little too picky there aren't we?
Would you really believe that Cyril Bateman would pull a bait and switch on published articles?
When I see a green phosphor scope display showing exactly one period of a sine wave occurring in 10 horizontal boxes, and a text caption which clearly states the frequency used, I don't even have to take my shoes off to do that calculation.
Also, his 1997 article does indeed show timebase numbers as well as stated frequency. granted, the numbers are messy down there, but he is showing 50 uSec per box, almost a full period on the display, and a statement of 1786 hz drive. While that math requires me to take my shoes off, it's still not difficult.
jn
Have you an ETA? I need an idea when to order my catering size bag of popcorn
hello..... T,
My question to my self was this --- we test into resistive loads for FR and THD/IM. Then we listen with a speaker connected. We could test with a speaker attached, I suppose. But there is no standard speaker. What would it sound like if the same speaker was like a resistor which could reproduce sound? Does the particular power amp sound different loaded as an almost pure R than a reactive speaker?
In other words, do these 2 Z curves sound different on any particular PA?
View attachment SPKR Damping R.pdf
Using an RC network only flattens some of the freq range. The shunt/parallel R flattens the whole range.
I didnt say this much about it because it is so easy for anyone to try and then listen.
THx-RNMarsh
Last edited:
It's too real, too complex and labor intensive to pursuit speakers (& room acoustics) for real audible improvement so some will stick to tweaking things that are much more layman oriented like cables, DACs and amps thinking that there is audible improvement to be had because they've been conditioned by marketeers to believe that way.
- Status
- Not open for further replies.