Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Richard Stringer]

I've always kind of wondered why toroidal transformer amplifiers are generall better for subs than switch mode power sply amps, and I kno, or rather think, that amplifiers like say the Crown XTI series aren't great on subs because their power supply isn't big enough and can't extract enough power out the wall to cope with the sheer power thats needed for running subs with styles of music that have sustained basslines. Am I right?

A Peavey tech guy told me that generally a lot of switch mode amps are rated in peak power and not continuous average power because the switch mode amplifier in general isn't capable of extracting as much power as toroidal amps, and so rate their switch mode amps in peak ratings because if they rated the actual continuous average power the power output figures, meaning specs, would look pretty shite compared to old school toroidal amplifiers like the Crest CA18, Crown MA5002, Crest 9001 etc..I was also told that older amplifiers have either bigger power caps or more of them so to store enough power ready for when it's needed.

What's your thoughts guys on why in general, toreoidal amps are better for sub than switch mode amps?

[Bob Leonard]

Richard,
It's not the toroid alone that will determine an amps capability. There are far more people on this site more qualified to discuss the subject than I, however, any amplifier will only be as good as the design of it's power supply regardless of type. Big iron is the way I go, but let's not fool ourselves either. You may have a point with entry level amplifiers, but I would say that if Lab Gruppen said they could get 1000 watts out of a battery you could bet your pay check on it.

[Ivan Beaver]

As Bob said, there is a lot more to it than just the transformer or the filter caps.

You cannot compare an entry level cheap amp (such as the XTI) to a Crown MA5000 or Crest 9001.

That would be like comparing a Chevette to a Corvette and trying to determine if the carb is the difference in performance.
 
There is A LOT more to it than that.

If you really want to compare-start with amps that are in the same price/wattage range and see what you end up with.

Switch mode power supplies recharge the caps faster than analog supplies do, so they can get away with smaller caps.

Yes most of the "big dog" amplifiers cannot sustain there rated output for real long.  You cannot put out any more power than you draw from the wall.  But that is "over time".  Large peaks are available with the "switching amps" that are not available with basic transformer amps (toroidal or regular).

There are advantages and disadvantages to each type-"better" means different thing to different people.

As was said many years ago (long before the big dog amps of today) it was said by one of the audio giants  "What we really need is a 100 watt amp that can pass 10,000 watt peaks".  20dB of headroom if you don't want to do the math.

Now that is for "normal" music.

If you are looking for something to be able to sustain long bass notes- you HAVE to do the following first.  Determine how long is "long"? 1 second-10 seconds-1 minute?  How much power do you really want to send to your loudspeakers (what is the point of power compression on your loudspeakers), how much money do you have?

Then you can start your search.

[John Roberts {JR}]

I've always kind of wondered why toroidal transformer amplifiers are generall better for subs than switch mode power sply amps, and I kno, or rather think, that amplifiers like say the Crown XTI series aren't great on subs because their power supply isn't big enough and can't extract enough power out the wall to cope with the sheer power thats needed for running subs with styles of music that have sustained basslines. Am I right?

Well PS matters but older or type of transformer used has little to do with it. FWIW switching PS often use torroidal transformers too. just tiny ones.

A Peavey tech guy told me that generally a lot of switch mode amps are rated in peak power and not continuous average power because the switch mode amplifier in general isn't capable of extracting as much power as toroidal amps, and so rate their switch mode amps in peak ratings because if they rated the actual continuous average power the power output figures, meaning specs, would look pretty shite compared to old school toroidal amplifiers like the Crest CA18, Crown MA5002, Crest 9001 etc..I was also told that older amplifiers have either bigger power caps or more of them so to store enough power ready for when it's needed.

Nah... and this is a pretty old story, there are slightly different design trade offs wrt reservoir caps on the mains side of a switching supply. Back decades ago appropriate HV capacitors were harder to source, expensive and large. Some early designs wimped out on mains caps and the reputation has stuck with all of them.

Modern amps keep getting better as all technology gets better, but there is no reason why any technology properly executed will ever behave differently than any other technology. If a technology is better or worse at some single aspect, the design engineer just needs to make adjustments for that technology. Properly executed is the magic phrase, and there are numerous examples of amps using all technologies that perform better or worse than typical.

What's your thoughts guys on why in general, toreoidal amps are better for sub than switch mode amps?[/font][/size]

I don't have any thoughts on why, because I don't believe they are better. There are only better and worse individual design executions.

JR

PS: be careful about listening to Peavey guys, especially ex-Peavey guys.. 

[Peter PAPP]

I've always kind of wondered why toroidal transformer amplifiers are generall better for subs than switch mode power sply amps, and I kno, or rather think, that amplifiers like say the Crown XTI series aren't great on subs because their power supply isn't big enough and can't extract enough power out the wall to cope with the sheer power thats needed for running subs with styles of music that have sustained basslines. Am I right?

A Peavey tech guy told me that generally a lot of switch mode amps are rated in peak power and not continuous average power because the switch mode amplifier in general isn't capable of extracting as much power as toroidal amps, and so rate their switch mode amps in peak ratings because if they rated the actual continuous average power the power output figures, meaning specs, would look pretty shite compared to old school toroidal amplifiers like the Crest CA18, Crown MA5002, Crest 9001 etc..I was also told that older amplifiers have either bigger power caps or more of them so to store enough power ready for when it's needed.

What's your thoughts guys on why in general, toreoidal amps are better for sub than switch mode amps?

Operational parameters of toroidal amps varies by the power line changes while most of the quality amps with regulated switchmode PSU are virtually independent from line.
Some of people like DJs better prefer fat "floating" bass which effect created by undamped resonating membranes of speakers or even amps clipping hard.
Many circuitry of conventional linear AB,H,TD class amplifiers have lower control of voice coil movements close to the rail Voltages. When the amp loses control of speaker at high levels cycle by cycle it could create extra bass sensation with freely resonating speakers.
Of course what you hear is NOT the original sound rather than something is "added"...
From the professional's view this effect is not good because when lots of speakers used in parallel freely resonating membranes without control mean random phase radiators which usually cancels each other. So bigger systems needs the tightest possible control of voice coils in the right phase otherwise the common output suffers.
The best amplifiers do not use the maximum range of rail Voltages, it means even at the peaks the load can be damped and the output impedance does not drop by pushing the rails higher.
The price of this solution is you need to have at least several tens of Volts above the actual output range. Most of the conventional amps can not work this way due to the unaffordable extra losses... some of the switching amps have this feature and gives tight, very precise phase controled low range for those who like it.

Péter

[Richard Stringer]

Thanks guys, that explains a lot to me, trying to understand it is hard but i'm learning about voltage slowly. I really apreciate you helping me on this matter.

Ivan,
I wasn't comparing cheaper amps with more expensive amps so much, but more comparing the switch moce supply to the toroidal supply in general.

[Ivan Beaver]

Thanks guys, that explains a lot to me, trying to understand it is hard but i'm learning about voltage slowly. I really apreciate you helping me on this matter.

Ivan,
I wasn't comparing cheaper amps with more expensive amps so much, but more comparing the switch moce supply to the toroidal supply in general.

If you compare the itech amps to the older "better" amps, it would be a better comparison, than using the XTI's.

Cheap amps will have less performance-no matter what the design.

I know many people think that they are getting the Itech performance for an XTI price.  They are not the same quality/performance etc.

[Richard Stringer]

If you compare the itech amps to the older "better" amps, it would be a better comparison, than using the XTI's.

Cheap amps will have less performance-no matter what the design.

I know many people think that they are getting the Itech performance for an XTI price.  They are not the same quality/performance etc.

Oh yeah I definately know that, i've heard the XTI series amps on subs a few times and not once have I been impressed. I heard a QSC RMX5050 though I was more impressed with it.

[Ivan Beaver]

Oh yeah I definately know that, i've heard the XTI series amps on subs a few times and not once have I been impressed. I heard a QSC RMX5050 though I was more impressed with it.

Just to be clear-I said what I did, because in  your origional post you compared a XTI to much better quality amps.

Had you compared a Lab Gruppen to those amps-your opinion would probably be different-or any other good quality "switcher".

But comparing a cheap entry level amp of one type of design to a much better quality amp of a different design is not exactly comparing.

For example, it could be completely turned around if you said "Why are the switching amps (like a Lab Gruppen-QSCV PL3 series-Itechs etc) so much better sounding than a standard transformer (toroidal or otherwise) like the Samson whatever #?"

Well, when you compare a good amp to a cheap amp-that is what you get.

[John Roberts {JR}]

Operational parameters of toroidal amps varies by the power line changes while most of the quality amps with regulated switchmode PSU are virtually independent from line.

Many switching power supplies used in power amplifiers are not regulated. They simply chop the mains voltage at a high frequency so they can use a small transformers, and those supply voltages will sag when the mains voltage sags. 

Most PFC (power factor corrected) amps OTOH are regulated, so they hold voltage and in addition to that, enjoy the extra benefit of spreading the current draw out over the entire mains waveform so literally can pull more power from a given distribution.

Some of people like DJs better prefer fat "floating" bass which effect created by undamped resonating membranes of speakers or even amps clipping hard.
Many circuitry of conventional linear AB,H,TD class amplifiers have lower control of voice coil movements close to the rail Voltages. When the amp loses control of speaker at high levels cycle by cycle it could create extra bass sensation with freely resonating speakers.

I am not familiar with this mechanism or what DJs prefer. Rumor is they like loud and don't mind clipping to get there. Most amps have clamp diodes across the power devices so even if clipping, back EMF from the loudspeaker can't swing more than a diode drop above the rail voltage and will see a low impedance at that point. 

Of course what you hear is NOT the original sound rather than something is "added"...
From the professional's view this effect is not good because when lots of speakers used in parallel freely resonating membranes without control mean random phase radiators which usually cancels each other. So bigger systems needs the tightest possible control of voice coils in the right phase otherwise the common output suffers.

again I am not familiar with this phenomenon.

The best amplifiers do not use the maximum range of rail Voltages, it means even at the peaks the load can be damped and the output impedance does not drop by pushing the rails higher.

Many analog amps use anti-sat diodes to prevent the output power devices from hard saturation. This is to speed up recovery time after clipping events (saturated power devices, stick to the rails and take more time to come out of clipping). So these will sound slightly cleaner and better able to ignore brief transients that clip. But clipping alters the signal output, so never clipping is always better.

The price of this solution is you need to have at least several tens of Volts above the actual output range. Most of the conventional amps can not work this way due to the unaffordable extra losses... some of the switching amps have this feature and gives tight, very precise phase controled low range for those who like it.

Péter

It is very impractical and generates way too much waste heat to have and not use, tens of volts of power supply rails.  By switching amps are you speaking of class D amps with switching output stages or linear amps with switching power supplies? Class D amps use saturated output devices so all of their PS voltage is always used, switching supply amps come in many flavors. 

Oversized amps that never clip can sound good, because they never clip.

JR

[Iain.Macdonald]

The price of this solution is you need to have at least several tens of Volts above the actual output range. Most of the conventional amps can not work this way due to the unaffordable extra losses... some of the switching amps have this feature and gives tight, very precise phase controled low range for those who like it.

Péter

It is very impractical and generates way too much waste heat to have and not use, tens of volts of power supply rails.  By switching amps are you speaking of class D amps with switching output stages or linear amps with switching power supplies? Class D amps use saturated output devices so all of their PS voltage is always used, switching supply amps come in many flavors.

Oversized amps that never clip can sound good, because they never clip.

JR

+1 to Peters post.

JR, I am not sure whether you are agreeing or disagreeing with the quote, especially the first sentence. What about tracking SMPS? The rails are usually held at very low level.

Iain.

[John Roberts {JR}]

+1 to Peters post.

JR, I am not sure whether you are agreeing or disagreeing with the quote, especially the first sentence. What about tracking SMPS? The rails are usually held at very low level.

Iain.

I am not sure either? What technology he is speaking of? I expect I will find out.   A tracking supply that is always tens of volts hotter than the output seems kind of wasteful.

Back in the '80s I designed a small amp with a tracking supply, that was capable of 4x peak, the continuous output power. One quirk of my design, that I suspect other similar designs may share is that the rate of change of the tracking supply directly impacted signal purity, so for acceptable distortion the power supply rate of change was slowed down. To keep up with signal transients, the supply had to be programmed with some HF lead and safety margin so the PS will be there in time to support fast audio signal transients. Of course this reduces the efficiency benefit available. Selectively making the PS anticipation lead, sensitive to rate of change allows for decent efficiency when playing LF content, but it was a juggling act between several variables. 

I suspect the technology available in the decades since I last played with this is better, both regarding PS speed and output stage rejection of PS artifacts (Miller capacitance in bipolar output devices was a vector for distortion from high PS edge rates). I used a pretty low tech approach for my variable rail and only made modest power, but it was enough to reveal the difficulty to execute this approach well (there is no free lunch).  I would expect similar issues in a big dog tracking rail amp to optimize for both low distortion and fast transient response (like maybe a bunch of extra JIC volts in the tracking PS). Tone burst testing should reveal any issues from a PS not keeping up. Steady state sine wave testing should be relatively easy lifting for this approach. 

I am generally supportive of this technology, if that is what we are talking about (analog amp with switching tracking supply). it wasn't clear from the post. 

Another amp idea I never got around to breadboarding is to use a variable intermediate rail in a class G/H amp. If this lower rail is regulated to deliver high  efficiency based on actual waveforms, and large transients could simply be handled by the high rail using the existing rail commutation inherent in that topology, we could milk some more efficiency from the good old class G/H without any performance trade off (just some extra PS complexity cost). This intermediate rail could be back slope regulated (kind of like a light dimmer once or twice every mains cycle) so relatively cheap and low tech. I haven't been in the amplifier trenches for a while so my apologies if somebody is using this variant too.

This is all kind of mooted by class D getting better and cheaper. A tracking switching supply, wrapped around a linear amp, could also be thought of as a variant  class D amp that uses the linear amp as an expensive output filter.

One more wing nut idea of mine, that I never bothered to breadboard was to drive one speaker lead with a dirty class D amp, and then drive the other speaker lead with a low voltage correction signal to subtract out distortion from the cheap class D amp, but this too is impractical. Cheaper now to just make the class D amp better.

Bigger amps that don't clip are always good.

JR

[Richard Stringer]

You're able to get a cleaner sound and not have to clip an amplifier by using an amplifier which can output 2x the continuous average power rating of your speakers or subs, and then operating the system at a lower level though aren't you, headroom it's called isn't it. I know for dance music a lot of engineers have to be careful because of the massively compressed tracks with sustained sub basslines, so they use amplifiers which use no more than 2x the speaker cab's rated continuous power. Am I right in saying that with rock music and country music because it's live sound and no pre-recorded and compressed to death, that you guys can get away with using amplifiers capable of outputting maybe 3x the speaker cab's continuous rating so you have masses of headroom? Or do you guys, because you can afford the speaker cabs, bring quite a bit more system than you'll need so you can achieve the same spl levels and not have to run the system so hard?

[Stu McDoniel]

I've always kind of wondered why toroidal transformer amplifiers are generall better for subs than switch mode power sply amps, and I kno, or rather think, that amplifiers like say the Crown XTI series aren't great on subs because their power supply isn't big enough and can't extract enough power out the wall to cope with the sheer power thats needed for running subs with styles of music that have sustained basslines. Am I right?

A Peavey tech guy told me that generally a lot of switch mode amps are rated in peak power and not continuous average power because the switch mode amplifier in general isn't capable of extracting as much power as toroidal amps, and so rate their switch mode amps in peak ratings because if they rated the actual continuous average power the power output figures, meaning specs, would look pretty shite compared to old school toroidal amplifiers like the Crest CA18, Crown MA5002, Crest 9001 etc..I was also told that older amplifiers have either bigger power caps or more of them so to store enough power ready for when it's needed.

What's your thoughts guys on why in general, toreoidal amps are better for sub than switch mode amps?

Here is a nice paper on amplfier power and loudspeakers

http://www.eaw.com/Info/EAW/Technical_Papers/AmplifierPower.pdf

[Richard Stringer]

Thanks Stu, I knew about headroom and power handling anyway, I was just wondering if you guys bring more rig than you need or use more powerful amplifiers to achieve the headroom.

[Peter PAPP]

+1 to Peters post.

JR, I am not sure whether you are agreeing or disagreeing with the quote, especially the first sentence. What about tracking SMPS? The rails are usually held at very low level.

Iain.

I meant two different state of operation when reactive energy occurs, i found very important how the amplifier recycles (if able to do that and not dissipates) reactive back energy.

1. The amplifer controls the reactive amount of energy by forcing the momentary output voltage on the voice coil. If you have a good class-D amp changes operation from buck mode--->>boost mode and pushes back energy into its DC rail capacitors. The Voltage of DC rail increases but the actual value of coil Voltage NOT. This amount of recycled energy could be reused again the next buck-cycle.

2. In a class-AB linear amp the reactive energy usually creates extra heat losses, sometimes over the safe dissipation factor of output devices. In a Class-H amp there are separation diodes between "steps" which effectively blocks the reverse energy flow.
If the AB class amp has enough momentary headroom in heat dissipation on its output devices It would be able to perfectly handling the situation and would be the best sounding solution despite of heavy heat losses. However I have not seen any of those analog amps.

The general problem is handling speakers as a resistor while they are not. Same is the situation with the tracking amps (class-TD) where switching preregulator used for coarse voltage production of the fine class-AB endstages.
This method has even lower losses than Class-H and distortion values are extremetly small with RESISTIVE loads.
Since these circuitry use almost always a simple buck-type switching preregulators as the reflected voltage (due to the reactive back energy) exceeds the coarse Voltage value (typically ~10V) the amp completely loses control of its output.
The best solution would be extending the simple buck to a synchronous buck converter which is able to recycle reactive energy by mode changes however it would makes twice as complicated the already two stage circuitry.

About the clipping of bigger amps: I think this is a very important feature that NEVER CLIP. When the signal reaches output rails means loosing control and the clipped signal sharply increases its rms value killing voice coils by elevated heat. When the amp is really big this effect is even more faster.. so for reliability aspects proper built-in limeters are very important.

P.

[John Roberts {JR}]

I am familiar with the hypothetical premise of back EMF from loudspeakers driving back into power amp outputs. Mostly from advertising material associated with class D amps.

I am not familiar with solid characterizations of how often this occurs with normal signal playback. I have a judgement about magnitude of load angle (from designing amps that didn't melt), but no empirical data to provide here.

If the triggering event for this back EMF is amplifier clipping, I think we all can agree that this is best avoided, and behavior around clipping is underestimated as a source of sonic differences between sundry amps.

Perhaps some serious speaker guys reading this can contribute some practical input about reactive energy available from their loudspeakers. 

JR

PS: I notice you are making a 3 phase powered amp. That is sweet if you can do it, but it seems with PFC you forfeit some of the benefit from the overlapping phases.

[Scott Wagner]

Thanks Stu, I knew about headroom and power handling anyway, I was just wondering if you guys bring more rig than you need or use more powerful amplifiers to achieve the headroom.

Rig size decisions are based more upon coverage.  SPL considerations are usually considered in rig selection (ie: which system to bring), as opposed to how many speakers to deploy.  The headroom increases are more a function of having a less compressed signal - less average levels equals more capacity for peaks.  This is precisely why DJs have a reputation for releasing the magic smoke.  The exact same rig can get much higher levels for live music that it can for over-compressed DJ tracks.

Scott Wagner
Big Nickel Audio

[drew gandy]

Do you make any amplifiers for export to the US (ie 120VAC operation)? 

[Richard Stringer]

Rig size decisions are based more upon coverage.  SPL considerations are usually considered in rig selection (ie: which system to bring), as opposed to how many speakers to deploy.  The headroom increases are more a function of having a less compressed signal - less average levels equals more capacity for peaks.  This is precisely why DJs have a reputation for releasing the magic smoke.  The exact same rig can get much higher levels for live music that it can for over-compressed DJ tracks.

Scott Wagner
Big Nickel Audio

Yeah most dj's I know run levels into the red lights, I personally think it's so stupid, I mean these dj's either don't know you get a cleaner sound and better sound quality when you don't clip the system, or they just couldn't give a sh*t. Personally I think it's because most couldn't give a sh*t. I've been a dj since 1993 and I never clip a system, I would rather have better sound quality than more spl.

[ProSoundWeb Community]