Rare earth/metal doped spark plugs are not unique to DS. Dick also has a patent that would contain more details about his claims. They are certainly less far fetched.
EDIT - US 5898257 Data presented using a 1995 MAZDA.
Last edited:
And at a mere $10 per plug, they're downright cheap. You could replug an eight banger for less than a single Bybee.
se
I am not so sure -135 is inaudible. Fletcher Munson is based on steady tone levels, allow 20 db or so for peak to average, factor in 10 db or so for above average human skill and you would want to design for at least -165 to -170db!
Also the distribution is not normal. The certainly is no zero and no infinity, there are lots more at -120 than -150.
Based on what evidence? All controlled tests (that I have seen) for audibility of any number of distortions had thresholds orders of magnitude higher. The sorting out of resistors by GEB members in uncontrolled tests does not cut it.
The test equipment (ancient dinosaur of a BMI, couldn't sell one if I tried) has already been determined to be unsuitable for high frequency capture. It was the equivalent of a 10 kHz DSO. I think SY's equipment is a tad better than that.
Additionally, SY's (preliminary) measurements of the device would indicate the test Jack made way back when was in error, and I will guarantee you one of two things:
1. The test Jack performed cannot be repeated.
2. The test procedure would not be accepted by peer review.
Take your pick.
If I am not mistaken, SY used the HP3581C that I loaned him 10 years ago and ultimately he bought it from me for 1/2 the price that I got it through a wholesaler of used test equipment. I tried to measure a Bybee device, about the same way with my HP3581A and my HP3580 and I failed to get results either. So, it would take better test equipment than I have to measure it. However the BMI was in use, allegedly at AMES RESEARCH LABS, and it was calibrated and had a date code associated with the test report. So be it.
What bothers me is that my judgement as to what constitutes a measurement is impugned by SY. For the record, I was making measurements on test equipment referenced to the National Bureau of Standards, 48 years ago, when I was a test technician for UL in Santa Clara, Ca. I made hundreds of measurements each day, of temperature, with thermocouples, power, voltage, and current readings with calibrated WESTON test equipment. I kept my Heathkit at home. Then, when I was Ampex, 6 years later, I made hundreds of measurements with HP and GR test equipment for distortion, frequency response, noise, Q, and frequency stability over temperature.
To imply that I would NOT question a test result that looked 'faked' is an insult.
What bothers me is that my judgement as to what constitutes a measurement is impugned by SY. For the record, I was making measurements on test equipment referenced to the National Bureau of Standards, 48 years ago, when I was a test technician for UL in Santa Clara, Ca. I made hundreds of measurements each day, of temperature, with thermocouples, power, voltage, and current readings with calibrated WESTON test equipment. I kept my Heathkit at home. Then, when I was Ampex, 6 years later, I made hundreds of measurements with HP and GR test equipment for distortion, frequency response, noise, Q, and frequency stability over temperature.
To imply that I would NOT question a test result that looked 'faked' is an insult.
I wouldn't say that. I have no idea what Jack's test was, how it was set up, what the error bars are, what was the DUT, how it was connected, nothing. We know the brand of the test instrument, that's all. So I can't say it was right or wrong, I can just say that I saw no difference in ripple current waveforms (which was what the second-hand graph "waveforms" looked like) between the reference resistor and the Bybee.
I'm not aware of anyone claiming the result was faked John - quit the injured pride routine.
The problem is that there is no supporting documentation for it - nothing to indicate how, when why etc.
That makes it a set of pretty lines, nothing more or less.
None!
I assume everyone understands we are talking about a level related type of distortion such as T.H.D. not one that becomes more significant as level drops such as crossover notch.
The model I have generally used assumes that our hearing mechanism is two stage, the first stage is an overall volume control with a time constant averaging levels. The second stage is fast acting and resolves the fine details. I allow 1/2 of the total dynamic range the plus 10db to each. The model has worked fine for my design use. It however has no basis in any research.
What is my motivation is complaints. I really do get 22,000 to 85,000 people at a time listening to my work. As a result what I look for is quite different than what most folks at this site consider.
I can tell you a Peavey Media Matrix DSP unit has enough distortion that even in an NFL stadium it sticks out! With the ambient noise level compared to the peak reproduced music levels there is less than 70db of slow weighted average SPL range. I haven't looked at the specs, but I am sure the device is much better than that.
In an NBA arena during construction, I have replaced the brand new DSP box with analog and even the construction managers noted the improvement. It was a new design DSP (Not your chips!) and was enough of a problem I spent my own money on the replacement. BTY to adjust the DSP system one could sit in a comfortable location, for the replacement you had to ride the cluster beams up into operating position, dangerous and much more difficult.
So when I set design goals I want to be at least 10db better or more than what can or has been a problem.
Please also note that average levels and peak levels are quite different. So -100 average could be -40 to -60 peak. Just throwing around numbers with out reference detail can cause a lot of yelling and screaming without useful content. So I apologize for any confusion.
Last edited:
- Status
- Not open for further replies.