• TalkBass has been independent since 1998. Add your voice.
    Create a free account to reply to discussions, view embedded media, and browse with fewer display ads.
    Join free Log in
    Want zero display ads or expanded classifieds tools? Compare plans.

Output transformer for SVT-CL that doesn't limit low frequencies

#84
Possibly would work, but the added feedback of ultra linear operation may create some instability issues unless the design considers these factors. If it were me, I wouldn't do it.
 
#85
upload_2015-8-11_22-38-13.png

Looks like clipping just starts to come into play but, around the zero line there is a strong artifact that may be easily overlooked. There is very well audible distortion due to bias crossover distortion. Although the trace does not indicate that but crossover distortion overwhelmes clipping. Spectrum is of odd order Harmonics
There is one interesting thing with crossover distortion. All of the Harmonics of a given audio program are bendend cause there is no more linear area. If there was only some clipping at the top then most of Harmonics remain mainly none distorted at this point.
IMO a good example where assumed time invariant systems become time dependent. That's very important for any amp modellings.

upload_2015-8-11_22-43-9.png

Increased clipping at the top. Looks like hard clipping but in addition there is now more amount of crossover distortion, easily seen as saddle points around zero line. This artifact is often mismatched with OT satuartion.
Spectrum is of prefered odd order Harmonics mainly.

upload_2015-8-11_22-49-31.png

Improved distortion,
output signal becomes an asymetrical characteristic thus output power is reduced slightly.
IMO an impressing behavior because everbody (sorry, let me correct, most folks) would by default expect more power the more the master or Volume knob is turned to full up
Spectrum is of odd order Harmonics and obvious amount of even order Harmonics.
 
Last edited:
#86
agedhorse said:
Saturating an audio transformer is not unlike current clipping an inductive load in a solid state amp where the VI protection is set with inadequate margin. About the most nasty sounding of all distortion. Made even more so because it doesn't happen at zero crossing which a (just a little) more benign.
Click to expand...
Yah, my Ampeg Micro VR into some 4 ohm loads nails that :vomit:. I could be wrong but that's my guess as to what's going on - sounds awesome on an 8 ohm cab...
 
#87
Crossover distortion at low volumes is very obnoxious because it dominates the signal, but at high volumes it just becomes a harmonic component as it's level will be pretty low relative to the rest of the waveform.
 
#88
Right.
IMO that's a somewhat fascinating phenomena how qualitativ "identical" properties are perceived totally different. In this case it all depends on the volume of appearance rather then qualitative characteristic of the artifact
 
#90
agedhorse said:
Crossover distortion at low volumes is very obnoxious because it dominates the signal, but at high volumes it just becomes a harmonic component as it's level will be pretty low relative to the rest of the waveform.
Click to expand...

This.

With serious "underbias" distinct crossover distortion would be constantly present in the signal and its magnitude would even be comparatively higher at low signal levels than at higher ones. It would sound like someone "layered" the signal with "buzz/fizz" and the effect would be more obnoxious the lower the volume.

But that's not what's happening there...

What's depicted in the pictures is a different mechanism of crossover distortion: When power tubes get sufficiently overdriven their current draw begins to turn very "asymmetric" and this shifts their DC offset levels. Since DC offset is equal to bias voltage the bias shifts "colder". Now crossover distortion begins to gradually appear when power tubes get highly overdriven AND for a moderate period of time.

So you have basically a time invariant mechanism that adds even more harmonics to power amp's clipping distortion, and the mechanism is also highly interactive with overall envelope of the signal. So it's a very "touch sensitive" process that adds a whole new harmonic pattern to the existing distortion. (Basically one can thus change the "timbre" of the clipping distortion by varying his playing dynamics). Since such process is merely adding more distortion to distortion the effect is indeed very different than that of crossover distortion which is constantly present and adds distortion to clean signals.

Also, whenever you push a tube power amp to such state you have other stuff happening as well: The screen grid current draw of pentode and tetrode output tubes suddenly ramps up (as the tube goes to griod conduction) and this will often modulate the screen voltage of the tube, which in turn results to gain compression within the tube. Plate voltage will also drop so the clipping threshold reduces even further.

Also, whenever one tube "deep saturates" in a push-pull amp the other will "cut off". Cutting off the current flow in a circuit that is transformer coupled to reactive load will create transient ringing, which then add high voltage "spikes" to corners of the clipped wave, corners of the crossover clipping, and corners of wave clipped by OT saturation. Basically to everywhere where current flow suddenly halts. In poorly designed systems these high voltage surges may even arc through and damage the amplifier itself.

SVT, on the other hand, is an interesting case because its output stage design tries to minimize many of these effects: The screen voltage is somewhat regulated to prevent screen current from modulating it, and the power tubes are driven by DC coupled low source impedance high current amplifiers (cathode followers) that actually handle driving power tube grids to conduction rather nicely. (For same reason there is not much of the aforementioned bias shift during excessive overdrive). Additionally the input has diode clipper to "clamp" input voltages below certain level so that it becomes somewhat impossible to drive stages of the power amp to "deep" saturation.

SVT power amp is a very good example of VERY different design goals and practices than what you usually encounter in generic tube guitar/bass amps.
 
  • Like
Reactions: cfsporn, nashvillebill, ThisBass and 1 other person
#92
In the meanwhile all ampeg svt series and v4b tube amps are equipped with these "peak level stopp" diodes.
The V4 amps never had cathode followers. IME there is no mandatory need for cathode followers on a 100-130 watt power amp.
 
#94
ThisBass said:
In the meanwhile all ampeg svt series and v4b tube amps are equipped with these "peak level stopp" diodes.
The V4 amps never had cathode followers. IME there is no mandatory need for cathode followers on a 100-130 watt power amp.
Click to expand...

It depends very much on what you are trying to accomplish. Not just in the linear range, but when it's intentionally driven into non-linear territory. Different circuit topologies have very different overload characteristics. Now wrap a feedback loop around these different topologoies and you have an entirely different animal. ;)
 
  • Like
Reactions: ThisBass
#95
agedhorse said:
It depends very much on what you are trying to accomplish. Not just in the linear range, but when it's intentionally driven into non-linear territory. Different circuit topologies have very different overload characteristics. Now wrap a feedback loop around these different topologoies and you have an entirely different animal. ;)
Click to expand...
No doubt about it.
That reminds me on a Boogie 2x100 watt stereo amplifier that showed nearly nothing of all these musically sounding "arifacts" discussed above. Indeed totally different designing goals.
 
#98
upload_2015-8-12_23-15-31.png


Currently in production V4B

Peak level blocking diodes
input stage
cathodyne PI
driver stage
pentodes

That's not so bad and IIRC very familar with svt design. Just thinking about the cathode follower?

There was a thread a couple of weaks before regarding these peak blocking diodes. The question was if the diodes impact the sound. IMO it depends. If you want to hear the impact you probably hear the diodes, if you don't think they impact the sound then you are also right. If there is an impact very likely the impact is subtile AND complements with different (other) effects.
 
Last edited:

Latest posts

Top Bottom
Back