Small Class-D Amp Heads. What validates the 'Warm' sound ?

[Killing Floor]

Warmth is in the nut width.

 

[VolverseLoco]

I'm surprised no one has mentioned the Genz Benz Streamliner 600 yet, a Class D amp with three tubes in the preamp section. I can't imagine anything but " warmth " coming out of that head.

I was trying to keep my question based around a 300-500 watt Class-D all solid state amp head with roughly the physical size of the Tone Hammer or Mark Bass head in that range. I looked at those Genz Benz's with the tubes and I guess I would have to find one used it I ever wanted one as the stores say its no longer available. Same thing now with the TC BQ-500 head, there isn't one singe one on E-Bay and Sweetwater says they just keep whatever they have on it up on their website for 'reference purposes'. Like the BQ-500 just dissappeard of the face of the Earth.

 

[VolverseLoco]

Warmth is in the nut width.

Yeah, a constant 98.6 degrees...right..

 

[Phaidrus]

Warmth is in the nut width.

Could you please elaborate on that.

 

[dave sands]

Wow, lots of "Feel" answers here. I'd say that the sound of the instrument you're playing should be what comes out of the power amp. That's a very subjective thing. But power amps are many and varied!
Some folks like tubes in class B, some like tubes in class A ( Me!!), some folks like solid state in a class A, while others prefer class AB, and others the class D. But doesn't look like there are many actual descriptions of the circuitry that will enable y'all to understand the why of the characteristics. Aside from a treatise on basic linear audio amplification, suffice it to say that class D amplification has been around since the early 60s, but was too expensive to implement easily until the advent of power MOS FETs (metal-oxide-semiconductor-field-effect-transistors). While MOSFETs absolutely suck at analog amplification, they are really good at fast switching. That's what class D is, a means of amplifying by switching the transistors between the power supply rails (+ and -) at a very high frequency, like 100KHz, and modulating the pulses that are created with the analog signal you want to amplify, in this case, my 1972 Fender Jazz. This creates a high power PWM (pulse-width modulated) signal that carries the sounds of my bass and the pulses are extracted and put into the speaker by a simple low pass filter that cuts out everything above about 17.5 KHz, like the switching freqs, and leaves only the nice amplified bottom end. What makes a Class D amplifier sound warm or harsh or clanky or... whatever, is that low pass filter. How well it is designed is how warm the sound will be. So you have to ask questions like, what is the Q (quality) factor of the filter, is it Butterworth or Tschebyschev and most importantly, how many orders is it. Filter design is based on mathematical polynomials that can have an infinite number of orders. A filter that has more orders is going to sound more natural, (similar to the way 16 bit digital audio sounds far better than 8 bit digital audio does) so always ask for an amplifier with a 5th order LP filter on the output of the class D. It's gonna cost more, but usually you get what you pay for...

 

[Phaidrus]

Wow, lots of "Feel" answers here. I'd say that the sound of the instrument you're playing should be what comes out of the power amp. That's a very subjective thing. But power amps are many and varied!
Some folks like tubes in class B, some like tubes in class A ( Me!!), some folks like solid state in a class A, while others prefer class AB, and others the class D. But doesn't look like there are many actual descriptions of the circuitry that will enable y'all to understand the why of the characteristics. Aside from a treatise on basic linear audio amplification, suffice it to say that class D amplification has been around since the early 60s, but was too expensive to implement easily until the advent of power MOS FETs (metal-oxide-semiconductor-field-effect-transistors). While MOSFETs absolutely suck at analog amplification, they are really good at fast switching. That's what class D is, a means of amplifying by switching the transistors between the power supply rails (+ and -) at a very high frequency, like 100KHz, and modulating the pulses that are created with the analog signal you want to amplify, in this case, my 1972 Fender Jazz. This creates a high power PWM (pulse-width modulated) signal that carries the sounds of my bass and the pulses are extracted and put into the speaker by a simple low pass filter that cuts out everything above about 17.5 KHz, like the switching freqs, and leaves only the nice amplified bottom end. What makes a Class D amplifier sound warm or harsh or clanky or... whatever, is that low pass filter. How well it is designed is how warm the sound will be. So you have to ask questions like, what is the Q (quality) factor of the filter, is it Butterworth or Tschebyschev and most importantly, how many orders is it. Filter design is based on mathematical polynomials that can have an infinite number of orders. A filter that has more orders is going to sound more natural, (similar to the way 16 bit digital audio sounds far better than 8 bit digital audio does) so always ask for an amplifier with a 5th order LP filter on the output of the class D. It's gonna cost more, but usually you get what you pay for...

Thanks for the input. Here's a simplified explanation of the various amp classes.

 

[spatters]

A filter that has more orders is going to sound more natural, (similar to the way 16 bit digital audio sounds far better than 8 bit digital audio does) so always ask for an amplifier with a 5th order LP filter on the output of the class D.

A bass speaker with a frequency response that falls off at 5K, and is zero above 12K, isn't going to tell you the difference between a second and fifth order filter above 20K. The primary reason to filter the output is because otherwise you're generating massive amounts of RF interference: the secondary reason is excessive heat dissipation in the speaker. Small Class D amplifiers, such as in Bluetooth speakers, often have no output filtering at all, and the typical Class D output filter is a simple second-order Butterworth.
https://www.ti.com/lit/an/slyt198/slyt198.pdf
Optimizing Class-D Audio Amplifier Output Filters

For another long, technical discussion, go here:
Link Removed

IIRC Chebyshev has too much ripple in the passband to use for audio, which is why Butterworth is universally used for audio. Frankly, I suspect that you're BSing.

 

[agedhorse]

Wow, lots of "Feel" answers here. I'd say that the sound of the instrument you're playing should be what comes out of the power amp. That's a very subjective thing. But power amps are many and varied!
Some folks like tubes in class B, some like tubes in class A ( Me!!), some folks like solid state in a class A, while others prefer class AB, and others the class D. But doesn't look like there are many actual descriptions of the circuitry that will enable y'all to understand the why of the characteristics. Aside from a treatise on basic linear audio amplification, suffice it to say that class D amplification has been around since the early 60s, but was too expensive to implement easily until the advent of power MOS FETs (metal-oxide-semiconductor-field-effect-transistors). While MOSFETs absolutely suck at analog amplification, they are really good at fast switching. That's what class D is, a means of amplifying by switching the transistors between the power supply rails (+ and -) at a very high frequency, like 100KHz, and modulating the pulses that are created with the analog signal you want to amplify, in this case, my 1972 Fender Jazz. This creates a high power PWM (pulse-width modulated) signal that carries the sounds of my bass and the pulses are extracted and put into the speaker by a simple low pass filter that cuts out everything above about 17.5 KHz, like the switching freqs, and leaves only the nice amplified bottom end. What makes a Class D amplifier sound warm or harsh or clanky or... whatever, is that low pass filter. How well it is designed is how warm the sound will be. So you have to ask questions like, what is the Q (quality) factor of the filter, is it Butterworth or Tschebyschev and most importantly, how many orders is it. Filter design is based on mathematical polynomials that can have an infinite number of orders. A filter that has more orders is going to sound more natural, (similar to the way 16 bit digital audio sounds far better than 8 bit digital audio does) so always ask for an amplifier with a 5th order LP filter on the output of the class D. It's gonna cost more, but usually you get what you pay for...

Man, there's a whole lot of fallacies and misleading if not inaccurate and incorrect power amp info. in this post IMO. I will try to address in order:

1. Regarding power amps, I would enjoy being enlightened about what power amps (tube or otherwise) are class A??? I think you will find that there are no true class A power amps above say 20 watts (with a couple of exceptions in the tweaker hi-fi world). Virtually all linear power amps that are used in commercial bass amps are class AB or class B. I can't think of a single exception.

2. MOSFETs absolutely suck for linear amplification? I don't think so, numerous companies (not just Mesa) have produced very highly regarded amps using MOSFETs of all types. I have used lateral MOSFETs for years but stopped using them because the availability of production quantities has become a challenge and screening for uniformity was a pain, not for the possible advantages to sound quality. Hartke, Ampeg, Ashdown, Trace Elliott, MarkBass and Warwick (just off the top of my head) have had successful, well received MOSFET power amp based products.

3. Warmth is generally not related to class D output filters, most extend pretty flat out past 18kHz. This can be easily see when using a cabinet without a tweeter, even with an extended range driver there's nothing meaningful being reproduced a full octave below the cutoff.

4. The order of the output filter is very limited by practical considerations. 2nd order is the most common, I have seen a couple of 3rd order filters but I can't say that in the hundreds of designs I have looked at over the years that I have ever seen more than that though 4th order is possible. I have never seen of, nor heard of a class D amplifier with a 5th order LPF.

5. A filter with more orders is going to sound more natural? How would you explain this? Your bit depth example makes no sense in context of filters, it's entirely unrelated.

 

[musicman556]

Man, there's a whole lot of fallacies and misleading if not inaccurate and incorrect power amp info. in this post IMO. I will try to address in order:

1. Regarding power amps, I would enjoy being enlightened about what power amps (tube or otherwise) are class A??? I think you will find that there are no true class A power amps above say 20 watts (with a couple of exceptions in the tweaker hi-fi world). Virtually all linear power amps that are used in commercial bass amps are class AB or class B. I can't think of a single exception.

2. MOSFETs absolutely suck for linear amplification? I don't think so, numerous companies (not just Mesa) have produced very highly regarded amps using MOSFETs of all types. I have used lateral MOSFETs for years but stopped using them because the availability of production quantities has become a challenge and screening for uniformity was a pain, not for the possible advantages to sound quality. Hartke, Ampeg, Ashdown, Trace Elliott, MarkBass and Warwick (just off the top of my head) have had successful, well received MOSFET power amp based products.

3. Warmth is generally not related to class D output filters, most extend pretty flat out past 18kHz. This can be easily see when using a cabinet without a tweeter, even with an extended range driver there's nothing meaningful being reproduced a full octave below the cutoff.

4. The order of the output filter is very limited by practical considerations. 2nd order is the most common, I have seen a couple of 3rd order filters but I can't say that in the hundreds of designs I have looked at over the years that I have ever seen more than that though 4th order is possible. I have never seen of, nor heard of a class D amplifier with a 5th order LPF.

5. A filter with more orders is going to sound more natural? How would you explain this? Your bit depth example makes no sense in context of filters, it's entirely unrelated.

I don't pretend to understand much of these technical replies from Agedhorse, but MAN I love reading them!

 

[agedhorse]

A bass speaker with a frequency response that falls off at 5K, and is zero above 12K, isn't going to tell you the difference between a second and fifth order filter above 20K. The primary reason to filter the output is because otherwise you're generating massive amounts of RF interference: the secondary reason is excessive heat dissipation in the speaker. Small Class D amplifiers, such as in Bluetooth speakers, often have no output filtering at all, and the typical Class D output filter is a simple second-order Butterworth.
https://www.ti.com/lit/an/slyt198/slyt198.pdf
Optimizing Class-D Audio Amplifier Output Filters

For another long, technical discussion, go here:
Link Removed

IIRC Chebyshev has too much ripple in the passband to use for audio, which is why Butterworth is universally used for audio. Frankly, I suspect that you're BSing.

I agree.

The EDN article is very well written and presented also.