feedback resistor and cap bypass to eliminate ice pick

[J M Fahey]

Fizz, fizz caps, icepick,etc. , are terms invented by bedroom rockers.

Guitar amps almost by definition and with very few exceptions are very bright, have bright caps either switchable or fixed, presence circuits which also enhance treble, treble skewed tone controls, 2500 to 3500 Hz 10-12dB peaks in their preferred speakers, etc.

Why?

Because all that sounds GOOD, with a sharp punchy guitar sound which cuts through when played at *band* levels, specially because "the" reference sound which all others must meet is a drummer, which is both very loud and liberally uses earsplitting cymbals, so Guitar sound was arrived at by Natural Selection, ´players chose some amps/guitars/speakers over others for good reason.

Now you play same amp/speakers *alone*, at bedroom levels and all that bright goodness becomes unbearable.

I´d say:bad use, not bad amp.

Some butcher amps and remove bright caps, add "conjunction filters" (a.k.a. snubbers) across OT primaries, bias EL84 with Leds, etc.

Why use tubes then?

They might straight use SS practice amps and call it a day.

[loudthud]

We're kind of in the dark here. Can anybody post pictures from a scope with and without fizz ? Is there a production amp or pedal that does it ? Can you record an amp that does it, then filter it out upon playback ? Let's be a little more analytical. Any threads on other forums ?

I just searched for "Fizz" on TAG and get 932 matches (63 pages) of hits.

[daz]

Fizz, fizz caps, icepick,etc. , are terms invented by bedroom rockers.

Guitar amps almost by definition and with very few exceptions are very bright, have bright caps either switchable or fixed, presence circuits which also enhance treble, treble skewed tone controls, 2500 to 3500 Hz 10-12dB peaks in their preferred speakers, etc.

Why?

Because all that sounds GOOD, with a sharp punchy guitar sound which cuts through when played at *band* levels, specially because "the" reference sound which all others must meet is a drummer, which is both very loud and liberally uses earsplitting cymbals, so Guitar sound was arrived at by Natural Selection, ´players chose some amps/guitars/speakers over others for good reason.

Now you play same amp/speakers *alone*, at bedroom levels and all that bright goodness becomes unbearable.

I´d say:bad use, not bad amp.

Some butcher amps and remove bright caps, add "conjunction filters" (a.k.a. snubbers) across OT primaries, bias EL84 with Leds, etc.

Why use tubes then?

They might straight use SS practice amps and call it a day.

I would agree with that because use my amps in both scenarios and really never seem to have any issues with fizz in a band setting. But at home it's a different story because at very low olume you lose the punch and cut in the high end and tend to turn the highs up creating the possibility of fizz. In fact, if you read my first post i said "Not really all that concerned (with fizz) because the amp isn't harsh, but i thought it would be a good idea for when i want a brighter sound to cut any ice pick that might occur with higher top end."
When at low home volume i use a lot more top end because of the relatively mushy sound at very low volume. In the band i usually have the treble presence and other switchable junk i have in the amp set for a much darker sound that at band volume isn't dark. But i DO like the amp to sound good in both scenarios and try to achieve that which at the moment i think i have for the most part. But the thread was more about just trying this because i always jump at the chance to try something in the back end of the amp in place of the same goal in the front, IE: cut top in the PI/output rather then in the preamp for example. Often trying to achieve the same goal with different methods often proves one works better then the other. Not that i'm telling you anything you don't know but just to note that i DO think of these things and often ask questions o0ut of curiosity rather than necessity.

[kg]

Those who care about fizz aren´t playing loud enough

100% this!

if you put a scope on your signal it's pretty obvious how massive the peak to average ratio of distorted tones are. those sharp spikey peaks are the higher frequencies and harmonics. when you turn up the volume the first thing to go into hard clipping is that higher order content, shifting the tone from treble to bass and rolling off that fizz.

this is happening in your speakers too... speakers have a f/r that looks more and more like a bell curve as you push them harder.

[Chuck H]

100% this!

if you put a scope on your signal it's pretty obvious how massive the peak to average ratio of distorted tones are. those sharp spikey peaks are the higher frequencies and harmonics. when you turn up the volume the first thing to go into hard clipping is that higher order content, shifting the tone from treble to bass and rolling off that fizz.

100% this to 100% that

This is one reason the amps designed with high gain preamps for "metal" usually have high power and don't push the power tubes that hard. The more stringent tone with peaks on either end of the clipped wave form is part of the sound. And also why there are guys that swear by clipping the power amp to get the tone they want where the clipped wave form is virtually flat. Clipping the power tubes nulls any preamp EQ effect and renders the tone stack to more of a 'distortion character' circuit where turning up the bass just adds flub rather than low end and turning up the treble adds harmonics rather than brightness.

this is happening in your speakers too... speakers have a f/r that looks more and more like a bell curve as you push them harder.

This (sort of) came up on another thread about this amp, but this aspect wasn't addressed. The inquiry was whether or not higher power rated speakers lack efficiency compared to lower power rated speakers for lower powered amplifiers. To which the answer was no. But your observation above may have some relevance to the matter as well. Implying that lower powered speakers may actually be LESS efficient (though subjectively better sounding) when pushed closer to their power rating. But given the nature of modern guitar speaker power ratings and that this is a twenty watt amp there aren't many choices.

[Helmholtz]

.. speakers have a f/r that looks more and more like a bell curve as you push them harder.

Is there any evidence by f/r measurements at speaker power limit?

[Helmholtz]

To add: I once asked electro-acoustics professor Manfred Zollner if speaker f/r significantly changes with signal level. The answer was: No.
Zollner has done many (guitar) speaker measurements and published the results in his books.

[Mick Bailey]

That's surprising, although I haven't read Zollner's books. I would have expected at least two factors that would affect response;

1. In a sealed cabinet the air behind the speaker has a greater damping effect on cone movement at higher signal levels compared to low levels, and affects lower frequencies to a greater degree due to the increased ratio of swept cone displacement compared to cabinet volume.
2. Xmax is given as the maximum extent of voice coil movement until a 10% deviation in linearity is measured. The fall-off in linearity affects lower frequencies more. Why would a fall-of in linearity at high signal levels not affect the response?

[Helmholtz]

That's surprising, although I haven't read Zollner's books. I would have expected at least two factors that would affect response;

1. In a sealed cabinet the air behind the speaker has a greater damping effect on cone movement at higher signal levels compared to low levels, and affects lower frequencies to a greater degree due to the increased ratio of swept cone displacement compared to cabinet volume.
2. Xmax is given as the maximum extent of voice coil movement until a 10% deviation in linearity is measured. The fall-off in linearity affects lower frequencies more. Why would a fall-of in linearity at high signal levels not affect the response?

I'm not an expert on speaker theory (Zollner certainly is).
You might ask your questions here: https://gitec-forum.de/gitec-community/fragen-stellen/

Ad 1: I understand that the air in a sealed cabinet (without acoustic damping material) has little damping (i.e. energy absorbing) effect, but I might be missing something.
Ad 2: No idea how increasing non-linearity affects f/r. If you're right, a speaker with a higher power rating should be affected less in the same amp.

My main question would be "how much?".

Both your points seem to refer to a change of bass response. Does it correlate with your listening experience? I think you would have to drive the speaker with clean program to exclude amp effects.
Also I think the OP is more concerned about treble reponse.
"kg" spoke about a bell shaped response. I interpret that as a relative drop at low and high frequencies.

Still need to see f/r measurements at full speaker power.
I'm an experimental physicist who likes to see data before digging into speculation.

[kg]

while i like data as much as the next guy (probably more) i am no longer equipped to collect that kind of information at a peer-reviewable level.

i am not familiar with zollner's work, but i believe we can all agree that musical instrument amplification and audio transduction is a different animal to hi fi. in the former, we are specifically requiring that the equipment modify the source tone in a variety of ways, whereas in the latter the opposite is desired.

mick hinted at this regarding xmax limits and what happens once that limit is exceeded. it seems reasonable that MI speakers are regularly pushed beyond the limits required by accurate T/S analysis. in other words, "there's a lot of nonlinear shit going on simultaneously."

excceeding xmax would mean that the magnetic field generated by the speaker coil has substantially left the magnet gap. then the question becomes what happens to F/R when the effective BL drops precipitously.

also, regarding comments on higher power speakers being affected less, i do not think that is necessarily the case, as excursions beyond xmax can be witnessed at vastly different power levels depending on frequency content and enclosure parameters.

studying/testing this might be easier if one were to do a F/R analysis with varying levels of DC offset which will statically push the cone to extreme excursions. this would replicate what happens with a HF signal riding on top of a LF fundamental (the LF in this case is 0hz). i cannot for the life of me imagine why the F/R would not be affected in these "extreme for hifi but mundane for MI" cases.

[Helmholtz]

The internet is full of seemingly plausible assumptions.
I need to see evidence.

None of my guitar amps seems to lose bass or treble at stage levels (of course bass response is always influenced by cabinet placement and room size).
But I might be fooled by my ears as (according to Fletcher-Munson curves) relative hearing sensiitivity at low and high frequencies increases at elevated SPL.
In the end it doesn't really matter.

[uneumann]

mick hinted at this regarding xmax limits and what happens once that limit is exceeded. it seems reasonable that MI speakers are regularly pushed beyond the limits required by accurate T/S analysis. in other words, "there's a lot of nonlinear shit going on simultaneously."

excceeding xmax would mean that the magnetic field generated by the speaker coil has substantially left the magnet gap. then the question becomes what happens to F/R when the effective BL drops precipitously.

also, regarding comments on higher power speakers being affected less, i do not think that is necessarily the case, as excursions beyond xmax can be witnessed at vastly different power levels depending on frequency content and enclosure parameters.

studying/testing this might be easier if one were to do a F/R analysis with varying levels of DC offset which will statically push the cone to extreme excursions. this would replicate what happens with a HF signal riding on top of a LF fundamental (the LF in this case is 0hz). i cannot for the life of me imagine why the F/R would not be affected in these "extreme for hifi but mundane for MI" cases.

I think the discussion may have mixed different things and good points are being made about all of them. However, it's worth saying clearly that F/R is being discussed in the context of other signals (bass or dc) and that means it's no longer F/R. F/R is measured with a single freq input. Once you mix other signals into the input (a chord for example), you can create all the xmax and NL issues already discussed and all of that has nothing to do with F/R anymore. You can perceive and measure the output of the mixed signal components (as well as the NL generated components) but again, this has nothing to do with F/R.

Someone earlier pointed out that peak xmax excursions can attenuate HF signals. That's exactly what happens with soft-clip circuit behavior, so one can reasonably expect it to happen in speakers too (although I claim no measured proof of that), however, again -- this has nothing to do with F/R. It's the behavior of a NL variable gain system.

[Helmholtz]

See here: https://www.gitec-forum-eng.de/wp-co...-materials.pdf

Fig.11.96 shows no tendency towards a bell shaped f/r when increasing power from 1W to 40W.

[Mick Bailey]

F/R is measured with a single freq input.

Wolfgang Klippel published a paper identifying the limitations of using a single frequency;
http://cyrille.pinton.free.fr/electroac/lectures_utiles/tech_haut-parleurs/Assessment_of_Voice_coil_peak_displacement_XMAX_02.pdf

Also more information on non-linearity (amongst other things) here;
https://www.klippel.de/know-how/literature/papers.html

[Helmholtz]

F/R doesn't need to be measured with single frequency input. An alternative would be noise analysis. Zollner uses both.