John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
Jump to Latest
#84,861
billshurv said:
You are saying that all LATFET amps have a soft bass. I can't accept that, not least because soft bass has no meaning beyond Frequency response abberations.
Click to expand...
I think the audible definition has to go a bit further than that.
When listening to wideband music/audio, one is always comparing different parts of the frequency spectrum to other parts. It is the memory of what you "think" that frequency comparison should sound like for continually changing signals.
I suspect it not down to ONLY frequency response as measured by a constant sinewave signal and taking a measurement.
I think there is a more subtle audible effect at play that becomes apparent in varying audio signals.

One possible technical explanation could be the effect of current delivery on the way a speaker driver gets fed with a voltage signal.
If a louder bass frequency is combined with a mixture of lower level mid and higher frequencies and the amplifier limits a bit of the current delivery that the speaker driver demands at that moment, then the audible effect could be a slightly reduced level of "APPARENT" bass level relative to the comparable other frequencies.
This effect would not be apparent at lower levels when lower currents are demanded. This effect would not show in a frequency response plot.
There are other possible explanations of why some amplifiers don't give comparable bass effect, even though they have an identical frequency response to other amplifiers that have full bass reproduction.

Again it comes down to what we measure and what we don't measure, either because we don't know how to measure that parameter, or that it is too diffficult to measure, or expose above some threshold.
 
Last edited:
#84,863
If you are hitting current delivery issues then you are operating the amplifier outside it's limits. But why should this affect a lateral rather than a vertical or a BJT? Only thing that i can immediately think of is different handling of thermals. But if you are getting to that point you either have a suboptimal thermal and SOA design or you are trying to arc weld an Apogee with a single pair of output devices.

Most intelligent people accept that clipping performance and recovery time can and will have an audible signature that will vary between topologies, but as long as you are a dB or so off that point you should be ok.
 
#84,864
I'm saying suggesting it is only a latmosFET problem.

I promote current delivery/capability as a major contribution to power amplifier performance.

For example I test an 8ohms rated amplifier into 4r0 and 2r67 dummy test loads and expect the maximum unclipped output voltage to droop within my set standards.
Others think my 3times maximum power delivery current is a nonsense requirement.
Maybe that's why I generally don't hear what I proposed as a possible explanation.
 
#84,866
vacuphile said:
Not really, noise is just the random part. You can distinguish a signal below random noise. Imagine a noise level just above your limit of perception, and a sine signal slightly below it. Without the noise, you would never be able to hear the signal. However, thanks to the introduction of noise, enough of the signal can be pushed above the threshold for your brain to distinguish the periodicity of the signal. Actually, you don´t even need a brain for this. I first came across it when I was studying neural networks a couple of decades ago. The artificial addition of noise can push some signals above detection thresholds and is thus used. There is some family relationship with dithering btw.

So, in stead of being buried under the noise, the noise can actually push signals above it. I don´t think it is a coincidence that ears are never completely silent, even in the absence of sound, and eyes always show noise, even in pitch black. Without it, our sight and hearing would be much less able to discern low level inputs.
Click to expand...

That s due to noise distribution, if the noise level is say 1mV and that the sine is 0.5mV it can be heard of course because the noise level in a band of 800Hz-1.2KHz is lower than 0.5mV, so no, the sine doesnt push neither is pushed but is simply of higher power within a given bandwith than said noise within the same bandwith.
 
#84,869
vacuphile said:
In the end it all came down to the speakers. Designed and build an active crossover to integrate some of the loudspeakers and amps I had and this was the real breakthrough. I have been running active systems for over 30 years now and am very disappointed that 'high end' still means trunks filled with coils and caps in speakers. The biggest and most obvious improvement that can be made to sound reproduction in the home is still largely neglected.
Click to expand...

Yes, active crossovers are the way to go, and I'm also surprised how few systems there are doing that. Fits will with dsp room eq too.
I think Bruno's latest class D offerings are active loudspeaker? Horrendously expensive probably...
 
#84,873
help me calculate this approximation?

Perhaps someone here can help me calculate this approximately?

Let's say that there is a driver that is spec'd at 0.1%THD @ 128db/1w/1m.
Assuming that our average listening level is 100dB/1w/1m (which would be somewhat loud in a typical room/system) what would the expected THD be at that level?

Thanks.
 
#84,876
billshurv said:
you need to chose a driver with better specs. Hint. Speakers are non-linear.
Click to expand...

billshruv are you serious?? say what!!!???

What will the THD for YOUR drivers look like at 128db SPL????

Aside from the smoke coming out...😱


AND, assuming they are linear, what number would you come out with??
What's the spec on the typical high performance HF driver that you know about, please state.
Given that they are non-linear, what result do you expect?
Substantially LESS THD at 100dB SPL for my "imaginary" driver?? By how much?
How would that number compare to the typical HF driver... (especially at the same 100dB SPL) curious, I am.
 
Last edited:
#84,877
SY said:
Did the words "frequency response" not register?
Click to expand...

OH, you mean measured response of the speaker(s) then?

I thought you were measuring amplifiers.

So, I guess all amps with the same DF sound the same WRT subjective bass impression. And, how much "frequency response" deviation or delta (difference between two amps) will be audible?

Let's get some numbers out on the table here. I don't want to have to guess about this stuff, when I putting together my new system.


_-_-bear
 
#84,879
billshurv said:
A single point measurement is pointless. As is replying to most of your posts. You are gagging for an argument I can see rather than a discussion.
Click to expand...

Explain.

Pick any sort of measurement you wish.
Specify a measurement as you think is best.
Say something of substance.
Enlighten, amplify.

I'm asking questions, if you have no way to provide input, explanation or expansion, then why post suggesting I am trying to argue? Do you think there
is something to argue about in asking a question about distortion figures for an imaginary driver?? Why do you believe this?

Why do you think that my general post is aimed at you??
It's not.

Any time things come down to anything real-world I get nothing except for avoidance. Why is that? If my questions were so "stupid" anyone could simply respond and put it to rest immediately.

The aim here is to start with something simple that relates back to things that were claimed by various posters/contributors earlier. Perhaps you've figured out where the linkage is, and don't want to be called out on these issues. Perhaps.
 
Status
Not open for further replies.
Top Bottom