If we follow this thread with atypical unrelenting distracting forum answers, it proves the same point: https://www.stereonet.com/forums/to...er-amp-do-i-want-more-volume/page/2/#comments Note only one forum member asks the right question which is "What is the input sensitivity of your power amp? " The owner had an equivalent input sensitivity to the Quad 306 with a Cambridge amp with 370mv, but sold it, and there the problems started because the amplifier he replaced it with,teased with more power but of course that power was never achieved because (the unstated but consumer found ) sensitivity was higher .
What is pertinent is he has a DAC device rated just over 0dBfs, but of course 0dBFs is never present, only consumer line level. You can work out how much money was wasted in moving away from an amplifier that had the correct input sensitivity
What is pertinent is he has a DAC device rated just over 0dBfs, but of course 0dBFs is never present, only consumer line level. You can work out how much money was wasted in moving away from an amplifier that had the correct input sensitivity
It's better to have an amp with a higher sensitivity and if necessary simply attenuate the output of the dac.
Even this is not needed as his 'passive' preamp also has switchable attenuators.
I suspect the reason for the OPs question boils down to his passive preamp volume control only being useable over the first 25% or so of it's travel due to the combination of the 306's high sensitivity and the dacs higher out.
Even this is not needed as his 'passive' preamp also has switchable attenuators.
I suspect the reason for the OPs question boils down to his passive preamp volume control only being useable over the first 25% or so of it's travel due to the combination of the 306's high sensitivity and the dacs higher out.
Both are correct. The digital stream has a certain level, not in the waveform amplitude but in the digital code. Has to do with amongs others how loud it has been recorded. Same as the level is 'encoded' in the groove amplitudes on an LP record.
But a 1V DAC output setting will definitely sound less loud than a 3V DAC output setting. A factor 3 less loud of course, about 8dB IIRC.
Jan
Thanks Jan.
It's reassuring to know that I'm not just imagining the volume differential between 1V, 2V and 3V output options, from my DAC. 😊
My original question from Post #34...
Whilst I was always aware of several ways in which I can reduce the overall signal in order to make full and best use of our stepped attenuator, I don't think my question is unreasonable.
It is my thinking that other QUAD 306 owners - who don't share my passive preamp and output adjustable DAC - might also find the answer to this question useful.
Whilst I was always aware of several ways in which I can reduce the overall signal in order to make full and best use of our stepped attenuator, I don't think my question is unreasonable.
It is my thinking that other QUAD 306 owners - who don't share my passive preamp and output adjustable DAC - might also find the answer to this question useful.
What is not said, is Only if playing a warned 0DbFS test track to occupy otherwise unused headroom, As 0.894V Peak to Peak or 0.316V RMS is available on a CD, It is why Quad a reputable manufacturer since 1936 provided 0.375V RMS sensitivity with the 306 , because 0.316V RMS is consumer line level, and audio is arguably far more enjoyable, not adding unnecessarily distortion with voltage amplification prior to a power amp.
No, from over 13 years of experience with the 306, and it never overloading playing any CD, all you need do is connect your CD player to your passive attenuator input, and Quad 306 from it's output and press play. Your CD outputs a maximum of 0.316v RMS as does your DAC because the maximum level is embedded in the media, you can press as many selections of your DAC as you like but it still outputs 0.316V RMS maximum ,... it may, as I alluded to earlier arrange slightly different current capability but that's not voltage ... noting though to never play any Test Disc..... , it couldn't be simpler. There is no need to attenuate because you already generally have the correct matching.
The matching can be improved further by having extremely high series resistance at lowest volume , and adjustable shunt resistance to achieve no adverse loading to the source component, but that awaits a change of attenuator.
Oh and be very thankful you chose the 306, (and your passive to make the most of your source component capability ) which demonstrates excellent compliance to known audio levels, which is what is expected from such a trusted manufacturer.
Last edited:
Chris, if that is the case, then why do we see output levels (far) above 1Vrms these days? Reputable DACs often can be jumper set to 1V, 2V and even 4V output levels. I know where the 0.316V comes from, but that has been left in the dust a long time ago.
The high sensivity of these components can cause awkward volume control ranges. The problem is real with those components.
Jan
Hi Jan
We see this in advertising, but it ends up we lose relationship to actual levels originating with the media, and begin believing that equipment creates levels
when its the media via the equipment that has provided consumer line level. Awkwardly there is a trend for amplifiers not to state their sensitivity and results in consumer confusion, with on one hand apparent large power output, but actually much less output because the equipment sensitivity is never reached, see #61
The 1v 2v and 4V levels are allowing headroom, that is ordinarily never present, and not as far as I know the DAC rearranging consumer line level to jump up in voltage. The only exception is if you happen to need to experience a Test CD providing 0DbFS which comes with adequate precautions, You would certainly select your DAC to the higher level, such as if measuring using a scope, the level 0dbFS can provide.
Why 0.316V RMS is used as the output of the media, is to preserve dynamic headroom for enjoyable music replay with digital source equipment.
This results from digital not being forgiving, and when you run out of headroom hard clipping can occur. We can find reference to that level with the Audio Engineering Society https://www.aes.org/par/l/ and is typically seen expressed as -10dBV and annotated as consumer line level where RCA connections are provided. "-10 dBV Standard voltage reference level for consumer and some pro audio use (e.g. TASCAM), equal to 0.316 Vrms. (Tip: RCA connectors are a good indicator of units operating at -10 dBV levels.)
We also find it with Streaming services , although not expressed as consumer line level, instead as Loudness Units Full Scale , where there are many selections available but most stick to minus -14 LUFS, noting the end user can have choice too of the level. There is some agreement that -14 LUFS is close if not the same as -10dbV
Loudness Standards for Music Streaming Platforms
Spotify: -14 LUFS (integrated)
Apple Music: -16 LUFS (integrated)
YouTube: -14 LUFS (integrated)
Tidal: -14 LUFS (integrated)
Amazon Music: -14 LUFS (integrated)
Deezer: -15 LUFS (integrated)
Of course had we adopted Murray Crosby's ideas and found as companding by Dolby and DBX we might have instead 2: 1 compression and 1:2 expansion, recordings, where the improvement would be in noise floor , particularly with 16 bit equipment. There would have been less need for preserving dynamic headroom with level , rather we would have all been entertained by what was possible.
We see this in advertising, but it ends up we lose relationship to actual levels originating with the media, and begin believing that equipment creates levels
when its the media via the equipment that has provided consumer line level. Awkwardly there is a trend for amplifiers not to state their sensitivity and results in consumer confusion, with on one hand apparent large power output, but actually much less output because the equipment sensitivity is never reached, see #61
The 1v 2v and 4V levels are allowing headroom, that is ordinarily never present, and not as far as I know the DAC rearranging consumer line level to jump up in voltage. The only exception is if you happen to need to experience a Test CD providing 0DbFS which comes with adequate precautions, You would certainly select your DAC to the higher level, such as if measuring using a scope, the level 0dbFS can provide.
Why 0.316V RMS is used as the output of the media, is to preserve dynamic headroom for enjoyable music replay with digital source equipment.
This results from digital not being forgiving, and when you run out of headroom hard clipping can occur. We can find reference to that level with the Audio Engineering Society https://www.aes.org/par/l/ and is typically seen expressed as -10dBV and annotated as consumer line level where RCA connections are provided. "-10 dBV Standard voltage reference level for consumer and some pro audio use (e.g. TASCAM), equal to 0.316 Vrms. (Tip: RCA connectors are a good indicator of units operating at -10 dBV levels.)
We also find it with Streaming services , although not expressed as consumer line level, instead as Loudness Units Full Scale , where there are many selections available but most stick to minus -14 LUFS, noting the end user can have choice too of the level. There is some agreement that -14 LUFS is close if not the same as -10dbV
Loudness Standards for Music Streaming Platforms
Spotify: -14 LUFS (integrated)
Apple Music: -16 LUFS (integrated)
YouTube: -14 LUFS (integrated)
Tidal: -14 LUFS (integrated)
Amazon Music: -14 LUFS (integrated)
Deezer: -15 LUFS (integrated)
Of course had we adopted Murray Crosby's ideas and found as companding by Dolby and DBX we might have instead 2: 1 compression and 1:2 expansion, recordings, where the improvement would be in noise floor , particularly with 16 bit equipment. There would have been less need for preserving dynamic headroom with level , rather we would have all been entertained by what was possible.
My assumption when designing a multi input source preamp is
Legacy inputs like old tuners, tape decks etc: 200mV
Modern digital sources like DACs and CD players: 1V
So the preamp line stage is set for 14 dB gain, and the modern 1V input sources are attenuated at the input connector by -14 dB. This gives 1V for all sources at the preamp output.
There is a spread in the source voltage figures I quoted above: some modern digital sources quote 1.6 to 2 V output (quite a few CD players over the years) while seen I’ve legacy sources quoted at 120mV (old tuners for example) all the way up to 350mV.
For phono preamps, I design for 35-40 dB for 5 mV input and that feeds into the 14 dB line stage.
With the gain structure mentioned, the volume control will easily cater for a 6 dB spread in input levels without requiring it to be too far off the nominal setting which on a log taper pot will usually be between say 10 o’clock and 2 o’clock on the volume dial. Amplifier rated full output for rated input levels is with the volume control at maximum.
I find it a bit strange that modern digital gear is coming out with 3 and 4 volt output levels.
Ultimately, the volume control on your preamp sets the final preamp output level and no source signal should be clipping at the preamp output level.
On legacy gear like the Quad being discussed here, c. 350mV input sensitivity was the norm, so the aforementioned gain structure would not have the +14 dB line amp - the gain would just be 0 dB.
In this specific case, I would not fiddle with the power amp gain - you could run into stability issues if you try to reduce it by a large amount which is clearly what would be required. I would suggest a circa 10:1 attenuator using a 10k and a 1k divider, the output of which feeds into the volume control. If the volume control is 10 k of better, the loading on the divider will be minimal, but you could always tweak it by reducing the upper 10k divider resistor to say 8.2k or 7.5k.
From what I’ve heard, digital attenuation always reduces resolution - resistive dividers do not.
YMMV
Legacy inputs like old tuners, tape decks etc: 200mV
Modern digital sources like DACs and CD players: 1V
So the preamp line stage is set for 14 dB gain, and the modern 1V input sources are attenuated at the input connector by -14 dB. This gives 1V for all sources at the preamp output.
There is a spread in the source voltage figures I quoted above: some modern digital sources quote 1.6 to 2 V output (quite a few CD players over the years) while seen I’ve legacy sources quoted at 120mV (old tuners for example) all the way up to 350mV.
For phono preamps, I design for 35-40 dB for 5 mV input and that feeds into the 14 dB line stage.
With the gain structure mentioned, the volume control will easily cater for a 6 dB spread in input levels without requiring it to be too far off the nominal setting which on a log taper pot will usually be between say 10 o’clock and 2 o’clock on the volume dial. Amplifier rated full output for rated input levels is with the volume control at maximum.
I find it a bit strange that modern digital gear is coming out with 3 and 4 volt output levels.
Ultimately, the volume control on your preamp sets the final preamp output level and no source signal should be clipping at the preamp output level.
On legacy gear like the Quad being discussed here, c. 350mV input sensitivity was the norm, so the aforementioned gain structure would not have the +14 dB line amp - the gain would just be 0 dB.
In this specific case, I would not fiddle with the power amp gain - you could run into stability issues if you try to reduce it by a large amount which is clearly what would be required. I would suggest a circa 10:1 attenuator using a 10k and a 1k divider, the output of which feeds into the volume control. If the volume control is 10 k of better, the loading on the divider will be minimal, but you could always tweak it by reducing the upper 10k divider resistor to say 8.2k or 7.5k.
From what I’ve heard, digital attenuation always reduces resolution - resistive dividers do not.
YMMV
Last edited:
You can have 32 bit or more digital attenuators that decrease resolution less than the S/N reduction with resistive attenuators.
I don't think you can make a general statement that one is better than the other, it all depends.
Jan
I don't think you can make a general statement that one is better than the other, it all depends.
Jan
With resistive divider you lose some SNR as the preamp will have the same gain for low input signal as for high.
I use non GNFB current conveyor in my preamps with two set of gain, high and low with the same SNR.
I am not sure I explained it clearly.🙁
Something like this. Gain P1/Rg1 or lower with Rg1 parallel with Rg2.
Attachments
Last edited:
Fair enough Jan, on 32 bit stuff I agree but a lot of it out there isn’t that good. The noise floor on a divider feeding a 10 k pot is still very impressive. Best of all, it’s pennies to implement 😊
Are you referring to 1V as peak to peak where the standard provided is 0.894Vpp or as RMS ? https://en.wikipedia.org/wiki/Line_level
The present state of affairs is sadly one of not providing the products input sensitivity in specifications, not showing leadership with what are well known line level standards, and having distortion introduced after source components, prior to the power amp. We see motivation to introduce confusion to the consumer, by stating levels that are unrelated to standards, as only headroom. Disturbingly where stated power levels of output are provided, are then subsequently found substantially less by the purchaser, because input sensitivity ( why it is hidden in data) bears no relationship to established standards. see #61
The Quad 306 remains a beacon of correctness. I would strongly advocate all power amps now comply with sensitivity essentially the same as the AES consumer line level specification. Quad as example provided since 1967 to 1993 sensitivity of 500mv, but bless them improved that further with the 306 to 375mv.The preamp if preferred over a passive can instead be a device not increasing voltage but delivering current, so the source impedance never has adverse loading. Noting #68 where this can also be done more easily with redesign of a passive attenuator.
Abiding by consumer line level standards for all power amplifiers as the Quad 306 does, makes perfect sense. .
Well, @Chris Daly and @batteryman were right about me not needing to worry about the very high sensitivity of our new QUAD 306. With our passive preamp, we still get more than 50% usage of the volume knob, which is plenty for our purposes.
I was really delighted with the sound of this charming little amp upon first listens - but I did have a few reservations. There was a pervasive "dryness" to the midrange, which detracted from the beauty of vocals and the like. As tube-amp lovers, this was a serious problem for us. Also, there seemed to be a slight lack of fullness to the sound - not that the upper bass and lower mids were MIA, it just seemed a little lightweight and anaemic. For us, it seemed like a big part of the music enjoyment was being held-back within this area.
I left the amp for a few more days, whilst we got seriously busy with Chinese Lunar New Year - a very big deal here in China - powered-up 24/7 and softly playing music in background mode during the rare moments we were at home.
I have since returned to the amp for a few weeks of long and critical listening sessions - and am really quite stunned at what has resulted!
The midrange dryness had disappeared utterly and completely and there is now a pleasing fullness to the lower mids and upper bass also. There is also a strong new sense of dynamic contrast, which was not at all obvious in earlier listens. The other pleasing aspects of the QUAD's sonic signature, have all remained and or improved.
In a final appraisal then, here is how I would describe the QUAD 306: (My personal and highly subjective view... YMMV!)
- The QUAD is utterly silent partnered with our very efficient loudspeakers; just how we like it. When music is not in play (apart from the little green LED hidden deep within our hi-fi rack) there is no way of knowing the amp is sitting there, fully powered-up and ready to go - even if the volume on our preamp is left at full listening position.
- The QUAD is very "open" sounding and very dynamic. We have found that it is very hard to find "open" sounding amps - either tubed or solid-state - and the QUAD is as good as the best in this regard. Vocals and other central music elements just seem to leap out of the speakers and I find myself wanting to listen a little louder than usual. Superb!
- The QUAD is detailed and revealing without a hint of brightness or listening fatigue. Spatial cues are well rendered and rim-shots, percussive elements and the like are highly propulsive. I'm sure that the excellent tops of the QUAD underpin many of the other wonderful sonic attributes of this amp. The balance is just perfect.
- The QUAD portrays correct richness of timber sitting exactly midway between euphonic and incisive. Instruments sound the way they should, without any sweetening. Vocals, solo piano, vibes and brass all tell the real story here - delivered with magical credibility accompanied by rich harmonics, but not without the all-important leading-edge bite.
- Imaging is wonderful and as good as the best. If the recording offers this illusion - not all do! - the QUAD paints the picture superbly with clean lines and solid images.
- The soundstage width the QUAD offers is clean and wide - and well beyond the physical limits of the loudspeakers and the front wall. Best of all, the extreme "open-ness" of the QUAD has the music inhabiting the free-space around our loudspeakers, rather than remaining trapped between them. This bodes well for "other room listening" also! [He says, typing from the office and thoroughly enjoying the music...]
- PRAT is as good as I have heard from any amp - and that includes some pretty impressive amps we have owned including a Nelson Pass-built FIRST WATT J2, LEBEN CS300F, CARY CAD SE300 Signature monoblocks and others. The PRAT is so impressive that time and time again, I've felt the need to jump-up to check our LP12 playback speed with the strobe to see if the platter is spinning too fast. Nope. The platter speed - with stylus in-groove - remains solid and right on the mark… The QUAD just sounds so pacey!
- Tops are extended, pure and clear with well-judged sparkle. The tops lack the going-on forever quality of the best tube-amps, but surprisingly there is no lack of "air" up top at all, with the QUAD. It is quick to reveal poor recordings and should not be regarded as an amp that makes everything sound like honey. The QUAD is one truthful little engine.
- The QUAD mids manage quite a balancing act, because whilst there is a hint of warmth and humanity in vocals, cellos, acoustic guitar bodies and the like, there is also an abundance of texture, crunch and edge. This type of midrange detail is often MIA in amps that don't shout in the mids. Not the QUAD.
- Bass is not of the cyborg variety, but it is full, quick and weighty - and has an impressive degree of slam. Importantly - in our system - there is no sense of ponder, hang-over or bloat, which are attributes we cannot tolerate. I’m guessing that the excellent and very deep QUAD bass - just right IMHO - is probably the reason PRAT is so very impressive with this machine.
Having inspected the QUAD internally upon arrival from the highly reputable Dutch dealer, I noticed that the main smoothing capacitors are all very new - 4700uF / 50V / not bypassed - but I had assumed that they had all been given a good run-in already. Perhaps I was wrong?
The only explanation I can imagine for this gradual and very favourable sonic development - since Day-1 and thereafter - is that perhaps the new caps were not yet fully formed, when the amp first arrived. This wee amp now sounds so darned good, we're in no hurry at all to return to our reference tube amps.
What was supposed to be a stand-in amp for the unbearably hot summer months here in China, could in fact serve as a full-time first choice. That is a very big statement coming from a house-full of tube-amp lovers - with some stunning tube amps standing by.
The QUAD 306, for our purposes, is a total winner - and a keeper. I'm a total fan!
The last few days, we have been down to -5 degrees Celsius outside - and our tube amps haven't been ignited, even once!
❄️😊❄️
Last edited:
I believe that the reason is so that it can be used in biamping with the 606, which is 500mV for 140W out. The same voltage level in gives the same watts out in bifh.
I have owned a 405-ii, 306, and currently a 909.
I was charmed by the 405 which I bought to drive my Martin Logan Hybrid electrostatics - a very difficult load.
The 909 was purchased recently to drive my Magneplanar MG6i planar speakers, also a difficult load and needing an amp with 4r capability.
I am somewhat surprised that these amps are dismissed by many, most of whom will never have audiotioned one.
Good to read that you like what you hear and thanks for your review.
With regards to consumer audio: you are right that the normal level is lower than 2V rms but 2V rms can occur (rarely but still) so anyone with normal thinking designs for 2Vrms. It is a very old unofficial but nevertheless adopted standard. I recall some early CD players with +/- 5V PSUs for the IV/output stages clipping with some media. The headroom is there to never have any issues not even with test CDs. Sane thinking! I am sure many would freak out when the test device on their bench chokes on playing a test CD 🙂
The even older standard was -10 dBV (0.316 Vrms). With the invention of CD it became 2V rms via RCA and practically every digital device, every DAC chip datasheet etc. of today mention this. Older guys here will remember the large difference between for instance FM tuners and cassette players with the then new CD players. Amplifiers started to have special CD inputs that simply had an attenuation network. Interested people that noticed such issues were quickly aware of the higher output level. Switching inputs could become a shocking moment with large volume differences.
You keep creating confusion with this. Also with the writing errors. It is V, Vrms, dB, dBV etc.
Last edited:
The original industry standard output level agreed upon for the introduction of CD was not 2V. It was actually 1.4V (3dBV) and was changed at the last minute as a few manufacturers (Sony etc) had decided to make their players "louder" on the sales floor to get an advantage. Yes! Loudness wars back in 1982/3 on the players themselves!
Plenty of 1st gen players had already been shipped and service bulletins were quickly dispatched so products already in transit and landed in the US and other countries could be modified before going to the dealers. It was a case of "they are doing it- we should too"
Here's a Hitachi example for their first player, the DA-1000, notice serial number 531 onwards. March 1st 1983- the official release date for overseas markets. Sony and others had already been selling their ~2.2V Sony CDP-101 player for 6 months in Japan.
A few different value resistors in the output IC feedback loops was all it took to make the magic 2V we all take for granted as being some special "standard". It wasn't.
As for the Quad's sensitivity, they were always ridiculously high, be it from the old days with their horrible 4 pin DIN connector levels right through to their eventual implementation of RCAs with a ~300mV sensitivity on power amps. The entire world had moved on to 150-200mV line levels for full rated power on integrateds and 1.0V-1.5V for full rated power on power amps.
Plenty of 1st gen players had already been shipped and service bulletins were quickly dispatched so products already in transit and landed in the US and other countries could be modified before going to the dealers. It was a case of "they are doing it- we should too"
Here's a Hitachi example for their first player, the DA-1000, notice serial number 531 onwards. March 1st 1983- the official release date for overseas markets. Sony and others had already been selling their ~2.2V Sony CDP-101 player for 6 months in Japan.
A few different value resistors in the output IC feedback loops was all it took to make the magic 2V we all take for granted as being some special "standard". It wasn't.
As for the Quad's sensitivity, they were always ridiculously high, be it from the old days with their horrible 4 pin DIN connector levels right through to their eventual implementation of RCAs with a ~300mV sensitivity on power amps. The entire world had moved on to 150-200mV line levels for full rated power on integrateds and 1.0V-1.5V for full rated power on power amps.
Last edited: