Give me a reason NOT to do this

[juanpmar]

Mani,

What about Spectron class D amplifiers to drive your Quad?. I used it with big SoundLab A-1 electrostatics and worked fantastically well.
Now I´m using the Spectron as monoblocks to drive my Von Schweikert VR-9SE and I´m very satisfied with the smooth and detailed sound, terrific bass and huge soundstage. The Spectron works in perfect combination with the NOS1.

You can get the Spectron Musician III MK2 for around $3.500 (at today currency rates that means less than 2.500€ or around 2.125 UK Pounds) and you can buy two and get them as monoblocks.

Here you have some technical information:

http://www.spectronaudio.com/tech1.htm

here as well:

http://www.spectronaudio.com/musicianIII.htm

and here as monoblocks:

http://www.spectronaudio.com/monoblocks.htm

I hope it helps, Juan

[Suteetat]

Hmmm... I have a slightly different take on Soulution. I heard 700 monoblocks and 710 stereo amp a few times with Magico Q5 and Q3 in that order.
First of all, I am not sure why the two are often shown together as I don't like the combination at all. My theory is that damping factor on Soulution is credibly high (> 10,000 per spec sheet). With Magico, I don't think I ever heard a drum that start and stop as quickly in real life. Listening to it for awhile, I felt that the sound was rather unnatural. I heard Quads a few times but not well enough and don't know if it would benefit from high damping factor or not. Magico Q3 sounds more natural to me with Lamm M2.2 (damping factor 82 if I remember correctly). On the other hand Lamm M2.2 could not quite control dual 12 inch Eton woofers in Usher Be-20 quite as well as would most likely benefit more from amp with more damping factor.

Soundwise, this really depends on the person. I find Soulution hyperdetail but in a good way, not very fatiguing but I really like my music a bit more sugar coated. Vocal music on Soulution just did not quite do it for me unlike Lamm M2.2. So I hope you will have a chance to take a good listen to Soulution with Quads yourself before making up your mind. It is definitely an excellent amp but not with a sound that is universally good for everybody.

[Flecko]

Up to some degree, sure yes. I mean, in order to suffice a certain bandwidth, first the slew rate has to be sufficiently high to not let mix up the subsequent "samples" (but notice this is about analogue). When this is met, next the slew rate can still be very much higher, while the bandwidth does NOT increase

So you can calculate the minimum slew rate from bandwidth but not the maximum. The calculation should be:
Maximum Frequency fmax=1.5 Mhz
Maximum output voltage Vmax=30V
Minimum Slew Rate [V/us]= fmax*Vmax=1.5MHz*30V=1.5*10^6[1/s]*30V=1.5[1/us]*30V= 45 V/us
It has to be considered, that the amplifier has to change the voltage from peak to peak in half of a period. So the result must be multiplied by 2.
Makes a slew rate of 90V/us
If the maximum voltage is given in RMS and not PP, which is the normal case, you have to multiply by 2*SQRT[2].
The minimum slew rate would be then 254,5 V/us.

[PeterSt]

Well, at least I made a mistake here :

So, for e.g. 30V this takes 6V/uV. Is the needed bandwidth 1GHz ?

This, of course, should have said 1MHz (because I wanted to calculate that - and did).

But

So you can calculate the minimum slew rate from bandwidth but not the maximum.

I don't see any of the subjects I brought up in your math ...

It has to be considered, that the amplifier has to change the voltage from peak to peak in half of a period.

Nope. This is not difficult for an amplification because this is a "swing". Maybe not exactly true for a square wave form (what this will be about), but if the length of the high voltage sustainment is short enough (the highest frequency = the highest bandwidth) it comes close to that. But since you leave this out of the equation anyway, it can go in all directions ...
It is to be from zero to peak and back. If this weren't so, what would be "fall time" ? it would be the same as rise, but in the other (voltage +/-) direction.

Peter

[Flecko]

So you can calculate the minimum slew rate from bandwidth but not the maximum.

I don't see any of the subjects I brought up in your math ...

I used the unit V/us becuase it is more intuitive. I derived this based on the definition of the slew rate, which is the rise time from voltage "a" to voltage "b" in a given time. If you start with this definition, the maximum frquency and the maximum voltage is directly linked to the minimum slew rate.

Quote
It has to be considered, that the amplifier has to change the voltage from peak to peak in half of a period.

Nope. This is not difficult for an amplification because this is a "swing". Maybe not exactly true for a square wave form (what this will be about), but if the length of the high voltage sustainment is short enough (the highest frequency = the highest bandwidth) it comes close to that. But since you leave this out of the equation anyway, it can go in all directions ...

It has nothing to do with the amp or how the total signal could look like, it is just math. If you have a sine, and that is what we always have no matter how the total signal looks like (because every signal is just a linear combination of sines), and this sine is at the maximum frequency your amplifier can handle (in this case 1.5MHz), the steepness of the curve, in other words the slew rate, you need to reproduce this sine is dependent on the "voltage change" per "time" unit. If you use one period as the time the voltage should change from Vo=0V to Vmax=30V, you use a number that is only half as big as what is actually needed.
I have made a scetch. You can see that to reach the first maximum of the sine you have 1/4 of one period. This maximum is at a voltage off pp/2. Hence your slew rate is PP-Voltage/2 times 4*Maximum Frequency or simplified PP-Voltage times 2*Maximum Frequency.

[PeterSt]

Hi Adrian - You make a nice difficult outlay of this !

I think if you go this direction, it is waaay too simple. Not that I would be able to make the story complete (on the real math I mean), but if you talk like this you'd have to give definitions for overshoot (limits), allowed ripple and what not, and I am not sure it works like this for the functional aspects we are trying to talk about.
If you, for example, want to talk about the as good as possible representation of 20KHz (assumed that is audible or at least senseable), you'd have to start with what comes out of the D/A converter. You'd have to start with no filter at all (and the amp representing that sh*t because it is good (on the slew rate)) vs. common filtering as needed and applied (I'm not talking NOS1 now), that not demanding anything from the amp (because it's stupid 20KHz sines always without real "impulse" data).

What we have been doing/trying, is approaching it the other way around : the amp has a bandwidth, and what does that mean without context. I think this is useless, because a figure comes from that which won't say whether it is good or bad for our purpose.

Theoretically one could start off with 30KHz as a "senseable" frequency because it easily can be there in properly done HiRes. But use 50KHz to be on the safe side, although I still must see a properly done HiRes which can represent that. This is the easy one. The more difficult one is 50KHz of impulses (On/Off sound) because that theoretically can exist too. Synthesized at least. But, stupid, plus I don't believe much in the "senseable" thing. So, maybe better stick to 18KHz as a realistic On/Off which can be audible for some. *Now* do that math ...

Don't forget in your math to incorporate the distance between the pulses, because one sample space in between them will physically be impossible and it will become a mess (it would need infinite rise and fall time). But, the better the "slew rate", the less soon it wil become a mess, or IOW the more close the pulses can be without disturbing eachother.
Once you are done with it, you will find to have done it without reason, because your DAC will never spit out those pulses, because the filter will have been ahead of it all and will have flatten it (to normal sines). Not so with the NOS1, but now there's the problem of images at higher frequencies. Aha. How high ? at what level ? and they sure are at higher frequencies ...
The higher the bandwidth, the more is let through. The lower the bandwidth, the more is nicely filtered out, that is, while in some frequency band (like 200KHz) the amp will distort. At what level ? dangerous ?

There is all a bit more to it.
Regards,
Peter

[Flecko]

I think if you go this direction, it is waaay too simple. Not that I would be able to make the story complete (on the real math I mean), but if you talk like this you'd have to give definitions for overshoot (limits), allowed ripple and what not, and I am not sure it works like this for the functional aspects we are trying to talk about.

It should be that simple because it is just an estimation of the minimum slew rate. The real slew rate must be meassured. But I think it is good to calculate a lower limit of what is to be expected from the amp. The real slew rate should be always higher than the calculated number. So we are on the safe side.

If you, for example, want to talk about the as good as possible representation of 20KHz (assumed that is audible or at least senseable), you'd have to start with what comes out of the D/A converter. You'd have to start with no filter at all (and the amp representing that sh*t because it is good (on the slew rate)) vs. common filtering as needed and applied (I'm not talking NOS1 now), that not demanding anything from the amp (because it's stupid 20KHz sines always without real "impulse" data).

If the amp should reproduce a perfect sine wave at a certain frequency, the slew rate must be higher than estimated, because the sine has a maximum slope at the point of zero crossing, that is slightly higher than what I have assumed.
I think it gets too complicated if you incorporate other equipment and it is not necessary. There are simple boundary conditions. The maximum Voltage and the maximum frequency. If you want to fullfill this boundary condtition, the amp must have a slew rate, that is high enough to reach the voltage in the expected time. The amp could have a higher slew rate but not a lower, otherwise the datasheet would say something wrong.

Theoretically one could start off with 30KHz as a "senseable" frequency because it easily can be there in properly done HiRes. But use 50KHz to be on the safe side, although I still must see a properly done HiRes which can represent that. This is the easy one. The more difficult one is 50KHz of impulses (On/Off sound) because that theoretically can exist too. Synthesized at least. But, stupid, plus I don't believe much in the "senseable" thing. So, maybe better stick to 18KHz as a realistic On/Off which can be audible for some. *Now* do that math ...

I am not sure what you mean by the on/off thing. You mean you have an outputvoltage of 0V and than suddenly you switch the siganl (music) "on" and imedeatly "off" again?
If this would be so, the shape of this pulse would be determined by the bandwidth you have. For very short pulses you would need very high frequencies. If your amp can not represent this frequencies, the sharp pulse would get flatter and wider. It depends of the frequency content of the puls, if anything will come through. The basic principle is, that no matter if we have digital data or analog data, the slope of any puls should be stricly limited by the bandwidth of the equipment. This also should be true for the "resolution" the amp has, to make it possible to differentiate between two pulses, which are seperated by a very small time. If the pulses are so close to each other that it crosses the maximum frequency your amp has, or what ever hifi equipment you want to look at, it will not be reproduced as two peaks. It will be smeared and maybe reproduced as one peak or it will not be reproduced at all if the bandwidth is too small.
There is this philosophie of having a high bandwidth, which is much higher than what you can hear. Reading the article I mentioned earlier, I would say this high bandwidth is a byproduct of getting the amplifier fast to get the negative feedback work well. Having higher bandwidth also should prevent phase changes in the "used" frequency range.
For the math so far, nothing should change, because we only have used basic principles that hold true under any situation.

[christoffe01]

But even used, the 710 will cost me 14K UK pounds! Now, I'm not a rich man and this is a totally absurd amount of money for me. But, I can just about afford to buy it without putting my family in a state of total poverty.

Someone... anyone... please give me a compelling arguement as to why I shouldn't do this. Please...

Mani.

Hi Mani,

in the German HiFI Magazin Stereo they tested the 710 in the beginning of 2010, and a Acoustic Arts AMP II MK2 a couple of month later

(http://www.accusticarts.de/pages/de_amp_ii.html)

The 710 was by a very, very small margin superior but the Acoustic Arts AMP II MK2 costs new appr. € 9.000,00 only.

best

Joachim

[manisandher]

What about Spectron class D amplifiers to drive your Quad?.

Hi Juan, thanks for the suggestion. Well, they certainly look interesting... and they're capable of driving a 0.1Ω load!

But if going the class-D route, I kind of feel a slight loyalty towards Hypex. I already have 8 of their UCD700HG modules that I used to use to quad-amp my previous speakers. And the yet-to-be-released Hypex 'Ncore' modules look very interesting (http://www.hypex.nl/docs/papers/ncore%20wp.pdf).

But maybe I need to re-evaluate this loyalty...

Mani.

[manisandher]

... I am not sure why the [Magico/Soulution] are often shown together as I don't like the combination at all.

I've read similar comments from other people. I haven't heard that particular combo but did hear the Magicos being driven by a pair of Spectral amps at the Munich High End last year. And I can't say that I was particularly impressed by that combo either... but then again, I wasn't impressed by anything there!

I find Soulution hyperdetail but in a good way, not very fatiguing but I really like my music a bit more sugar coated. Vocal music on Soulution just did not quite do it for me unlike Lamm M2.2. So I hope you will have a chance to take a good listen to Soulution with Quads yourself before making up your mind.

I think I know what you mean. When I listen to a real instrument, I want to be able to 'hear' the size of the instrument being played. Too often, instruments like violins sound as if the strings have been placed on 2D MDF, rather than a gorgeous 3D waxed body.

BUT... ultimately, I'd rather have a 'thin' sound with 'accurate' timing and touch than a 'blossomed-out' sound with just beautiful harmonics - I mean, I'd rather listen to Keith Jarrett playing a cheap stand-up than me playing a Steinway!

Mani.

[juanpmar]

But maybe I need to re-evaluate this loyalty...

Just in case here you have a nice review of the Spectron:
http://www.positive-feedback.com/Issue44/spectron.htm

Juan

[christoffe01]

Hmmm... I have a slightly different take on Soulution. I heard 700 monoblocks and 710 stereo amp a few times with Magico Q5 and Q3 in that order.
First of all, I am not sure why the two are often shown together as I don't like the combination at all.

Hi,

The combination 710 stereo amp with Magico Q5 or Q3 (metal housings - no resonances from the casing) is a very fast, super detailed system with an unheard fine resolution. You will not hear this fine resolution with any other speaker (except YG Acoustics).

If you like a fast, clean, detailed, analytical sound this is the combination. (preference Jazz music)
The most musical power amp is the Dartzeel NHB-108, but with 100W at 8 Ohms only. (favoured by many active musicians)

If you like a warm sound, tubes is ...............  !

Peter has an amazing, warm sounding system with a floor shaking bass.

best

Joachim

[PeterSt]

I like to add to this that Joachim has more experience than listening to Magico's on some show. So, FWIW in here, I think we can very well trust his help.

Peter


PS: Joachim, it seems offtopic (but it is not) ... what about your relocating ?

[juanpmar]

with Magico Q5 or Q3...You will not hear this fine resolution with any other speaker (except YG Acoustics).

And except the Von Schweikert VR-9SE  :

http://www.stereotimes.com/showreport05cespage6.shtml

[manisandher]

The most musical power amp is the Dartzeel NHB-108, but with 100W at 8 Ohms only. (favoured by many active musicians)

Hi Joachim, yes the Dartzeel still holds a lot of appeal for me. But it performed pretty badly on John Atkinson's bench, so I'm wondering whether it simply has a 'nice' distortion characteristic that makes it sound euphonic. If so, this isn't really the route I want to take... for now.

What I really want is to be able to hear the NOS1 with as little added by the amp and speakers as possible.

Mani.