See this is where you went wrong.
Exact factual reality is perfectly real, and may I add, measurable thus factual.
Magura
I don't think so! Be careful what you measure!
Jan,
With the right tools and approach you probably can measure a difference in just about any physical product. Aside from measuring things that should not matter such as color (Yes color alters perception) even on something such as THD there may be a difference, even if it is below any reasonable expectation to influence perception.
Why people get hot and bothered over what somebody else hears is an interesting exercise. So I will not bother to comment on the semantics and general argument on what is reality. That is even more involved than phantom audio.
I've said all this before and will continue my practice of actually measuring things. Every so often I get whacked over the head with what shows up versus what I would have expected.
For example power cords, seems ridiculous a power cord should make a difference. But when I measure the impedance of my AC line it stays well below .1 ohm. Electrical code allows for a 5% voltage drop from the circuit breaker panel to the outlet at full load. So a 20 amp branch circuit could be as high as .3 ohms. I of course being cheap have wired my outlets with a heavier gauge wire than code requires so I don't have to spend as much on electricity. Also my runs are longer than you would find in a house. So the range of .1 to .15 ohms is probably typical with .3 ohms worst case.
When I plug an IEC power cord that came with some audio equipment into a Leviton midgrade outlet and also into an IEC inlet from Sprague filter (Surplus! 1000uh x 2) the reistance measures .179 ohm! So just the power cord connection provides more than 1/2 of the total power line loss.
I don't need to tell you that since a power supply only charges at peak voltage the current draw is often 10 times what the average current would be. Of course the effective line impedance is reduce by the internal power supply transformer. So allowing for a 30 volt transformer that would still be putting almost 1/2 ohm in series with my power supply just from the power cord! I know from previous postings that you understand this is corrected by the feedback in most but not all amplifiers designs.
As an interesting note switching amplifiers have the most problems with power supply variations.
So I'll continue to measure and be amused by uneducated comments.
ES
Last edited:
Yes, you have to be careful with that stuff. My uncle Mortimer forgot to take his meds one morning and died of an overdose.
What's that mean - "I am, therefore I argue"?
Seem's appropriate but which 10 people were you referring to specifically?
D'oh - did you have to mention that?
Scott,
I owe you these two measurements. One is of a Mills 5 watt resistor the other of a Dale 1/4 watt metal film type (Green Curve). Both are measured at the same power. So when common sense says to use a higher wattage resistor to minimize thermal distortion in a feedback resistor, sometimes it just doesn't work that way. Although across families of the same process it does seem to be more in line with common sense. But does not exactly follow expectations.
For those who haven't been reading old posts the tallest spike is the 1k test signal that has been nulled. The rest is either hum and noise pickup or harmonic distortions.
ES
I owe you these two measurements. One is of a Mills 5 watt resistor the other of a Dale 1/4 watt metal film type (Green Curve). Both are measured at the same power. So when common sense says to use a higher wattage resistor to minimize thermal distortion in a feedback resistor, sometimes it just doesn't work that way. Although across families of the same process it does seem to be more in line with common sense. But does not exactly follow expectations.
For those who haven't been reading old posts the tallest spike is the 1k test signal that has been nulled. The rest is either hum and noise pickup or harmonic distortions.
ES
Attachments
Last edited:
Moving on, back in the '50's and '60's, there was an audio marketing war, where AR won, so to speak. Large speakers, especially corner mounted ones (usually the best) were great for MONO, but lousy for STEREO. However, smaller speakers like the AR-1 or AR-3 (tweeter added), sounded lousy in MONO, but good in STEREO. Also, smaller speakers were more easily placed in the room at an optimum location for some effective STEREO EFFECT, and not just a ping-pong effect. There was an effective marketing effort by AR that literally convinced the public that they had made a 'breakthrough' and there was no other convincing way to generate deep bass.
Let us remove some of the marketing hype that still lingers. The AR speakers did not go super deep and they were not a very clean loudspeaker. They just were designed to have more cone movement than many other similarly priced speakers. This LOWERED the efficiency, and increased the instability of the speaker. It also increased the FM distortion drastically, although the AM distortion was reduced. Some tradeoff!
Also, their patent was broken by Harry Olsen, showing prior art, so it wasn't an exclusive 'break-through' either. Please remember, I lived with this loudspeaker for years, believed in the propaganda, had heard the live-recorded demo at GCS in NY, and had used the speaker in 4 separate distinctly different rooms over a period of time.
I did add a tweeter onto my AR-1, but it was a Bozak (sp) 2 speaker tweeter, that I found more subjectively pleasing at the time. AR did offer a separate tweeter to 'update' the AR-1, but I found it harsh and annoying, so I went looking elsewhere. I think that the AR marketing was the most effective million dollar rip-off in my audio history, and I swallowed it, 'hook, line, and sinker', for a few years.
Let us remove some of the marketing hype that still lingers. The AR speakers did not go super deep and they were not a very clean loudspeaker. They just were designed to have more cone movement than many other similarly priced speakers. This LOWERED the efficiency, and increased the instability of the speaker. It also increased the FM distortion drastically, although the AM distortion was reduced. Some tradeoff!
Also, their patent was broken by Harry Olsen, showing prior art, so it wasn't an exclusive 'break-through' either. Please remember, I lived with this loudspeaker for years, believed in the propaganda, had heard the live-recorded demo at GCS in NY, and had used the speaker in 4 separate distinctly different rooms over a period of time.
I did add a tweeter onto my AR-1, but it was a Bozak (sp) 2 speaker tweeter, that I found more subjectively pleasing at the time. AR did offer a separate tweeter to 'update' the AR-1, but I found it harsh and annoying, so I went looking elsewhere. I think that the AR marketing was the most effective million dollar rip-off in my audio history, and I swallowed it, 'hook, line, and sinker', for a few years.
Last edited:
Again Ed your logic escapes me. You compare resistors that might well have >100x difference in TC, who knows. I would not have any expectations beyond measuring the TC computiing the thermal properties of the resistor and making some kind of prediction as to what I should measure and then comparing an actual measurement. I have postulated several experiments to try and separate out the TC from true voltage coefficient of resistance. Of course us engineers are accused of having no curiosity.
BTW looking at that Dale plot you would have a hard time making an arguement for naked Vishays.
It was Ed Villchur and your buddy Mitch Cotter might have some good things to say about him.
And to compound the TC issue, the last few decades they have been increasing the power rating of all resistors by increasing the alumina content, making them all run hotter.
And a separate issue with mercury wetted contacts is the all the metals and alloys used inside the glass capsule MUST have low solubiity with mercury, that leaves a small choice of contact materials, usually Iron contacts...
Wrinkle
Scott
We are not disagreeing. I don't understand your comment, do you think I am trying to provide misleading data? I ain't selling nothing. I picked the Mills 5 watt as that is touted as an audio resistor. 5 watt metal film resistors do exist but are not as available. I ordered a bunch of resistors sold as either audio or precision and tested them. So from my data set I showed that power rating alone is not a guarantee of lower distortion. And why use a 5 watt resistor when a 1/4 watt will do?
The Mills resistor is rated at 50ppm as was that Dale sample, so it was not apples to bananas. The interesting issue with the Mills was the third harmonic distortion. The voltage distortion does not change with frequency as does the thermal. Since I have not yet found in my limited testing any resistor that has second harmonic distortion not following predicted thermal changes, it may be the voltage effects show up better as the third harmonic levels. Don't know for sure, but third is certainly an interesting case.
The problem with just looking at tempco is that you don't really know thermal mass. (For those who don't understand this, a high thermal mass would take longer to heat and cool so the distortion would be less. This shows up in the previously posted 19 & 20khz IM distortion plots.) I would have expected a wirewound to have more than a film type, but the measurements show less distortion with the metal film.
So even with identical rated tempcos there are still significant differences in the thermal distortion as well as other distortions!
As to the Dale working as well as the naked Vishays, certainly seems that way, worse yet there is a 1.6 cent resistor that measures much better than many others!
ES
Paul Klipsch had his own speaker design philosophy, and it was 'flawed' as well, but John Meyer and I did not catch it, for years. We did read Harry Olson and Beranek, in fact, these were virtually reference 'bibles' for us, when we designed speakers for PA reproduction, which both of us did for years.
Klipsch's FM distortion WAS important, but HORN THROAT DISTORTION, and path length mismatch were important too! For home operation, horn throat distortion is a SMALL contributor, but for PA, it is AWFUL! This is why Klipsch got away with cheap drivers and small throats (1/2") for his K-horn. In the early '50's, Klipsch did a darn good job, but later, in the '60's, in order to keep costs down, he compromised a bit too much.
more later
Klipsch's FM distortion WAS important, but HORN THROAT DISTORTION, and path length mismatch were important too! For home operation, horn throat distortion is a SMALL contributor, but for PA, it is AWFUL! This is why Klipsch got away with cheap drivers and small throats (1/2") for his K-horn. In the early '50's, Klipsch did a darn good job, but later, in the '60's, in order to keep costs down, he compromised a bit too much.
more later
The biggest problem with the K-horn was the effective acoustic path length between the drivers. It was about 4 feet between the woofer and the mid, and maybe 1 1/2 feet between the mid and the tweeter. With large musical events such as symphonies, it was OK, but with human voice, it was not so good. Klipsch maintained, from work in the 1930's and perhaps '40's that the ear was deaf to this, but it just didn't wash. more later
Sorry Ed it's hard to have a discussion without all the information. It's perfectly possible for a resistor to have a voltage coefficient, that's just part of the equation. You still won't convince me these sub-ppm effects are audible in a controlled test. In this case you show <-150dB thirds.
Last edited:
Scott,
I do count on you to poke holes, as a fellow professional, I expect that. There are more than enough dull wits in the world.
I don't recall saying that -150db of third harmonic can be heard. I have said that given two parts costing the same it would be silly to use the one with higher distortion. In a system using -80db parts where the harmonic do add coherently at some point that would directly be a problem.
In a musical reproduction system I expect -20 of second harmonic may be heard, possibly -30. Of third I think it is in the same range. So if you could manage to put 300 or so not very good resistors in the signal path....
What I do think may be heard is when very high energy signals such as record warp, power supply bounce, mechanically induced low frequencies, EMI and even extremely "Hot" low frequency musical notes interact through Intermodulation Distortion can cause artifacts that appear in the frequency ranges our ears are most likely to notice. That is why I am for very low distortion levels in my designs. I can control many of the noise sources but not all of them.
I don't dismiss some of the claims of what people can hear, but I think we both often laugh at the explanations.
One of my early expensive lessons was with reproducing speech. At 20 db signal to noise ratio and a low reverb time (say 1 second) speech recognition is excellent. However in a room with a noise level of typically 30 db (NC curve) and a signal to noise ratio of 30 db, the sound reinforcement system will be considered inadequate. That because when the sound level drops below 70 db "A" weighted fast reading, people start to fall asleep. Or in other words things you didn't look at may bite you in the...
Enjoy the holidays!
ES
"There was an effective marketing effort by AR that literally convinced the public that they had made a 'breakthrough' and there was no other convincing way to generate deep bass."
Actually that's not what Villchur claimed. What he said was that to build a horn speaker that would equal AR1 in low frequency capability, it would have to be at least 40 feet long. For speakers that rely on resonant air columns for deep bass, the LF cutoff is related to the dimensions of the air column. Klipschorn is one of the most efficient speaker I know of. I think it will produce around 95 or 97 db from 1 watt at 3 feet. Many other speakers today and AR3 was among them were around 83 or 86 db. That's around 10to 20 times as efficient.
The folded horn was a large complex cabinet to build. It was big and heavy and expensive. The AS design only required a sturdy 2 cubic foot sealed simple box to trap the air. The Klipschorn design also relied the corner for optimal placement, the side and back walls being extensions of the horn itself. There are actually considerable advantages to corner placement. Bass reinforcement is only one of them. If the goal is maximum usable listening space in a room with uniform high frequency coverage, seemingly the opposite of the goal today, corner placement is best, you can't get more than 45 degrees off the tweeter's axis.
We've disputed whether or not Olson's claim on Villchur's patent was valid or not. I'm not going to choose sides, I didn't read Olson's work. There are also some who claim Villchur didn't invent the dome tweeter but that seems to be accepted by many. I think it's also widely accepted that AR was the inventor of the ferrofluid cooled tweeter.
"The AR speakers did not go super deep and they were not a very clean loudspeaker."
AR1/3/3a et al. were 3db down at 42 hz with output falling at 12 db per ocatve and 5% THD at 30 hz. That was champion class for many years and even today is better than many so called subwoofers. But amplifier power was expensive and limited. The most powerful consumer amplifier I'm aware of until the Crown DC 300 came along was the McIntosh MC275. I'm not sure when the 300 watt MC3000 monophonic amplifier appeared. The Riverside church demo used 4 AR1s and 4 150 watt Western Electric amplifiers.
"I did add a tweeter onto my AR-1, but it was a Bozak (sp) 2 speaker tweeter, that I found more subjectively pleasing at the time."
This was not a good tweeter. As I recall it had a nasty peak at around 2 khz and was very beamy. Rudy Bozak himself used them in a line array in his large speakers like the Concert Grand. Before AR3, many tried JantZen electrostatic tweeters which were beamy too. Others tried the Microstatic 4 tweeter array. I don't think AR ever got it just right. The high frequencies always sounded muffled to my ears. No amount of rationalization can change that impression. However, equalization may be able to mitigate it to a large extent. Limited power handling was another shortcoming of AR tweeters. They were blown up easily.
"Moving on, back in the '50's and '60's, there was an audio marketing war, where AR won, so to speak."
Yes they did. I think they captured something like 27% of the component hi fi speaker market at one point. Nobody else I think has ever come close.
" I think that the AR marketing was the most effective million dollar rip-off in my audio history, and I swallowed it, 'hook, line, and sinker', for a few years."
And here I thought you once said Bose 901 was the biggest ripoff.
Resistor distortion
If you guys wants to measure some real resistor distortion just for the holidays, then make a voltage divider with a LDR/LED combo. Expect THD in the 0.1% to 1%. region.
I used LDR's for VCA's in limiters, voice-over circuits, crossfaders etc back in the eighties. But not in critical applications.. And every time someone attached a THD-meter, i just closed my eyes
If my memory serves me right, I think Audio Research used one in series with the line-output in one of their preamps. Nasty!
Can someone explain to me why LDR's still are used in audio gear?
If you guys wants to measure some real resistor distortion just for the holidays, then make a voltage divider with a LDR/LED combo. Expect THD in the 0.1% to 1%. region.
I used LDR's for VCA's in limiters, voice-over circuits, crossfaders etc back in the eighties. But not in critical applications..
And every time someone attached a THD-meter, i just closed my eyes If my memory serves me right, I think Audio Research used one in series with the line-output in one of their preamps. Nasty!
Can someone explain to me why LDR's still are used in audio gear?
Well, your numbers are somewhat off, the K-horn had a specified sensitivity of 104spl at 1W. We have not gotten to Bose yet, how can I?
It is true that you could put a subwoofer under a K-horn and hear an extra note played with modern (at the time) direct disc records. I did it myself. However, Klipsch, being from the old school, clung to his opinion that there was no bass below the pass band of his k-horn to amplify. I could agree with him, if it were 1950.
It is true that you could put a subwoofer under a K-horn and hear an extra note played with modern (at the time) direct disc records. I did it myself. However, Klipsch, being from the old school, clung to his opinion that there was no bass below the pass band of his k-horn to amplify. I could agree with him, if it were 1950.
- Status
- Not open for further replies.