tube-like SS power amp

[tinitus]

Actually Im not really sure its that simple at all

I have a small AB amp that seem to just follow the signal very truthfully

When my speakers are not working properly(xo) it just sounds bloddy awfull, and I begin to think whether amps are broken, making them oscillate

Seems I have just now been real lucky with speaker adjustments and now it sounds very pleasing, natural fullbodied, lots of drive and rythm, fluend with impressive ambience and superiour sonics, if there is any
Sound is simply very "responsive" and you cant get enough of it

But its apparently not the amp that does it at all
It just does what its told to do, which is also very important to achieve the above

 

[MJL21193]

Very high voltage SS with an output transformer.

 

[mosfets]

SS amps can be made to sound almost like a good tube amp if H2 and H3 is injected into the mix in very small quantities, around 0.05%.
Hugh

Hugh, are there any(tubish SS) designs in you're opinion, that stand out in the DIY arena? Perhaps specifically made with the intentions of replicating the classic tube harmonics? (outside of Nelson Pass as he clearly has A few in the bag)

Cheers

 

[c2cthomas]

Just curious if there is any SS power amp out there that may have a tube-like warm sound or qualities.

This might be of interest - and no, I do not have any connection with Grant Fidelity

http://grantfidelity.com/site/B-283+Tube+Processor+Buffer

 

[GK]

That's where tubes are usually better than SS - in overload.

This is a myth in my opinion. Just about every commercial valve amp intended for HiFi use that I have played with and have had in my system has had mediocre overload/recovery performance – particularly objectionable when overdriven to a degree that would still sound quite OK on a well designed SS amplifier with clean clipping performance.

This is especially so for valve amplifiers that employ capacitive coupling to the grids of the power output valves and even more so for those that use both capacitive coupling the output valves and global negative feedback.

In such amplifiers even moderate overload can result in a large spray of objectionable high order harmonics due the loss of bias that occurs when the inevitable forward-biased grid currents charge the coupling caps (blocking distortion). And then there is the recovery time proportional to the degree and duration of overdrive.

With such valve amplfiers (the vast majority) I would always want an adequate reserve of power to avoid clipping as much as possible.

Cheers,
Glen

 

[Mooly]

The original poster asks for amps with "tube like qualities"
I am unashamedly going to give mine another plug. The original thread got far to long and wandered around.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=119151&perpage=20&highlight=&pagenumber=1

Just say that if it's for a domestic setting it's unbeatable IMO. Don't let the simplicity of the design fool you, and don't get hung up over specs... it all comes together to give a totally outstanding and believable musical result. If I had heard better I would want it... !

 

[Wavebourn]

SS amps can be made to sound almost like a good tube amp if H2 and H3 is injected into the mix in very small quantities, around 0.05%.

Translation:

"A water from ocean can be made to taste almost like a good Coke if to add a brown pigment in small quantities"

 

[Lumba Ogir]

Mooly,
and it did NOT happen by chance.

 

[Lumba Ogir]

Glen,
but I don`t like your valve amp.

 

[Bigun]

I thought clipping behaviour is mostly important when the input has a wide dynamic range such as with microphone amps, guitar amps etc. ?

It looks to me that 90% of the 'established opinion' is in line with Hugh; you want to have control over H2/3 and noone harmonics above those.

Last week I was leafing through a British hi-fi magazine full of amplifier tests. They nicely provide a written description of the sound (subjective) and a graph of the harmonic performance. I found a strong correlation between 'sounds nice' and amplifiers that had harmonics that decreased monotonically (i.e. H2 highest, then H3 and then the others falling off really fast).

The classic LTP input topology gives you all the control you need over this. If the rest of the amp is designed OK to avoid higher order harmonics the LTP can give you as much H2/3 as you want to dial-in. Nelson shows how a JFET can be used to do pretty much the same thing.

A discussion over what makes an amp sound like a tube amp will be a long thread without a conclusion so I say follow the crowd first time around and rely on the collective wisdom !

An very well balanced LTP produces low H2 and some H3. Let it come out of balance and you can 'dial in' the H2. If you look at what published information there is on Hugh's AKSA you find that the LTP has no current mirror to force LTP balance, runs the LTP at a low current without emitter degeneration - all the 'wrong things' and if you simulate this you find that the current isn't balanced particularly accurately. So you get H2/3 ! No doubt Hugh has many other tricks up his sleeve, but this is a good starting point.

I'm unashamedly on a path to borrow from those that went before with my TGM 1.0.

 

[Mooly]

Mooly,
and it did NOT happen by chance.

Indeed it did not ! Greetings most learned one I hope you are well. Haven't posted much lately.

 

[Bigun]

Here's a few links talking about LTP distortion...

http://www.angelfire.com/ab3/mjramp/sym.html

http://www.angelfire.com/ab3/mjramp/ampdes1.html

and some thoughts on FETs + BJTs

http://gabevee.tripod.com/sstube.html

 

[GK]

Selective reading.

Also, umm, what valve amps are we talking about here? The majority of valve amps (push-pull configuration) inherently cancel even harmonics (ie H3 higher than H2) and can have harmonic spectral well above H3 - especially pentode output stages.

Class AB valve amplifiers also generally have a much higher class A threshold (relative to maximum power output) than do solid-state amplifiers - which has, in my opinion/experience, probably more to do with their characteristic sound than anything.

For example, a 50W SS amplifier with a single pair BJT power output stage and a bias current of 100mA will exhibit a class A threshold of only 160mW (8 ohms) and at typical and low listening levels will generate a rich spray of high level, high order distortion that would have pretty much zero correlation to the THD spectra Vs signal/power output level of a typical 50W class AB valve amplifier - no matter how much a misguided tweaker fudges up the input stage.

 

[Mr Evil]

To emulate a valve amplifier, won't you also need high output impedance? That can affect the response of a speaker quite a bit.

 

[hafp]

The last model SS amps that were produced by Conrad Johnson got a lot of comments of sounding "tube like"
Think it used JFET input circuit with bipolar output to achieve this.
Maybe someone with more technical knowledge could chime in on this and comment.

Cheers

 

[Bigun]

- no matter how much a misguided tweaker fudges up the input stage.

Isn't that the whole point, this thread is about 'tweaking' the SS amp to sound like something that it isn't

 

[Lumba Ogir]

Glen,

The majority of valve amps (push-pull configuration) inherently cancel even harmonics (ie H3 higher than H2) and can have harmonic spectral well above H3 - especially pentode output stages.

Any push-pull configuration does that.

A threshold of only 160mW (8 ohms) and at typical and low listening levels will generate a rich spray of high level, high order distortion

Although D. Self calls it linearity.

 

[Efren Noble]

Good day! May i suggest JLH 1969 ver. dual rails, capacitance multiplier p.s., idling at 1.5 amps/ch. (Be ready to spend more on hi grade heatsinks!) I use ONKYO AX37 speakers on my JLH. A friend once commented on my amp as "SOUND-A-GRILLE" (imagine using it in a tropical country like ours) but upon hearing it said it really "taste exellent"!

 

[unclejed613]

Selective reading.


For example, a 50W SS amplifier with a single pair BJT power output stage and a bias current of 100mA will exhibit a class A threshold of only 160mW (8 ohms) and at typical and low listening levels will generate a rich spray of high level, high order distortion that would have pretty much zero correlation to the THD spectra Vs signal/power output level of a typical 50W class AB valve amplifier - no matter how much a misguided tweaker fudges up the input stage.

you can get just as much distortion from OVERbiasing as from UNDERbiasing the output devices, which is why it's best to set the bias with a distortion analyzer. same thing goes for tubes. the source of the distortion is from the variations in output impedance at the crossover point. the crossover point with both SS and tubes is a very nonlinear region, and the best you can do is to minimize the nonlinearities. if the output devices overlap too much, the output impedance dips. if they don't overlap enough, the output impedance peaks. there's a "happy medium" between the two, but some crossover effect will still be present.

 

[AKSA]

"A water from ocean can be made to taste almost like a good Coke if to add a brown pigment in small quantities"

Alternatively:

'When we are thirsty, some of us like a little lime with our cold water'.

It is merely a preference for some people. It is not some ludicrous call to arms in a religious, purist war. I see a lot of engineering here; I see no interest whatever in psychoacoustics, how the human animal perceives sound, and what happens to it as it is processed within the ear canal and the brain.

It seems that amps are designed by tradition along zero distortion grounds. This is fine as far as it goes, but as is pointed out by the cognoscenti, any push pull amp, tube or SS, will generate a spray of artefacts at the crossover point. These are, by Fourier synthesis, high order, odd harmonics, as they are largely symmetrical about the crossing point. The objectionable nature of these harmonics is subliminal, not directly perceived, and it is almost impossible to rid them. Class A is the classic solution to this problem, but is highly inefficient. For real world Class AB amps, we cannot rid them, so we should mask them.

One very effective principle by which this can be done is to inject benign, low order harmonics. That tends to null them from a psychoacoustic POV. It is fighting fire with fire, put another way.

These days it is becoming clear that the THD is of limited use in identifying the sonic virtues of an amplifier. We need the distortion profile, the relative amplitudes of the first twenty or so harmonics.

Before people tear this argument apart in an apoplexy of engineering outrage, do it. Build such an amplifier, and do the AB testing. No amount of ridicule can substitute actually having done it and listened, and many outrageous ideas survive the test of time rather better than convention. A round earth, for one.

This is an opinion. It is not an armed threat, or a small bomb.


Hugh