Of course. It would be a hard work to design a bass reflex and chose an accurate speaker without knowledge of Helmholtz Resonance and how to interpret Thiele and Small parameters... But this knowledge is not sufficient to get the optimal proportions of the enclosure, best internal damping and how to reduce vibrations of the walls of the enclosure. Not to talk about WAF factor ;-)
But definitely not a result from a random walk. Empirically doesn't mean that there was no clear idea in which direction to go, supported by experience/knowledge/science!
Jan
It was scarcely different; I will have to go back and look, or better remeasure with the cleaner setup. I just bagged it to get back to some other work.
There is reason to suppose that some resistive loading (at the input of the cascode that is, edit) of a given folded cascode arrangement might reduce even order. It can work well with a JFET and bipolar folded cascode, although the resistance gets noisily low.
Last edited:
Times have changed for amateur and professional speaker builders. There is a lot free or not expensive measurement and design software available to speaker makers these days as we can see just by rummaging around in Diyaudio.
This very interesting thread, which woke up after going to sleep in 2006, (I'm quite sure availability of of inexpensive SW is reason for thread's revival), is a very good example of their use:
http://www.diyaudio.com/forums/multi-way/88135-filler-driver-ala-b-o-5.html
One participant, xrk971, has been prototyping his enclosures using foam board - cheap, cheap, fast, fast.
Danley and Geddes make excellent speakers without access to anechoic chambers.
Earl Geddes has been posting information here for years on the importance and psychoacoustic effect of speaker/room interaction, on application of multiple subs and on psychoacoustic effect of time domain artifacts.
Some participants here have been posting about 3D printing of waveguides or their components.
Cheap DSP processing makes prototyping of crossovers and running cheap prototype speakers much easier than it was 10 years ago.
Very sophisticated, cheap resources that can be applied to speaker making are available at Diyaudio and other enthusiast sites.
I don't think designing a loudspeaker is any more difficult than designing electronics. But, rather, it needs different skill sets and knowledge.
But if you are designing an IC with a reticule set costing $1M plus you would sim the hell out of it to make sure it would work first time. Having access to the process designers to make your Sim accurate is of course a huge bonus compared to us great unwashed who have to limp along with the models that we are given.
But I agree with you that for us DIYers some iterative empirical work is the way to do it. It's also more fun as a hobby to lob up a breadboard to test a theory.
Aside: Has anyone got a link to one of Bob Pease's legendary rats nest prototypes?
Edit: found this pic of one of his less monumental efforts http://www.rfcafe.com/references/electrical/bob-pease-breadboard.htm
But I agree with you that for us DIYers some iterative empirical work is the way to do it. It's also more fun as a hobby to lob up a breadboard to test a theory.
Aside: Has anyone got a link to one of Bob Pease's legendary rats nest prototypes?
Edit: found this pic of one of his less monumental efforts http://www.rfcafe.com/references/electrical/bob-pease-breadboard.htm
Last edited:
Dahlquist knew what he was doing when he designed the DQ10. It was not a perfect speaker, but it removed the BOX from the equation, to a first order. The best listening example that I ever heard was Saul Marantz's personal pair in his home (Saul was president of Dahlquist for many years).
Jim Williams had the best, even threw some inside jokes in with text messages on the boards.
https://www.edn.com/design/analog/4411017/4/undefined
I don't know how Pease or Williams could even find their protos with these desks.
How messy is your desk? | EE Times
Last edited:
The Pease cover for Troubleshooting Analog Circuits shows such a breadboard. He used linkages to approximate actual metallization resistances, among other things, for circuits that were going to be integrated. The bench is actually pretty uncluttered by comparison with the one on the cover of one of the Williams compendia shown below.
I find that when I am really nervous about a circuit I procrastinate by cleaning the bench. Otherwise it does get pretty messy.
After the 1994 Northridge quake the Harman campus was fairly discombobulated, and we were admitted back in with hard hats strictly enforced. One person shot a picture of me at my desk pretending to be working, on the phone (which didn't work yet) and holding a Tonka truck as if it were associated with some automotive audio issue. The overhead cabinets in the cubicle were leaning over ominously, and lots of books and papers were all over the floor. Of course the gag going round was that my cubicle was strangely untouched by the temblor.
Attachments
I don't know about Mad Man Muntz (probably a salesman more than anything, c.f. Noel Lee?) but the Dahlquist story made some sense at the time.
Supposedly he had a background designing phased array antennas (probably true) and mapping from antenna radiation to speaker radiation is not that difficult a stretch. Further, understanding the real world effects of reflections etc. from antenna design would be a major benefit at the time since the whole issue was more black art than established design. The best knowledge at the time was from Harry F. Olsen and even RCA did not exploit it fully.
The important part of design is making sure you are solving the right problems. Its easy to focus on something like reducing distortion in an amp or flattening the on axis frequency response but if the end result doesn't do what the "customer" wanted you blew it.
Today's broader market customer seems to want a small single box wireless speaker and really has a limited understanding, if any, of the concept of stereo, otherwise why would a Sonos Play 1 or Amazon Echo sell at all? He/she is not interested in a stereo image (or bass or. . .) Pleasant, undistracting background music from streaming services is what sells. Our focus on reproducing a live experience seems odd in that context. What really spooked me was talking so some mid-30 tech workers and hearing that silence is uncomfortable. They were in need of music/noise to even sleep. At that stage the content matters little.
This sort of thing comes in waves of many years' period. Muzak was indispensable for a while. My father made an FM subcarrier receiver that was remarkably reliable even located in hot attics. It put a little food on the table for several years.
This too shall pass.
I mentioned this in the "bybee fraud" thread. That generation
was the first to have a constant corporate stream (MTV) beamed into
their brains. Not the 60's and 70's content , but the nagging - irritating
"post cable" era (80's). They need the constant din of media as they were
conditioned to it from birth.
They have never experienced the "contrast" of life before this time.
They accept lies as truth , junk as acceptable ...everything is disposable
and convenience oriented. (quality is a foreign concept)....
OS
There's a natural law concerning complex air-wire type prototypes. It involves a concept called 'terminal complexity', the most complex state any prototype can ever exist in and still work.
Terminal complexity arises when parts drop off or wiring faults develop at exactly the same rate as new parts or wiring is added...........no air-wire prototype has ever been known to exceed this theoretical limit for more than a few seconds
That was an issue with vacuum tube early electronic brains. Parts are generally reliable enough that the complexity limit is going to be schedule, or even lifetime!
One of my prototypes used 3000 parts, no issues what so ever besides circuit design. One manufacturer didn't like it as they were unable to build units that complex reliably. The folks who ended up building it have been selling the little brother product for better than 30 years.
One of my prototypes used 3000 parts, no issues what so ever besides circuit design. One manufacturer didn't like it as they were unable to build units that complex reliably. The folks who ended up building it have been selling the little brother product for better than 30 years.
Yes, the same principle applied there though the issue there was intrinsic unreliabilty of thousands of devices. The issue with air-wire prototypes is entirely of our own making though - for example needing to expand part of the circuit three layers deep in a 3D wire-space array of components !
Prototypes expand to overspill available space, is another natural law. Just like any wire cut to exact length will be too short ! What other laws are there? It'll always get hotter than you thought ?
- Status
- Not open for further replies.