Hi Naf, keep on the spirit
I understand what your goal (experience is the best teacher)
You already made the lateral one but you want to know how the BJT version "sound" to compare with
& more importantly with all BJT there will bring more DIYers that can try SSA
maybe we can call it naf's SSA someday you can make it
Can you post your latest sim file so the other can help you to make it better
Hi Jay, what the best version of SSA do you like?
I really want to see (pictures) of what you have been built so far, please
"curious mode on"
Which one may sound good & which one should be avoid?
Regards
Hey thanks a lot, it is a good solid state amp and yes there will be even more sons of SSA present here in near future.Hi Lazy Cat, Sorry for looking at this circuit so late in the day. Great job
Regards L.C.
The best experiences are learned step by step on a personal level. As others I also got an impression that naf would like to go this way. Otherwise he can go and grab a complex sch from some service manual and build himself an Accuphase on the spot, for instance. Not the right way to learn basics. Also at basic SSA he can immediately make an upgrade from single BJT outputs to Darlingtons, experiment with some different types of triple, etc.
Not at all.
Yes, most probably, err... no, I didn't use it. But I don't understand about something about this input filter. For example why, stability-wise, there is a situation in a simulation where no filter is better than with filter.
Possibly, but I like to see it in a different way. I mean, when objectivists say that the distortion adds something, I see it as the low THD circuit takes something from the circuit.
I think you or somebody else have mentioned about the speed of the current source (or power supply) and it makes sense to me. I think Salas has a very good understanding about power supply effect on audio amplifier.
Yes Jay, I would also very much like to see any of your work, honestly.
It is a difficult question, as I have different "taste" and approach in finding a good sound. For me, an amp is only a small part of an audio system.
In an amp, similar circuits may give different sound so I always use very good power supply and parts (to make sure that they are not the bottleneck). And I rely on ears to find out if I enjoy the sound (sonically musical and there's no hint of fatigue). And it is difficult as new circuit (or ears, who cares) requires burn-in time
SSA is a simple circuit. It ensures good sound to me, from the simplicity alone. In SSA I'm looking for certain sound character that can be described as "direct energy" or "speed" (what the hell is this
), which I think is associated with the feedback topology.I tend to find the best performance in a typical circuit, without having to increase the complexity. For example, the basic circuit as used by Nico and Shaan, cannot be used with high supply voltage. There's a trade off between power and quality. The highest I think I want to go is 38V.
Of course, complex auxiliary can help lower the noise floor which is very important for "higher level" sound. But usually there is more trade-off so is not easy to tweak.
Current mirror and cascades have been known to lower distortion, but I have never "satisfied" with amps implementing cascode. Same with commercial amps, even tho specs are better on paper, they tend to not excel on the market. May be another driver stage is required between cascaded VAS and output stage. But as I'm only a newbie, achieving stability is not easy for me, and I don't want to risk my laterals
. You want pictures? Hehe I don't build finished amps so no fancy looking amps here. But I can show you a stack of empty aluminum enclosures
and also PCB modules...
It is all about TIM.
If the slew rate of the signal is higher than the one of the amp, it begin to create TIM distortion.
If you slow the input signal under that spped, i believe you don't have to worry about the speed of anything, including CSS.
If you have the measuring tool, i suggest you first linearize response curve (No peak at HF) with cap parallel to the feedback resistance, then set the input filter for no overshoot at 20Khz square wave. The value of this last cap will be the mini, that you can increase in listening, to see if any further improvements.
Last edited:
About stability, we are on a very fast amp, with Mhz bandwidth. The input cap will add a cap to the ground with the one between collector and base of the input transistor, so it will change a little the phase in the loop. This will decrease a little your phase margin, but increase the stability with very high frequency square waves. That you can simulate.
I see, I will remember this the next time I implement a ccs.
I was mentioning that input filter doesn't always contribute positively to stability, in simulation.
I think simulator is sufficient to linearize the peak at HF, also the pulse test. I have done the same thing, yes. But there is compromise against stability against capacitive load, which is not so "accurate" I believe, but scary enough if the amp fail in the middle of casual use.
Strange that many good amplifiers simulated well (precisely) as if they are designed in simulator. Or, good amp must be simulated well but well simulated amp is not always a good amp?
You can just use the simulator as a guideline, to see if the currents are ca Ok and how the thermals are. you can use it across topologies to see how they match up, and other rough work, But replace Real test circuits it can't. and being a judge of sonic qualities in amplifier is simply not possible, as there's so much a simulator does not tell us any thing about...!!
there has been some quite muddy discussions on VFB vs CFB going on based on simulations that took us nowhere... What if CFB (like i SSA and TSSA) are simply better at dealing with loudspeaker return currents...? so what we here in real-life is simply not visible in sims as the dynamic load-case is not the mimicking real life situations... (just a thought)
there has been some quite muddy discussions on VFB vs CFB going on based on simulations that took us nowhere... What if CFB (like i SSA and TSSA) are simply better at dealing with loudspeaker return currents...? so what we here in real-life is simply not visible in sims as the dynamic load-case is not the mimicking real life situations... (just a thought)
I second that. Simplicity and real behavior on loudspeakers may be the key to good sound.
I rarely see measurements done with a loudspeaker connected. Here a guy has at least done that with a tube amp :SUPRA-Triodenschaltung von Frank Kneifel
I rarely see measurements done with a loudspeaker connected. Here a guy has at least done that with a tube amp :SUPRA-Triodenschaltung von Frank Kneifel
Interesting link, but i suppose his model is not enough.
I wonder why LTSpice does not provide models for loudspeaker charges. I mean real models, with the current mechanically induced while the cone is not fully stopped in its resonances at transients.
The second thing i do not understand is why they do not provide an easy square waves generator, where you set the frequency, the level, and the offset... and, by default potentiometers, trasfos, and some other very ordinary parts we use all the time.
I'm bored with their stupid interface...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.