John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
Yes it is funny, obscure and mad. In some amplifiers (SE tube power amps e.g.), the same people would not care about 5% (-26dB) distortion and will say they sound the best. In resistors etc., the fight will be about -140dB. This proves the only thing - non existent engineering method to support results of subjective evaluation and crazy attempts to explain it by non-linear distortion.

At home in UK I have been listening to this amp:

http://www.diyaudio.com/forums/sito-audio/217616-tssa-v1-6-docs-buyers-list.html

On paper the levels of distortion look truly frightening but in reality once I had attended to some PSU ringing - which I really didn't like the sound of - actually it sounds ok - it has "character" but that character does not detract from enjoying the music.

Simple truth is pure 2nd HD does not sound unpleasant even in large quantities - personally I do not try to add in 2nd HD on purpose ( unless it's a trade off for better stabilty ) but it's last on my list of distortions / noise that I worry about.

Bad sounding resistors are quite high on that list.

and now you bring it up I think I would rather have a theoretical amp with 1% 2nd HD but no other noise or distortion of any kind compared to the same amp but with no 2nd HD and made with cheap resistors.

The former could sound silky smooth when signal sounds silky smooth but by comparison, with the same signal, the latter would sound rough.
 
Last edited:
I use IC's mostly for servos, and I have done so for the last 30 years. However, only the jfet input devices are useful for this.
I have sampled IC's since 1966. I buy them by the 100's when I have a project, as well. When Walt Jung was at ADI and Scott Wurcer more friendly, they both would sample me their best audio offerings. That is how I got to use the AD825, AD797, AD711-2, and many others. I still have dozens of each, if not 100's in my lab.
I have nothing personally against IC's any more than I have against American Cars.
They both serve a significant purpose in what they are designed to do. AND they really help to make low distortion audio measurement equipment.
Then why don't they work 'perfectly' for hi end audio? Well it is not because they measure so low in lower order harmonic distortion, we have been beyond that barrier, for the last 30 years or more. It is the HIGHER ORDER distortion, or more fundamentally, the 'dropout' in the transfer function that generates higher order distortion that is what the ear seems to not be tolerant of. ANY hint of higher order distortion, 7th and 9th especially, and you sink to mid-fi. Why this is so, is partially based on the dissidence of these harmonics and their corresponding IM products as being easily detectable by educated listeners, who are sensitive to that sort of thing. Many here, probably accept mid-fi as their listening standard, just like some guy might judge his pick-up truck as an automotive standard, not wishing for more, and laughing at the extra cost of some vehicles, that make no sense to him. I cannot help that some people can hear audio differences and others are more forgiving. I don't design audio products for the 'more forgiving', they would be better off at Best Buy, or the internet to get their audio equipment, as I buy a lot of my video and computer stuff. Great bargains! '-)
 
I would LOVE to know what to do with them and put them to meaningful work.
So do-i :)
If they have little slew-rate, like TL072 or NE5532/4, i believe you can't do anything else impressive with them than servos, unless they are the kind of VERY low distortion brands you are talking about. :)
I had send to garbage thousands of them from various mixing desks.
Hopefully, the one inside the PGA23XXs is this kind of very low THD animal.
Limiting the bandwidth before input, can be, sometimes, a solution, as well as adding a discreet current booster to their output inside the loop... as you know very well. But most of them are boring, i agree.
This said, i was never disappointed by video OPAs for audio, specially Current feed-back ones. Did-you tried some ?
Bad sounding resistors are quite high on that list.
It would be interesting to see in witch context (power margin, brand, place in the circuit, you find them so 'audible' ? May-be we could find some reason ?
 
Last edited:
Well it is not because they measure so low in lower order harmonic distortion, we have been beyond that barrier, for the last 30 years or more. It is the HIGHER ORDER distortion, or more fundamentally, the 'dropout' in the transfer function that generates higher order distortion that is what the ear seems to not be tolerant of. ANY hint of higher order distortion, 7th and 9th especially, and you sink to mid-fi. Why this is so, is partially based on the dissidence of these harmonics and their corresponding IM products as being easily detectable by educated listeners, who are sensitive to that sort of thing.
I agree 100% with this (IM). Why do-you think i'm insisting so much on slew rate or limiting speed of the input signal?

My ears are not tolerant at all to any kind of IM, believe-me. I need air and space between each instruments, i need to can follow any instrument in a mix without any other mask or modulate-it, and every detail of it (specially the little details on attack, like nails or mediator on guitar chords, sticks on drums, metal wire on basses etc. ). I want no confusion between kick drums and bass playing in the same time too. And i want no sensation of heaviness from my speakers.

What kind of IM did i have to fear with 1000V/µs of an OPA at 2V output, and -70db of HD at 5Mhz ?
 
Last edited:
Then why don't they work 'perfectly' for hi end audio? Well it is not because they measure so low in lower order harmonic distortion, we have been beyond that barrier, for the last 30 years or more. It is the HIGHER ORDER distortion, or more fundamentally, the 'dropout' in the transfer function that generates higher order distortion that is what the ear seems to not be tolerant of. ANY hint of higher order distortion, 7th and 9th especially, and you sink to mid-fi.

Got some data? What is the 7th and 9th for (say) a 797 or 4562 as a line amp driving a standard 10k load compared with (say) a Blowtorch or Parasound?
 
It would be interesting to see in witch context (power margin, brand, place in the circuit, you find them so 'audible' ? May-be we could find some reason ?

I use cheap metal film resistors from mostly Maplin Electronics UK rated 0.6W & Parts Express US rated 0.5W + a few others types.

I didn't distinguish between them.

I also avoid trimmers ( 22 turn cermet types ) as much as is possible or design the circuits so that their o/p is filtered or diluted - I think they sound even worse than fixed resistors but I didn't do a study on this - studying these resistors does not really interest me - the solution is simple - buy better resistors - I hope I can find some good ones that are cheaper than the Caddocks - that would be great ! I will research more on this later this year.

As I explained already I heard a difference in all amplifier circuit positions - even as VAS CCS ( dual BJT ) feeder resistor ! The overall effect when all are changed is not small.

If anyone reading this finds it hard to believe - why not just try it just once - one amp with all cheap metal film and another with Caddocks TF020 or MK132 - you may be very pleasantly surprised . . . or you may not hear any difference - but personally I would be too curious not to try it.
 
AD797: sample
 

Attachments

  • Picture 19.jpg
    Picture 19.jpg
    196.8 KB · Views: 194
Did I say that the AD797 or the 4562 were mid-fi quality?

Yes, you did.

I have nothing personally against IC's any more than I have against American Cars.
They both serve a significant purpose in what they are designed to do. AND they really help to make low distortion audio measurement equipment.
Then why don't they work 'perfectly' for hi end audio? Well it is not because they measure so low in lower order harmonic distortion, we have been beyond that barrier, for the last 30 years or more. It is the HIGHER ORDER distortion, or more fundamentally, the 'dropout' in the transfer function that generates higher order distortion that is what the ear seems to not be tolerant of. ANY hint of higher order distortion, 7th and 9th especially, and you sink to mid-fi.

They ARE ICs, right?
 
John,

As a matter of interest - how low would H6 - H19 have to be that you think it is no longer audible and spoiling the sound ?

In spice when I was designing a servo circuit I could see that a poor design raised the harmonic noise floor from -150db to -125db with H1 10V peak o/p and in reality I could hear the sound had lost it's transparency.

So I try to aim for -150db H6 - H19 noise floor in spice but I have no idea what is happening in real life.

Lookin for a reality check !
 
Partial schematic with 'loading'
I don't understand.
The output load is 2K (less the charge of the following circuit input i don't know).
The feedback resistor is 530 Ohm: Why ?
As it look like an inverting configuration, we need to know the source impedance to have an idea of the gain + the charge induced by this strange low value for feedback resistance, while they suggest 6 x more in TI application notes.
Notice TI measure the best figures for 2K load, i would use this value as a minimum, and certainly not drive a 600 Ohm load without an additional buffer..
 
Last edited:
Status
Not open for further replies.