I am so enjoying to read this thread as it was just started, still surprising me deeply when new guys shows up, determined to make SSA according to their version of sch or already made one and pics showing really amazing DIY SSA amps. Looks like this thread encourage people to go to deeper waters to explore and still feeling safe. If something goes wrong they can expect help and our mutual support.
Best regards to all of you guys, have a nice day L.C.
Best regards to all of you guys, have a nice day L.C.
I agree with both of you here. Imo this is highly affected by tweeter performance. Bad tweeters with distortions (due to amplifier output or by itself) tend to give a "life" perception. Better tweeters, due to the absence of distortion will tend to give "airy" sound which actually, imo, IS SOFTER than life. May be only a really "good" tweeter that can give "life" sound without distortion, hence no fatigue like tweeters with distortion. But this opinion is still uncommon because most people are used to the "soft and airy" sound of high end tweeters (especially planar, which I don't like but many do).
My toughs about this treble presence question.
As opposite to what can be expected, a fast amp with low distortion will produce kind of "less treble" on a psychoacoustic point of view. First hight slew-rate apply mostly to high energy, low frequencies instruments, basses, kick drums. If they are improved, your first impression will be better basses, means less relative treble. Second, distortion lies in high frequencies (harmonics), so, mostly in treble. reduce them, you will decrease the trebles level.
More precision. Turn your signal to a square wave near 15 000hz. Some can object that, because our ear is not able to hear higher frequency than 20 000 (for the youngest of us) the ear will not make the difference with a sinusoidal signal, there. It is not true. The ear is able to feel the "Surface" = energy" of the square wave, and all sound engineers with such a corrector knows how you can change the weight and presence of an instrument, just by adding some dbs at 40khz, without changing its tonal character.
Last, slow amps are not linear in phase, near the high end of the spectral frequencies. If you null phase shift at 20kz, you will add more coherency to the instruments. treble will sound less "separated" from the other part of their spectral, and you will notice a "less treble" presence.
On my experience, better an amp, more natural, means less underlined, the trebles.
As opposite to what can be expected, a fast amp with low distortion will produce kind of "less treble" on a psychoacoustic point of view. First hight slew-rate apply mostly to high energy, low frequencies instruments, basses, kick drums. If they are improved, your first impression will be better basses, means less relative treble. Second, distortion lies in high frequencies (harmonics), so, mostly in treble. reduce them, you will decrease the trebles level.
More precision. Turn your signal to a square wave near 15 000hz. Some can object that, because our ear is not able to hear higher frequency than 20 000 (for the youngest of us) the ear will not make the difference with a sinusoidal signal, there. It is not true. The ear is able to feel the "Surface" = energy" of the square wave, and all sound engineers with such a corrector knows how you can change the weight and presence of an instrument, just by adding some dbs at 40khz, without changing its tonal character.
Last, slow amps are not linear in phase, near the high end of the spectral frequencies. If you null phase shift at 20kz, you will add more coherency to the instruments. treble will sound less "separated" from the other part of their spectral, and you will notice a "less treble" presence.
On my experience, better an amp, more natural, means less underlined, the trebles.
My life (Totally off topic ?)
When i was a young sound engineer, with ears able to feel sins up to 25 000 hz, i had a lot of problems with cymbals. Trying to add some level near 10 or 15 000hz. While i admired some cymbals from some old engineers.
At the same time, i was unable to afford violins, with the noise of friction of the bow on the strings hurting my teeths !
When i was old, able to afford and even appreciate violins, i never tried to add anything anymore to the trebles, but, on the contrary, to remove little levels from treebles adding basses on it, instead, young drummers used to compliment the sound, natural and energy, of my cymbals.
About loudspeakers, i was never satisfied with "Tweeters". Incoherency, undefined "pshh pshh" sounds, like a modulated noise. No dynamic, no "material" in the instruments ends. Specially with all those awfull, awful, awful soft domes.
Now, i use a horn from 1500 to 20 000 (spherical waves, circular, kind a JM Lecleach). It does not goes flat higher than 16000Hz, but instruments are perfect, never notice abnormal treble, or even asking-me questions about this part of the spectrum. Cymbals are just metal, and you can hear all the weight and stick high energy impact on them.
I tried to add a teeter, just a little (flat) energy from 15 00 to 30 00. Never worked, incoherency was back again. I left. After all, in real life, who notice any "trebles " ?
When i was a young sound engineer, with ears able to feel sins up to 25 000 hz, i had a lot of problems with cymbals. Trying to add some level near 10 or 15 000hz. While i admired some cymbals from some old engineers.
At the same time, i was unable to afford violins, with the noise of friction of the bow on the strings hurting my teeths !
When i was old, able to afford and even appreciate violins, i never tried to add anything anymore to the trebles, but, on the contrary, to remove little levels from treebles adding basses on it, instead, young drummers used to compliment the sound, natural and energy, of my cymbals.
About loudspeakers, i was never satisfied with "Tweeters". Incoherency, undefined "pshh pshh" sounds, like a modulated noise. No dynamic, no "material" in the instruments ends. Specially with all those awfull, awful, awful soft domes.
Now, i use a horn from 1500 to 20 000 (spherical waves, circular, kind a JM Lecleach). It does not goes flat higher than 16000Hz, but instruments are perfect, never notice abnormal treble, or even asking-me questions about this part of the spectrum. Cymbals are just metal, and you can hear all the weight and stick high energy impact on them.
I tried to add a teeter, just a little (flat) energy from 15 00 to 30 00. Never worked, incoherency was back again. I left. After all, in real life, who notice any "trebles " ?
After the one in the pics is going to be tuned best with CCS will be re-assessed and decided if it will remain resistive or active current sourced. There is TSSA 1.6 in the pipeline too to listen in the same system at a point, plus VGeorge's Le Cleach up to 10K with Great Plains large format driver and Fostex bullet for last octave speakers to finally visit and assess on. Through those impressions I should be better positioned to comment about this interesting life like optimum treble discussion and SSA input topology in the future I hope.
Sure I agree with you. What I mean is that a supposed to be a "good" tweeter should not sound "softer than life", it should sound more "pronounced" but not because of distortion.
Cymbals are metal. It is very hard, if possible, to produce such sound with soft domes or planar. So people mostly have to choose (or have to be used to) between distorted sound or softer-than-life sound.
I have a tweeter that at the first listen I can hear a "weight" with the treble. I thought, this must be a bad tweeter and must be fatiguing with long listening. Then for some reason I put the speaker in my bed room, and then I realized that it is not fatiguing at all (I cannot live with distorted system, especially because it will disturb my sleep. Putting the system in bed room is a standard test for me to check if a system pass the QC or not
).I then had a hard time to decide whether I want to use the tweeter for a better speaker design or not. Still undecided.
The tweeter uses some kind of material (still a dome) which is named, if I'm not mistaken, SELENIUM. Seems like a metal, but soft like fabric. This dome material is popular in PA speaker system.
Yes. Expensive tweeters tend to have a very smooth response, low Fs and high power handling that make it easier to be designed without break ups. Tweeters with very good response at very high frequency (40KHz or more) can display a long cymbal sound, at the high end portion.
I think even average amps can display good signal at HF (with good tweeters). So if there is something special to be expected with SSA, it must be the phase related issue which is not yet clear for me (why and how).
dear Salas and Shaan,
channel 1 done, current feed change by paralelling 3k3 || 2k4.. Vas current down to 14.2mA
Hi Naf,
[edit] Can you show it in the schematic? I may get hold of my 25vac transformer this coming weekend, and of course it will be close to +/-32vdc.
TY!
Last edited:
this is my change 1k2 to 1k4 by paralelling 2k4 with 3k3
r 220 increased to 267 by series 220 + 47
Attachments
Adding an offset to the input signal DC coupled, signal = 1/2 max level, offset = +1/V or -1/2V. That way, the signal will go from 0V to + (or - ) V.
We are not interested, here, with the rejection itself, but the distortion induced by asymmetrical signals.
Of course, we need a cap in the output before to measure the distortion. Then compare distortion with offset vs distortion without.
Last edited:
Hi Michael, I wonder which o/p device combination you preferred in the end ? In spice I found irf610/9610 followed by BJT looks best but for a low power amp a BC546c followed by a lateral MF also works very well - this is what I'm using at present but I will upscale soon.
Do you think a fixed resistor from a perfect supply would still spoil the FB operation ? In spice I see little difference . . . or do you also think the high impedance of a good CCS contributes to ideal working conditions ?
mike
The fixed resistor sets the working (driving) current flow in the input pair... when you inject the current from the feed back it will add to the driving current. if the current source is ideal the process will be purely additive and precise, but a 1k Ohm current source (the resistor)..is hardly ideal and so the feed back process is unlinear with amplitude. in steady state (sines) (spice) this effect will not be visible in terms of higher distortion, but in dynamic conditions with ever changing amplitude this effect is easy to understand.
Listening vise you get an amplifier with better focus and a stable character.. when inserting a CCS.
My SSA has changed quite a bit now i set the VAS current through mirrors and use Hawksford cascodes on both input and VAS.
My output is a combination of Fets and Bjts...
Listening vise you get an amplifier with better focus and a stable character.. when inserting a CCS.
My SSA has changed quite a bit now i set the VAS current through mirrors and use Hawksford cascodes on both input and VAS.
My output is a combination of Fets and Bjts...
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.