We sure as heck don't have a perfect speaker. How easy can it get? Just some wire, a dixie cup and a magnet, or so said Mr. Rice and Kellog. Was the DC-3 a perfect airplane? A Morgan Plus 4 the perfect car? Bic the perfect pen? AK-47 the perfect gun? Rhodes 19 the perfect daysailer? Well, maybe. Anchor Porter the perfect beer? Almost. We don't even have the perfect beer glass. Maybe you could point to a Corell dinner plate and say to yourself, what could I change to make it better?
On-Semi has not created the perfect transistor as far as I can tell. That leaves the esteemed gentlemen mentioned here with room to continue. What got me fascinated is what looks deceptively simple is really difficult to do very well. I am glad human nature drives some to exceed what is minimally acceptable.
I am also really a big fan of the industrial engineering art. Combining great function with great style. Eames chair, Jag E type, and I have to admin it, iPad. Sorry to say, I can't list any high end audio product in that list. Some are "very nice" where most are something to hide unless you are looking for bragging rights. I kind of liked the original Aragon preamp. B&O tried. Speakers are either just plain ugly or weird. Tanburg Fasetts fit into a few decors in the 70's, a pair on my desk right now, but would not fit in most rooms today. Worth a study. Maybe one of the first molded speakers. Very listenable and made of quite ordinary drivers. Good design.
I'm building one, but of course it will not be very loud. More like a Whisper from Nirvana.
Attachments
Very hard compared to the electronics (which are quite easy). Marketing and product differentiation are the challenge in electronics. Speakers are a whole different issue.
Confirm Avago (which was HP). They have kept the "standard red" as a product, which afaik is the best match to a silicon Vbe tempco. They are very quiet.
LED "lifetime" used to just be about the light output decline. I'm sad to hear that you can make them shoddily enough to have them degrade in days!
I think I've already told the story somewhere within diyaudio about my object lesson in signing up for a change before the actual specification was revealed, and then having to do some fancy design to come in at the budget or near. I wound up using an LED (an HP part) to bias a common-base stage level shifter. Everything went well until they bent the leads before wave solder (which they did not have to do, as there were holes on the PCB into which they fit perfectly).
It used to be conjectured that audio quality of recordings and musical performance quality were anticorrelated. I wonder if there is a comparable inverse relationship effect for audio equipment ID and sound quality?
At least tvr, you are LEARNING how difficult it is to make virtually 'perfect' electronics. You are fortunate that your wife keeps you on track, because your mind and your measurements would certainly throw you off.
With SY, he believes that the differences are an illusion, and he can PROVE this to both you and your wife, if you subject yourselves to his testing. I, however, tried this sort of 'testing' myself about 1/3 century ago, and came up with a conclusion: Either I am crazy, or the test is flawed in a serious way. Therefore I ignore ABX type double blind tests, and their near relatives. Seems to work for me! Holman and Spigel went the other way and dropped out of hi fi design. That was their choice, after subjecting themselves to ABX testing. What do you want to do?
With SY, he believes that the differences are an illusion, and he can PROVE this to both you and your wife, if you subject yourselves to his testing. I, however, tried this sort of 'testing' myself about 1/3 century ago, and came up with a conclusion: Either I am crazy, or the test is flawed in a serious way. Therefore I ignore ABX type double blind tests, and their near relatives. Seems to work for me! Holman and Spigel went the other way and dropped out of hi fi design. That was their choice, after subjecting themselves to ABX testing. What do you want to do?
Very hard. That's the point. I even built one once with no moving parts. It used the corona effect. It actually made sound, along with a lot of ozone.
Besides just being hard, they are only one part of a system including the room and the amp. Part of why the amp is so hard as it has to be built without knowing the speaker.
Besides just being hard, they are only one part of a system including the room and the amp. Part of why the amp is so hard as it has to be built without knowing the speaker.
Not clear to me what you mean by "back-to-back". Schematic perhaps?
The advantage of LED bias is (1) the tempco match (2) the extra voltage developed available to the emitter resistor, and (3) the low voltage noise of the LED and low impedance, which means you can have reasonable power supply rejection.
Disadvantage: the amount of available voltage means that you are limited as to how low the noise in the output current can be --- the resistor can only be so large. So transistor voltage noise modulates the current, and the thermal noise current of the resistor is significant. However, in most applications these are not significant noise sources.
Another supposed disadvantage is the light sensitivity of the LED. If this worries one, put a hood over it. The efficacy as a photodiode is poor anyway, and represents a very small fraction of the mA or so you should be putting through it as bias. Bluer parts are quite a bit less sensitive to light, but I'm not sure how the noise or tempco is. But again, I think this is generally a silly objection. And besides, it's kind of cool to see them glow --- why should hollow-state have all the fun?
I take your point, but really, that Citation (?) amp from 1980 that everyone seems keen on, doesn't seem much different from today's money-no-object over-the-top box with fins. And, as a relatively disinterested (but not uninterested) observer, reading the posts here, it often seems that the biggest excitement comprises moving one end of a cap from the base of pre-ordained-function transistor X to the junction of pre-ordained-function resistor Y and pre-ordained-function capacitor Z. And even then, there always seems to be a potential downside, but measurements are not deemed capable of ending the argument - strangely enough.
That, or just using even heftier or more polished pieces of metal for their psychological effect.
It would be nice to feel a sense of progress, or to simply declare that a certain design of amplifier is just about good enough, so we could all build or buy one without constantly worrying about whether we just heard some transient intermodulation that can't be measured meaningfully, apparently.
Or do you amplifier designers see yourselves more like craftsmen (e.g. violin makers) making the same thing over and over again, with artistic flourishes and embellishments to keep it interesting, but no ambition to ever make measurable 'progress'?
No kidding. I love classical guitar and the best performances are really poor recordings. ( groveling for recommendations) All time low mark: 461 Ocean Blvd.
If you look at Mr. Curl's designs and Mr. Pass's designs, you will see very different directions. That a lot of amps look very "Self" is pretty basic and DS tells us why. For the available parts, it is the good enough design. What that means is to do better is hard. It is small things.
I think you are reading the luthiers wrong. I think they all were, and still, work for more power, more music, better. Many advances from the sound post to the iron piano frame.
From an engineering standpoint, no. From a marketing standpoint, certainly.
Unfortunately, you are 'clueless' Coppertop. The Citation XX was high end, the best that could be made at the time. You just cannot see the differences, just like some people can't note differences in automobiles. After all, they all have 4 wheels, are steered by a driver, have lights on the front and back, and drive the same roads as everybody else. I guess, every car is just about the same as any other car. '-)
With the last sentence in mind I wonder why active speakers (in which each amplifier can be designed for a specific driver, and the crossover done at line level instead of for high power) aren't as popular among hifi and especially high end as they are in recording studios. Is this due to the same bias for "separates" and against "integrated receivers" that has existed since probably the dawn of hifi? There's often enough truth in integrated hifi's being inferior, but it seems to be more because of designing to a price point than negative effects of putting more stuff in one box. Active speakers have lots of advantages over separate amp-and-speakers (with the rather trivial exception of each speaker needing both audio and power cords). Why haven't I seen more of these compared to separate amp-and-speakers, or even generic (not designed specifically for the speaker it's connected to) monoblocks-and-speakers?
At least in part, because it removes a degree of freedom from the customer. People want to play with their equipment.
I auditioned a massive home theater system prototype recently. The designers and the investor were proud of their accomplishment, and it did sound pretty impressive in many ways. Essentially unlimited and ~undistorted SPL, among other attributes (I quickly regretted not bringing earplugs).
But the amps were stated to be part of the package. And they were things that however decent in measurement terms, would be sneered at by many audiophiles. And the cost of a fully-loaded system was 1.4 million dollars.
Now, there are a few who would take the package and enjoy the system. But. Most would want to have a hand in its constitution. That would include amplifiers.
And to make matters worse: the venue was a surplus dealer near the Van Nuys Airport. You had to walk through piles of great junk (to many of us, gorgeous old equipment and parts) to get to the listening room. How many billionaires would be able to see past that, and imagine what this system would look and sound like in one of their sprawling mansions?
Precious few.
- Status
- Not open for further replies.