Single-ended triode amps

[vrao]

My understanding about amplifiers

1. Current vs voltage school of thought - effects the speakers differently.

2. Power - a dark kept secret. Power rating are for a single signal. And power response drops with a complex signal. Power reserve can be "relatively" addressed for amplifiers with less than 60W, without affecting the outgoing signal. Add a power robbing crossover and more variables in the mixture.

Relax response: a very primitive description .... when the necessary parameters are met by the human auditory complex, brain will relax. Most will go to sleep ....

Because most are sleep deprived .... and once the brain feels safe, it will proceed to the next important thing on the list!

Best,
VJ

[manisandher]

Relax response: a very primitive description .... when the necessary parameters are met by the human auditory complex, brain will relax. Most will go to sleep ....

Because most are sleep deprived .... and once the brain feels safe, it will proceed to the next important thing on the list!

Ah, so it's not that the music is boring! Thanks.

Mani.

[vrao]

Mani,

Try comparing MP3 and hirez files. It may answer your question 

Best,
VJ

[vrao]

Mani,

Also look into this if not already done so!

http://www.crownaudio.com/en/tools/calculators

"Amplifier Power Required"

http://www.crownaudio.com/how-much-amplifier-power

"Peak headroom

Because music has transient peaks that are 6 to 25 dB above the average level, the power amplifier needs to produce enough power to handle those peaks without distortion.

For example, if you need 100 watts continuous power to achieve the desired average SPL, you need 1,000 watts continuous to handle 10 dB peaks, 3,162 watts to handle 15 dB peaks, and 10,000 watts to handle 20 dB peaks. Clearly, the peaks require far more power than the average levels. In the calculator's Peak Headroom field, enter 6 dB for rock music that is compressed or limited, or enter 20 to 25 dB for uncompressed live music. If you can live with some short-term clipping which may be inaudible, enter 10 to 15 dB.
 

[manisandher]

VJ, a few really dumb questions (and a lot of dodgy maths, I'm sure - a few edits already!):

1. Why would I need such high ratings of continuous power for momentary peaks in the music? Edit: Ha, I should have read the article more carefully. "According to Crown's chief amplifier engineer, Gerald Stanley, amplifier continuous power and amplifier peak power are nearly the same. Typically, peak power is only 1 dB higher than continuous power, and depends on peak duration."

2. My new SET amp will be only 5W pc. So in order to achieve 20dB peaks, the continuous power will be only 0.05W. But, my new speakers are 109dB/W@1m (across the full range). With only 0.05W, I can still achieve an SPL of 96dB @1m (109dB-13dB) - is this correct?

3. Let's consider the Orelo speakers now. We've got something like 60W of power going into the bass unit. So, for the same 20dB peaks, we have only 0.6W of continuous power. I'm assuming that the 'true' sensitivity of the bass unit is around 100dB/W@1m (it's just 'servo-assisted' to match the mid/high unit). So what does the SPL work out to with only 0.6W? 96dB (or so) @1m? In any event, around the same SPL as my Animas with only 0.05W of continuous power going to the bass unit.

4. It seems to me that the sensitivity of the speaker (across the full range, not just the mid and high) is way more important than amplifier power, because 10,000W+ amps just aren't a realistic proposition for home audio use.  In which case, 99.9% of audiophiles are 'stuffed' as they use woefully insensitive speakers. Is this correct?

I'm sure you're busy, so if anyone else would like to chime in, I'd appreciate it.

Mani.

[vrao]

Hi Mani,

I'm no expert with electronics.

In one of the links there is a program to add your desired listing level, distance, speaker sensitivity etc with head room to estimate power requirements. With a high sensitivity speaker I believe you will not have any issues within the normal listening levels. I've seen over 20dB peaks with Phasure DAC.

In which case, 99.9% of audiophiles are 'stuffed' as they use woefully insensitive speakers. Is this correct?

Most likely .... so.   

Best,
VJ

[manisandher]

VJ, I found the 'Amplifier Required Power' calculator thanks. With a 20dB headroom, it looks like I'm good for 86dBSPL at my listening position with my 5W amps and 109dB/W@1m speakers. Good enough for me.

It's all very well talking about the need for high powered amps to cope with headroom, but I think there is another potentially very important consideration: the effect these high-powered (non-class A) amps have on the AC mains. My feeling is that they might 'modulate' the mains with their current drain as the music plays, which may affect the source components adversely. Pure speculation on my part, but it was part of my reasoning to go the high-efficiency fully horn-loaded route, without servo-assisted bass.

Mani.

[manisandher]

Also look into this if not already done so!

http://www.crownaudio.com/en/tools/calculators

This gives a whole new perspective to the 'loudness wars'. It's just as well that much rock/pop music only requires 6-10dB of headroom as the majority of 'systems' out there would start clipping with more. You're sitting at home with your 50W amp thinking all is good. Well, for a 10dB headroom you're going to need a 500W amp! With a typical pair of 87dB/W@1m sensitivity speakers you'd get 84dBSPL in a listening room like mine. With 'better' mastered music requiring a 20dB headroom, you'd get just 74dBSPL at the listening position before clipping.

Really interesting stuff...

Mani.

[acg]

And valve clipping is soooo much nicer to listen to than SS...as a generalisation...that is if you must listen to something clip.

That is the beauty of an active system when using low powered amplifiers, they do not have to drive the passive crossover components.  Mani, you have passive crossovers which also have to be driven by the amplifier, so your 0.05W calculation is likely to be quite underestimated, it is likely to be quite a bit higher.  Then there is the unknown impedance behaviour of the speaker and crossovers which will affect the power that the amplifier is able to deliver.

[PeterSt]

VJ and others,

I am going to read into those links you handed, later. For now I can only contribute by saying that the bass drivers of the Orelo MKII are 101dB (1W/@1m) sensitive natively and that something like 9dB can be added because of the "horn loading". Tbh that was so for the Orelo MKI, the MKII should be a bit better on it.

Btw, generally, the power of an amplifier is determined by its possibility to deliver the current needed over the period of time the music depcits. So might an amplifier be able to deliver 20 Watts continuously then it may well be that it can do 200W as well (depending on the power supply). The question though is : for how long. Is that 0.1 second ? is it 2 second ? or can it handle 200W for 3 seconds before it melts down. Because it is about that : the heat (I'm thinking Solid State - Tubes are a different matter I think).
In the same realm we need to think about the impedance. The lower the impedance (and again it is the music which depicts that), the more current will (need to) flow, the shorter the "peak" is allowed to last. This obviously depends on the speaker too; can it go to 1 Ohm ? does it never drop under 4 Ohm at any frequency ?

Something else is that I learnt that the power music requires can not be calculated really, because it is too complex.

Regards,
Peter

[manisandher]

... something like 9dB can be added because of the "horn loading".

Peter, my understanding was/is that the Orelo wings are not there for horn-loading - they are simply way too small to achieve this at such frequencies. True horn-loaded bass units look more like below, no? The second pic gives an idea of the length required for true horn loading at low frequencies.

Mani.

[manisandher]

Mani, you have passive crossovers which also have to be driven by the amplifier, so your 0.05W calculation is likely to be quite underestimated, it is likely to be quite a bit higher.

Hey Anthony, the sensitivity of my speakers was measured at the output of the speakers, so I'm assuming it included the 1st order crossover components.

But I totally agree that running speakers actively is the best way. I've been there and done that (but not with horns), see attached. 4x Hypex HG700 amps per speaker, so plenty of power. And yes, the results were much better than with the passive crossover. But the gear takes over the room and I'm just not prepared to go there now. That's the beauty of the Orelo speakers -it's all packaged beautifully together.

Mani.

[PeterSt]

Hi Mani,

Peter, my understanding was/is that the Orelo wings are not there for horn-loading

That is why I put that "horn loading" in between quotes. Still it is to some degree because of the compressed and accelerated air the lot as a whole implies. Otherwise it is sufficient to measure the difference (SPL) between the back and the front. The "phase plug" is there for a reason too. The partial covering of the woofers, same.

And ... would you call the bass section of the Anima a horn ? Edit : with loading as such, I mean.
So let's say that loudspeaker designers are always bound to space constraints, and that compromises lure from many angles.

Best regards,
Peter

[acg]

And ... would you call the bass section of the Anima a horn ? Edit : with loading as such, I mean.

Yes, it is a horn.  The backchamber adds compression to the woofer as does a horn throat smaller than the size of the woofer cone.  The height  that the downfiring horn is set above the floor dictates the horn mouth area into the room.  I assume the height from the floor can be adjusted.

[manisandher]

Yeah, I don't know how else you'd get a 109dB/W@1m efficiency from a single 12" driver if it wasn't horn-loaded.

Listening to sine tones at my seating position, I'm getting useful output to about 40Hz, after which it drops fairly quickly. Interestingly, I'm getting very good output at 50Hz with the Animas - there was almost zero output with the Orelos at 50Hz... in the same room and at the same seating position. Weird...

Mani.