Output BJTs for buffer

[analog_sa]

The simplest test: use (double) mono-psus. And then connect these.

I don't seem to understand even the simplest test

Use and then connect??

 

[adason]

Do not feed the troll.

 

[cumbb]

Hearing as opposed to what actually hits the eardrum involves interpretation by the brain. I have a good understanding of what people can think they hear, and basically the brain exerts a far greater influence on what you think you hear than what is actually "heard". This is your problem.

Expectation bias. This is your explanation pure and simple. Couple that with confirmation bias and we have a person who ignores all other evidence that opposes what they have become invested in.

Universities only ever teach the basics coupled with how to learn. It is further education and reading plus direct controlled experience that brings a full understanding.

Now, the human body makes a very poor instrument. We are designed very well how to survive in dynamic situations. We do not have built in references, we can only compare. But our mind overrides and processes all inputs. We recognise danger or safe patterns and act accordingly. that is what we are optimised to do. In addition, we have a limited s/n ratio as our bodies are noisy. Then there is the environment which will have a noise floor greatly above what you say you are hearing. Finally, there is your experimental setup. It can be extremely difficult to design an experiment that eliminates or at least controls factors outside of the thing we are investigating.

You will disagree with what I am saying to you of course. I would suggest you quietly assess what has been said, then research the facts before saying anything else. This is not going well for you.

-Chris

;-)
The simplest test: use (double) mono-psus. And then connect these. With a switch at the listening position, if you like. And switch back and forth as long as you need it.
All co-writers: please test!
Further on after that.

 

[cumbb]

Do not feed the troll.

Experience Denier;-?

 

[cumbb]

I don't seem to understand even the simplest test

Use and then connect??

Use = listen music. Then connect (both together; your "double-mono-psus") and listen music. Then disconnect and listen music. And so on.
;-)

 

[anatech]

Hi cumbb,
You're stuck on this and too invested.

There isn't any point in doing this and it will not show what you are saying. You are now trolling.

 

[cumbb]

I sort of share some of the concerns above, but also find it difficult to make hard generalisations.

As it happens, the two most recent amps i have built and enjoyed are both followers, but more different than alike. Both are built as mono blocks.

The first one is based on a THF-51 SIT running at around 3A/30v with a CCS underneath, so a 60v PS. Two coupling caps, a polystyrene Multicap at input and a 10mF Cerafine at output. No pcb. Separate transformer/rectifier/regulator for setting the biasing voltage.

The second is a diamond buffer with 4 pairs 3281/1302 based on a circuit by diyaudio member SandyTodorov1. No coupling caps. 2.2A @ +-34v

Both have very low distortion at low power, and the SIT gradually increases in distortion towards its terminal 15W 4ohms while the PP amp preserves sub 0.01% distortion at higher powers. Tried running the diamond buffer at various currents between 1.6 and 4A before settling to 2.2A. The highest current did not sound particularly pleasing to me.

So, how do both amps compare at low listening levels but into a pretty demanding load which occasionally dips below 3ohms?

The SIT certainly has a charm and highlights voices and cellos. Does not do particularly well at either of the frequency extremes, has a narrower soundstage and significantly less detail. Still, very enjoyable.

The PP amp has a slightly cooler tonal balance, still does justice to voices and can suit a wider selection of musical genres. It also sounds faster, taller and wider but perhaps shallower.

There are some build differences between the amps. The SIT uses 4320 based rectifiers while the PP amp has the cheapest chinese bridges money can buy.

Some of my prejudices like SE operation and coupling caps seem to be justified in this comparison, while others are not. Neither the plastic cases, nor the multiple parallel devices, nor the pcb with a mask seemed to subtract enough from the PP amp performance. Ditto for the emitter degeneration resistors.

I intend to still try 4320 and Ge rectification in both amps and also to build a version of the PP amp with Sanken outputs instead of ON semi.

Below are the general circuits used.

;-)
Thank you for this constructive post first.
"... significantly less detail." Where did all the details go;-?

 

[anatech]

cumbb, you are now trolling. Stop.

 

[analog_sa]

"... significantly less detail." Where did all the details go;-?

This is a pretty good question, been pondering the same myself. Especially taking into account the PP buffer, although simple, is a lot more complex than having a single transistor. Always thought that paralleling reduces immediacy and has a homogenising effect on sound, but not the case here. Of course it is apples to oranges comparing a single SIT to multiple BJTs. A single pair of BJTs is just not a viable option under these operating conditions.

For the record, the voltage amplifier has a 100ohm output impedance so the input capacitance of the SIT, which in the case of a follower is not that high, should not be an issue.

 

[Citizen124032]

And switch back and forth as long as you need it.

That's not a double blind test, it's not even a blind test.
You're conforming your biased perception over and over.
It must be that you're not in control to what you are listening to in whatever order.
Even on sub-atom level does particals seem to behave according to the expectation of the observer.
I hope you attend live performances of authentic instruments unamplified by electronics regurlarly to have your hearing reference 'calibrated'.

 

[cumbb]

That's not a double blind test, it's not even a blind test.
You're conforming your biased perception over and over.
It must be that you're not in control to what you are listening to in whatever order.
Even on sub-atom level does particals seem to behave according to the expectation of the observer.
I hope you attend live performances of authentic instruments unamplified by electronics regurlarly to have your hearing reference 'calibrated'.

You could do as x times double double blind deaf dumb test: let your sugar switch.
And don't let it tell you which switch position. But give your hearing at least 10 seconds break, silence, for a processing time, memorization time.

And everyone could do it that way. And whether with or without expectation, the same experiences will, without that we will have agreed, be written on the pieces of paper:
1. change,
2. at position x the character a, at position y the character b. And with all, but a few exceptions - the reasons for this can be clarified later -, the same assigned characters will be on the notes.

 

[cumbb]

This is a pretty good question, been pondering the same myself. Especially taking into account the PP buffer, although simple, is a lot more complex than having a single transistor. Always thought that paralleling reduces immediacy and has a homogenising effect on sound, but not the case here. Of course it is apples to oranges comparing a single SIT to multiple BJTs. A single pair of BJTs is just not a viable option under these operating conditions.

For the record, the voltage amplifier has a 100ohm output impedance so the input capacitance of the SIT, which in the case of a follower is not that high, should not be an issue.

I would invert too: Two questions: does the SIT "lose" details OR does the PP "generate" "details".
The answer could be found in the experiment described above.

 

[analog_sa]

Two questions: does the SIT "lose" details OR does the PP "generate" "details".

Since the last time i posted here things changed. The SE amp now plays a lot better. Generally small, "insignificant" changes transformed its sound. In this light your question makes a lot more sense.