Forums :: Resources :: Features :: Photo Gallery :: Vintage Radio Shows :: Archives
Support This Site: Contributors :: Advertise


It is currently Jul Tue 23, 2019 2:11 pm


All times are UTC [ DST ]





Post New Topic Post Reply  [ 25 posts ]  Go to page Previous  1, 2
Author Message
 Post subject: Re: Questions about Laplace transforms in LTSpice.
PostPosted: Mar Fri 29, 2019 2:59 pm 
Member

Joined: Oct Thu 04, 2018 2:11 pm
Posts: 106
Location: Suburban Chicago
Flipperhome wrote:

Look at these two scope traces. I bet you'd call both 'near perfect' but the second has 5% second harmonic distortion.


Try it again with a third harmonic. Your eyes will have no trouble with that! I used to spend endless hours testing two way radios. I forget what the THD spec on them was but it was well below 5%. I could set up a radio and start cranking up the volume into a dummy load and monitor the signal on an oscilloscope. When I stopped and told you it was at the spec limit you could measure it with a distortion meter and I would be right. Your eyes can be very good at seeing certain kinds of distortion and you ears can be blind to certain kinds.

If you asked me to whistle a 1kHz tone while you measured it with a frequency counter I could do that too!

How much distortion do modern speakers cause? Do they get much cleaner at low levels? I honestly don't know, just asking. The point is that determining the minimum distortion that a human ear can detect is one thing. Determining the minimum amplifier distortion that the human ear can detect when the speaker is at 1% distortion is a much, much harder task. Speakers today may or may not (again, I wouldn't know) be better than those of the 1940s and 1950s but speakers from that era had their limitations and so one should not be overly critical of Olson or other workers from the era if they were unable to nail down a number for the distortion limit. They could only determine a limit to the extent that the equipment they had allowed. That remains true today whether our equipment is significantly better or not.

Laboratory experiments have shown that under some conditions a human eye can detect an event consisting of a single photon of light a small percentage of the time. That does have some bearing on what an amateur astronomer, for example, can detect at the eyepiece of a small telescope. I once made a determined effort to see the planet (or not!) Pluto with a telescope only slightly larger than what is considered the minimum for its brightness at the time. Armed with a printed chart that showed where Pluto was in relation to the surrounding stars of similar and greater brightness I was able to see it. Both Pluto and the dimmest stars on my chart danced in and out of view. I would see them sometimes but not others. But I was sure I was seeing it because when it appeared it was always right where it was supposed to be. So the same sort of thing (a statistical detection, not a steady view) was going on but the brightness of Pluto that evening in that scope was orders of magnitude greater than a single photon!

Your ears and eyes can both do quite amazing things and yet at times they can fail to do things that seem like they would be rather simple. And you will never sell a zero distortion amp to a guitarist. Some bassists would buy one, others would not....


Top
 Profile  
 
 Post subject: Re: Questions about Laplace transforms in LTSpice.
PostPosted: Mar Fri 29, 2019 4:25 pm 
Member

Joined: Oct Thu 04, 2018 2:11 pm
Posts: 106
Location: Suburban Chicago
Harry Olson, eh? The name doesn't ring a bell but the description of his work above and that found in ]the Wikipedia article about him:

https://en.wikipedia.org/wiki/Harry_F._Olson

remind me of something I read years ago in, I believe, Scientific American. The article was discussing how FM broadcasting came to have a 15kHz BW. It claimed that when the FCC was devising the standards for FM there was a strong contingent that though that a 5kHz audio BW was sufficient. It would allow for more stations in any given band devoted for FM broadcasting and consumers preferred the narrower bandwidth anyway. They had experimental data to prove that. They would set up a "radio" which was really just a loudspeaker for an audio amplifier in a listening room and ask for the listener's reactions as they varied the BW of the audio chain. Virtually everyone preferred a BW of 5kHz or below.

The 15kHz supporters were on the verge of losing the argument when they hit upon an experiment that from its nature must have either involved Harry Olson, directly or by knowledge of and inspiration from his work. The article may have mentioned his name in fact but I don't recall that after all these years.

What they did was to conduct an experiment much like the one described in the Wikipedia article with one difference. They had a large room divided in two by a partition. On one side was a small orchestra (my memory says it was a jazz orchestra) and the other side was a listening area for the test subjects. The two were connected by the type of acoustic filter that Olson had invented. But instead of an opaque divider between the two halves as the Wikipedia article describes, the top half of the divider was glass. The audience could see that they were listening to a live orchestra. The results were rather stark.

The audiences preferred a full audio BW by a wide margin when they could see that the music was coming from a live jazz orchestra. Yet when the same audience listened to similar material coming from a speaker that looked like an AM radio they preferred the narrow BW! The conclusion seems to have been that people of the day were letting their experience driven expectations of the "radio sound" color their natural inclination to prefer full BW audio. They preferred the narrow bandwidth of the "radio" because it matched what they were accustomed by long experience to expect from a radio. When they could see that the source of the music was a live band they preferred the BW they would get when they went to the club on Saturday night to listen to some cool jazz.

According to the article it was this experiment that convinced the FCC that the radio bandwidth wasted by using a 15kHz audio bandwidth on the transmitted signal was not wasted bandwidth at all. I wish I could point everyone to the article. All I can remember is that I think it was in Scientific American and I think it would have been published some time between 1974 and 2000.


Top
 Profile  
 
 Post subject: Re: Questions about Laplace transforms in LTSpice.
PostPosted: Mar Sat 30, 2019 3:51 am 
Member
User avatar

Joined: Nov Sat 26, 2011 4:09 am
Posts: 9518
Location: Texas. USA
khutch wrote:
Flipperhome wrote:

Look at these two scope traces. I bet you'd call both 'near perfect' but the second has 5% second harmonic distortion.


Try it again with a third harmonic. Your eyes will have no trouble with that! I used to spend endless hours testing two way radios. I forget what the THD spec on them was but it was well below 5%. I could set up a radio and start cranking up the volume into a dummy load and monitor the signal on an oscilloscope. When I stopped and told you it was at the spec limit you could measure it with a distortion meter and I would be right. Your eyes can be very good at seeing certain kinds of distortion and you ears can be blind to certain kinds.....
Okay. you asked for it. This is 5% third harmonic distortion.

Attachment:
h3.gif
h3.gif [ 3.38 KiB | Viewed 139 times ]

Now, for someone with endless hours of experience and 'super sensitive' eyes like you its pretty obvious but for the casual scope observer I think they'd call that "near perfect" as well, especially when its all on its own with no 'perfect' sine wave to compare against.


Top
 Profile  
 
 Post subject: Re: Questions about Laplace transforms in LTSpice.
PostPosted: Mar Sun 31, 2019 3:05 pm 
Member

Joined: Jan Thu 01, 1970 1:00 am
Posts: 286
Location: Arlington, TX, USA
Flipperhome wrote:
khutch wrote:
Flipperhome wrote:

Look at these two scope traces. I bet you'd call both 'near perfect' but the second has 5% second harmonic distortion.


Try it again with a third harmonic. Your eyes will have no trouble with that! I used to spend endless hours testing two way radios. I forget what the THD spec on them was but it was well below 5%. I could set up a radio and start cranking up the volume into a dummy load and monitor the signal on an oscilloscope. When I stopped and told you it was at the spec limit you could measure it with a distortion meter and I would be right. Your eyes can be very good at seeing certain kinds of distortion and you ears can be blind to certain kinds.....
Okay. you asked for it. This is 5% third harmonic distortion.

[picture snipped]

Now, for someone with endless hours of experience and 'super sensitive' eyes like you its pretty obvious but for the casual scope observer I think they'd call that "near perfect" as well, especially when its all on its own with no 'perfect' sine wave to compare against.


Agree completely.

That third harmonic distortion waveform looks like a normal sinewave to me even though it actually has 5% third harmonic added. I'd only be able to tell the difference with the undistorted original right next to it for comparison, and the difference would still be slight. In fact I'd wager that a scope display with even order distortion present in the signal would be easier to see since it affects the sinewave peaks unequally.

One of my electronics instructors told us in class that the oscilloscope is a useful tool but it is not a precision measurement device.


Top
 Profile  
 
 Post subject: Re: Questions about Laplace transforms in LTSpice.
PostPosted: Apr Mon 01, 2019 9:31 pm 
Member

Joined: Apr Tue 17, 2012 7:58 pm
Posts: 113
Location: Metamora MI, 48455
benman94 wrote:
I've looked at some other references and they seem to align very closely with what you've suggested. One source said that the lowest frequency cut during mastering would be about 40 Hz for the reasons you described...


I've been doing more digging and discovered that Victor and Columbia both adopted 50 Hz shelving standards at some point for their 78s. In fact, both Columbia and Victor manufactured test tone discs covering 50 Hz to 10,000 Hz for use by manufacturers in the design of equipment. They considered a 78 to have a useful bandwidth of 9.95 kHz, from 50 Hz to 10 kHz, and thus I'll be rolling off everything below 50 Hz and above 10 kHz.


Top
 Profile  
 
Post New Topic Post Reply  [ 25 posts ]  Go to page Previous  1, 2

All times are UTC [ DST ]


Who is online

Users browsing this forum: No registered users and 4 guests



Search for:
Jump to:  


























Privacy Policy :: Powered by phpBB