I built my first version after reading the IEEE paper and couldn't get any where NEAR the awesome performance. Then I built one with regulated supplies on the early stages and the awesome performance in the IEEE paper and the databooks appeared.
You might like to get a MAPLIN kit and try this out for yourself ... that's assuming these kits have genuine Hitachi MOSFETs. You'll need a separate transformer etc for da regulated supplies.
Can't remember if I had a HMA 7500 to look at. It was the 70s
The sound of a parts is the same at all points in a circuit. Whether as a power transistor, driver, diode, power supply or whatever: a BD139/BD140, for example, always sounds "bad" everywhere. You mention an experience that has not been analyzed and interpreted accurately enough: the development of complex circuits (concepts): the addition of further parts (objects) leads to a reduction in resolution and the merging of "sounds". To most listeners, this seems like an improvement: gray and flat and disembodied (e.g. Hitachi MosFets) becomes "darker" and "simpler" - roughly spoken. A Hitachi HMA 7500 mentioned above, for example, is a sonic disaster: coarse, choppy, colorless noise. I also bought, tested and converted this one - I have just checked the statements of others. I advise you to check my statements, for example. To check practically. To gain practical audio experience;-)
Sorry, but you keep showing your audio engineering inexperience)-;
I also advise you to do homework, to test my statements (so called "science", testing of statements;-). First of all this:
You will find out why this is a key experiment for audio engineering. Almost the entire audio electronics discourse can be analyzed by means of this experiment;-)
My experience is that long before any difference is audible the measurement gear has detected it orders of magnitude
below the threshold of audibility, to hear a diference you ll have to use something that has vast amounts of non linearities,
at wich point you ll be capable to make the diference between 5% and 0.5% distorsion, and not sure that it will be the latter
that will sound better for some dire hardcore "designers" by ears...
To add... Under +/-40v rails BD139/40 is the most popular driver transistor pair i know.
I, and others, hear differences between different batches of a transistor. Peek-measurement methods fail here, too. And YOU also would hear differences between parts of ONE batch: see mentioned "homework". If this experience is not made practical, lines and numbers and concepts will be misinterpreted as audio, which is again un-scientific and un-engineering!
I am in favor of opening a subforum: "diy audio", so that the 10 to 20 people interested and dedicated and committed in audio engineering are undisturbed. The rest could continue to talk about lines and numbers and audio distant concepts;-)
You have NO IDEA what homework is until you get an advanced degree, buddy. All that sedentary studying you’ll be forced to do will make you put on 25 pounds and develop insulin resistance. Interested?
All the tinkering, listening to the “sound” of components, and going through simplistic design exercises is actually quite valuable and can teach you things you’ll never get out of a classroom or textbook alone. But eventually you need to move beyond that - and discover why things behave or sound as they do. The above mentioned advanced degree helps, but one can learn it on their own without paying tuition if they put their mind to it.
All the tinkering, listening to the “sound” of components, and going through simplistic design exercises is actually quite valuable and can teach you things you’ll never get out of a classroom or textbook alone. But eventually you need to move beyond that - and discover why things behave or sound as they do. The above mentioned advanced degree helps, but one can learn it on their own without paying tuition if they put their mind to it.
There was one of Murthy’s corollaries that sates “If nobody uses it there’s a reason”.
The opposite is also true - if everybody uses it there is also a reason.
There are quite a few very well regarded amplifiers using this topology that don't have regulated supplies on the front end. The Perreaux and Australian Monitor amps come to mind. The Perreaux doesn't even use a CCS on the LTP.
I have a Perreaux. It has the same faults I found in my previous century prototype.
The fact that these Hitachi MosFets are not suitable for (home) audio is due to their TO-3 casing (shape and diameter) and materials. This is a problem with all TO-3 transistors. At best, they are suitable for high-current: "audio" applications, i.e. for disco sound systems, gymnastic halls, that sort of thing. Cleanliness, low signal deviation, are no longer important here.
I also refer here to a categorization into at least 2 application areas that are audio relevant - mentioned here in sufficient detail: low current and high current. Home audio is mostly low current.
And also: the development of transistors is also a history, a process (not analyzed and considered) that is NOT based on audio experience!
And: most electronic parts are still not oriented towards audio performance. Because: current and signal are not understood as complex process - still not!
And that is not taught either. Still not;-)-;
I also refer here to a categorization into at least 2 application areas that are audio relevant - mentioned here in sufficient detail: low current and high current. Home audio is mostly low current.
And also: the development of transistors is also a history, a process (not analyzed and considered) that is NOT based on audio experience!
And: most electronic parts are still not oriented towards audio performance. Because: current and signal are not understood as complex process - still not!
And that is not taught either. Still not;-)-;
Nonsense! The Hitachi MOSFETs had their TO-3 cases Hand Carved from Unobtainium & Solid BS by Virgins. You just have to ensure your other bits, especially the Mains Cables are also Hand Carved from Unobtainium & Solid BS by Virgins.
Today, the best audio stuff comes from China. That's cos they are the only country with any Virgins left. Without Virgins, you can't get cleanliness and low signal deviation,
I'm out...
You too have to decide: do materials have an influence on current, signal or not;-)
I am sure you will also prefer copper to steel or even wood;-)
- Status
- Not open for further replies.