As of writing this, Little Star - Madonna plays from the following setup at excellent quality heard from any room. Bandwidth sounds below 100Hz to at least 15KHz.
Signal path:
Little Star - Madonna from CD - Sony Walkman line out bridged - Pleiades R,C filter - Pleiades 2N3055 power amp - linoleum Pleiades Auratone 5C type cube speaker - reverberant room
It sounds amazing, natural on any room. Why should it it not?
It is just one 2N3055 transistor operating in class A (electrons flowing all the time), (electrons flowing all the time through speaker as well). No capacitors in the signal path. Just a bias resistor from collector to base.
There is no crossover network on these speakers as it is just a single fullrange drive unit.
The speaker box is small so there are minimal difraction effects.
Madonna's voice comes from one place, not half of its bandwidth from a woofer and the other half or whatever from a tweeter at a different point in space. So accuracy is pin point.
Objective frequency reponse is + or - 3dB from 100Hz to 15KHz. See previous posts on acoustic measurements.
A huge magnet Tesla 5in speaker type TVM ARX-130-20/4 is used.
Linoleum box vibrates as a whole possibly giving efficiency for bass frequencies. Mid coloration seems small as there must not be a lot of trapped mid range energy in the box due to the absorbing qualities of linoleum.
So what is missing?
Missing is the option to listen to below 100Hz.
This is why the Pleiades R,C shelving low pass filter is added before the power amp.
It's purpose is reducing everything above 100Hz by say 6dB so that deeper bass is revealed.
It is perhaps defeating the purpose of monitoring Auratone 5C type cubes.
The idea is creating a filter inside the cube box that is disconnected by a switch reverting operation to normal for those speakers 100Hz to 15KHz.
This filter might not be easy to make but various attempts were done today with the electronic cooking ingredients. Inductors, capacitors, resistors.
First tests were done directly at the speaker out of the JVC amp while cube speaker being connected.
At your risk.
Capacitor values of the order of 2200μF. There was distortion. Possible reasons the JVC radio amp does not cope, they are electrolytics etc. They were also tried in series with R for the shelving effect, but R is already there by the connecting alligator cable's resistance. Experimenting at 4Ω impedance is tricky. Possibly less than 2200μF is needed. Preliminary tests were done by listing to a test tone CD and observing SPL difference below 100Hz or above 100Hz while connecting the filters including the filters below. It is tricky as the filter reduces harmonic distortion too.
Then inductors is series with JVC amp out were tried. It was found by surprise that just a few dozens of millihenries do the trick. This is because the time constant τ is L:R or τ=L/R and R is small so τ gets easily large. Even connecting 1mH in series reduces sparkling treble. A resistor in parallel will create a shelving filter bringing the sparkle back. For doing this above 100Hz 60mH might be needed. Creating such an inductor as air core would create a great, series resistance. So this option was abandoned for the time being.
It was decided to play at 300Ω or 150Ω which is the output impedance of the Sony Walkman at line out.
Suitable capacitor value by trial and error was found to be about 470nF. And a series resistor of about 300Ω.
When connecting the Sony Walkman to Pleiades 2N3055 amp which has a lower input impedance suitable values so far found are 200Ω,2.2μF. The bias resistor from collector to base is 10KΩ. A series to base potentiometer adjusts volume so it is tricky as I out impedance varies as it is rotated to adjust volume. (A possible equal loudness curve compensating filter?)
More experiments needed. Including of course fine tuning with a Sound Presure Level meter at test warble tones from the Alan Parsons Sound Check 2 CD
No comments:
Post a Comment