Do you need a Preamp?

Supratek Chardonnay vs Mytek

This is not really a review but my observation of a tube and solid state preamplifier. I have been using Supratek Chardonnay valve preamp for about five years. I fell in love with tube preamplifiers for the liquid and silky mid range after hearing the famous Blue circle preamp and decided to get one. After some research I found one of the longest surviving thread in Audiogon (since 2002) for Preamp Deal of the Century where various users compared Supratek with many high end preamplifiers.
 
This preamp was and still is Edger Kramer's reference preamp. The strength is the nice push in the low mids and sparkling highs. An audiophile’s recipe for good sound.
 
Sadly, like all tube/valve equipment it too got its own “cons”.  The one I hated most was tube rolling. The sound changes with different tubes and it is hard to make up my mind which one supposed to be the best. The magic of the microphonics nature of tubes which makes distortion pleasant and likeable did not meet my Ambiophonics requirement for a simple clean signal path. It was a fight between accuracy and musicality.
 
So came in the Mytek DSD 192 DAC as an substitute. With  4 Y-splitters, I managed to convert the single output to three. Setting the volume control to analogue and connecting directly to Amp gave the cleanest possible signal path.
 
With new discovery, I listened to all my familiar tracks but…. Mytek failed to give the magical mids and the scintillating highs. Bass seemed lacked but it did go lower although presence wasn't felt as the lack of boast in the low mids that valves do. The only consolation is that the preamp is accurate and without any flavor. Perhaps, too sterile. Bordering too dull. I guess, accuracy is not everything.
 
Anyway, having made up my mind to go fully digital I settled for Mytek now because of I need the accuracy. Mytek turned out to be a reliable Preamp and bring out all the details. You may not like it at first but eventually you get used to it and then when you move to tubes your ears need re adjustment.
 

My dilemma now is what to do with the Supratek.  Mick Maloney now makes the US$6000 or US$10000 version only and that to only if he feels like it. Should I keep it or sell it? Or forget about accuracy and get valves flavor? Ahh…it is not easy to be an audiophile. So many decisions to make….  

by the way STC, I saw you changed your speaker too. What speaker you are using now?

Bro, already long time..

SoundLAb

wow, that is very nice speaker bro.

77006688 wrote:wow, that is very nice speaker bro.

Thanks Bro..Oklah...can do.

Continuing with other tracks, Mytek seemed to be lacking the highs. It just got not enough juice to drive three line out and the Miniambio causing serious voltage limiting where the output is limited to less than 1V.

Listening to Time from DSOTM confirms that the ESL not getting the right amount of juice. I guess, I have to go back to a dedicated preamp.

Having said that , I think Mytek would work great with high sensitivity amplifiers.

And to those who are interested to get the remote for Mytek, you could buy RM20 remote which looks like Apple remote.

Having disposed of the Supratek amplifier, I am now using the minimalistic approach with no preamp. It every audiophiles' dream to have as little as possible from the source to the speakers.

So do you really need a preamplifier? Unless, we are using analogue source or a Pro power amplifier then the use of preamplifier is simply redundant as the function of the preamp is not to amplify the signal but to attenuate.

Previously, I mentioned that Mytek lacked the juice which after exploring the reason for it turned out to be simply the case low level source and without a preamplifier there isn't enough juice to bring up the level. But this also brings another problem. How often are we clipping our amplifier?

Going back to Mytek, the maximum output is 4.5Vrms for the unbalanced output. While the amplifier requires on 1.6V for maximum output of 500W at 4Ohm. To prevent the amplifier from clipping, I have to reduce by 9dB so the maximum output from Mytek never exceeds 1.6V.

By using this level, most of the well recorded classical and some LINN recordings play at the comfortable level of 70 to 80db and peaks often reaching 100dB. This has worked with most albums. Most pop albums with small dynamic range require attenuation of -14, -16, -18 and -21dB. This is constant across the board.

You have to hear it to understand what is loudness compression, clippings and distortion. So far, it worked well and for one or two album where the level is relatively low, I use the inbuilt JRivers equalizer preamp to bring the level not exceeding 100% output with Mytek level at -9dB (I usually leave it at -10dB, just to be be on the safe side).

Give it a try. Less is more.

You said "I am now using the minimalistic approach with no preamp"

Actually the Mytek DAC has a built-in preamp, either analog attenuation (from the pictures, very likely a resistor-based attenuator inside an integrated circuit) or digital attenuation by ESS Sabre DAC.  It also has volume trim.  So you are indeed using a preamp which is built-in inside the DAC box.

I will share my own experience with preamps and I will skip the entry level models.  I tried Belles 28A (bought used) which uses ALPS potentiometer, then went straight to Audio Research REF 5 SE.  I was initially happy with it, especially the first time using a tubed preamp; the emotions and musicality made me happy.  But due to impedance mismatch with my McIntosh power amp, I am not getting the best yet - the highs are rounded and the bass lacking.  Also, since my stereo system is integrated with my home theater where the AR preamp HT bypass is used during multi-channel listening, I am wasting the tube hours in HT bypass mode which is undesirable.

So I tried building my own passive preamp using components like slagleformer (actually autoformer) and relay-based volume control circuitry from Bent Audio, Canada.  I started with the copper version, built a fully balanced version (2 autoformers per channel) and included 3 additional autoformers inside the same chassis for center and rear channels gang volume control.  It's a pure 2-channel and multi-channel passive preamp in a box.  No need for HT bypass.  And this autoformer-based preamp outperformed the AR REF 5 SE.  Higher resolution and very quiet background.

http://www.audiocircle.com/index.php?topic=129936.0

Earlier this year when conversion rate to USD was around 3.6, I built another one, this time using pure silver wire autoformer (also from Bent Audio) and silver internal hook-up wires for the front channels.  This time, I put the center and rear channels copper autoformers inside another box with volume control cable connected externally.  Now the music just flow out naturally - I can't hear the existence of a preamp.  The resolution and control are now at different level.  I think I have found the end game for a preamp.



Note: the same autoformer by Bent Audio is being used in exotic preamps like Purity Audio where the stereo preamp with copper autoformer costs USD22k (but it has a lot more exotic components).  Due to OEM customers' pressure, Bent Audio no longer sells the components to DIY enthusiasts but only to their OEM customers.

http://www.purityaudiodesign.com/comparison.htm

A preamp is always required in an audio system.  We can choose pure analog active or passive line level preamps or digital-based attenuation typically implemented inside DAC chip.  Some DACs and CD players have built-in preamp.

If your source components have enough voltage to drive power amp directly, please don't be afraid to try passive preamps, either transformer-based or LDR.  If you are interested in transformer-based passive preamp, you can even start with a locally made product, Promitheus.  I have not tried it myself but occasionally I see the used ones advertised here.

Anwar wrote:....

Actually the Mytek DAC has a built-in preamp, either analog attenuation (from the pictures, very likely a resistor-based attenuator inside an integrated circuit) or digital attenuation by ESS Sabre DAC.  It also has volume trim.  So you are indeed using a preamp which is built-in inside the DAC box...

I am actually referring to the use of preamp in the digital domain. In analogue age, we needed a preamp to amplify the low signals from the source but in the digital world most of the outputs are at least 2V. 

There are not many modern amplifiers that require more than 2V for full rated output. What we doing with a preamplifier in this chain is to attenuate the signal to bring the signal lower. In a normal domestic listening environment, we would be hard pressed to use more than 1V for a comfortable listening level of 80dB.

I am no DIYer, but IMO, Mytek doesn't amplify signal. It just attenuate the signal and therefore no (pre) amplification is involved. The output trim attenuates in the digital domain to prevent clipping and the gain pad reduces another 6dB via resistors. I use none of that as the Classe amp and most high end amplifiers power rail could take the higher signal level of the DAC ouptput.

In the case of a passive preamplifier, the role is simply to attenuate the output level without amplification. This is fine provided you do not have impedance mismatch. But if you could already do the attenuation at the earliest stage in DAC then why add extra connections to do the attenuation with a passive preamp? That itself would degrade the signal.

....
But due to impedance mismatch with my McIntosh power amp, I am not getting the best yet - the highs are rounded and the bass lacking. ...

Thanks for highlighting this. I am also having problem of understanding what should be the correct impedance. Your McIntosh input impedance is 22/47kOhm for regular/balanced, the Belles output impedance is 50 Ohm. The ratio is more than the minimum recommended 1:10. Technically, if the cables are within the normal specification and reasonable length there shouldn't be any attenuation of high frequencies. Can you do some measurement to confirm this? I am still confused about this part.

I too did experience missing highs initially.

If there is no gain in the line level signal, it is called passive preamp.
If there is a gain, it is called active preamp.
As long as there is volume control, there is a preamp.

It seems the Mytek is a resistor-based passive preamp when using pure analog input.

But in digital domain, it does have a high current output stage right after ESS Sabre DAC.  The resistor-based volume control comes after the output stage.  You can check the signal flow diagram in Mytek website.

The impedance mismatch I had with McIntosh was with Audio Research REF 5 SE preamp which requires output impedance > 100k Ohm.  It was sold to another audiophile who uses Audio Research tube mono block power amps and with proper impedance matching, it works flawlessly.

I always thought a volume control is just an attenuator. It cannot increase a signal on its own.

Quoting Mytek.

Volume Control
There are several methods of volume control available:
Analog - The output level is controlled by a programmable analog
fader. This also allows independent level control of the main output
or headphone output.
Digital - The output level is controlled digitally with ultra precision
32bit digital attenuator...

You can control (attenuate) the volume either digitally or via the programmable analog. (I am not sure how this is analogue.). If you look at the flow chart, digitally it is done before the DAC chip and the analog fader bypassed. Or bypassing the digital volume control and use the analogue fader.

The purest signal should be straight without any faders to the amplifier. In my case, that would be too loud for some albums and would definitely overdrive my amp.

Going back to your Acoustic Research preamp, the specs given by the manufacturer...

Output Impedance: 600 ohms Balanced, 300 ohms SE Main

.

The 600 Ohm been there since the telephone era and way below the input impedance of the Mcintosh.

Anyway, I still don't understand how this is relevant to have the cleanest and simplest signal path without any degradation or coloration.

You may prefer the passive or other preamps in the chain but what would be the best way to preserve the signal? More connections or less? Subjectively, I prefer the Supratek preamp but that's not the subject matter here. High fidelity is to be faithful to the original signal.

See the real-world measurement of AR REF 5 SE by Stereophile.  Frequency response is flat with 100k impedance.  Any impedance lower than that means there will be some frequency roll-off.

http://www.stereophile.com/content/audio-research-reference-5-se-line-preamplifier-measurements#6Zy3feLLSkverIGI.97

Audio Research technical support responded to my email stating that the tubes will work harder with lower impedance, hence the recommendation of 100k ohm or higher even though the specs said otherwise.

Back to your question about how to keep the signal faithful or pure, I will avoid trim or attenuate in digital domain since it will reduce resolution.  It is best to do attenuation in analog domain.  That means please use the analog fader and don't do any digital trimming.  You should also set the -6dB jumpers if it is overloading your power amp.



Programmable analog fader means when changing volume level, the controller microprocessor sends signal to the analog volume control IC to select which resistor value to use.  The signal path is purely analog despite going through an IC.  A lot of preamps are using this method.

Anwar wrote:See the real-world measurement of AR REF 5 SE by Stereophile.  Frequency response is flat with 100k impedance.  Any impedance lower than that means there will be some frequency roll-off..

But in the same page John Atkinson ( who took the measurements) also stated

I recommend the Ref 5 be used with power amplifiers having an input impedance of at least 10k ohms, when the preamplifier will be at its most linear at typical operating levels.
Read more at http://www.stereophile.com/content/audio-research-reference-5-se-line-preamplifier-measurements#rIJc3dW75YtJ6sCX.99

. So any amplifiers with input impedance of at least 10k Ohm should be fine.

Btw, the review was made using a Mark Levinson 50 kOhm amplifier and a VTL amplifier.

Thanks for the analog volume explanation.

The internal analog fader (volume control) is the last thing when it comes to purist design. This is just another additional circuitry (most probably yet some op-amps) in the signal chain. Bypass the analog fader and use a high quality passive volume control for minimalist purist approach. As rightfully pointed out, matching is crucial here to get the best out of the music reproduction.

Anwar,
Like you approach on the passive volume

Thank you vt4c.  Luckily I completed the projects before the current level of exchange rate.  Here is the link for Bent Audio components.  

http://bentaudio.com/index2.html

vt4c wrote:The internal analog fader (volume control) is the last thing when it comes to purist design. This is just another additional circuitry (most probably yet some op-amps) in the signal chain. Bypass the analog fader and use a high quality passive volume control for minimalist purist approach. As rightfully pointed out, matching is crucial here to get the best out of the music reproduction.

Mytek's own findings seemed to suggest that people prefer analogue volume over their digital. To my ears, I am happy with digital for their precision.

Limiting to DAC as preamplifer, I cannot see the technical basis that an additional component in the signal path could produce more pristine signal than dealing everything in the digital domain at the DAC level.

Getting a passive analogue preamplifer without channel inbalance is a challenge. After crossing that bridge, we are faced with multiple connections.

One pair of interconnects to the passive preamp. Then another connection for the selected level. Then we have another connection for the selector. After that we have another pair of interconnects to the amplifier.

Somehow, this doesnt concern the purists. Double the length and multiples breaks between the signal.

Your concern on multiple connection and interconnect is valid.

I've tested some of these electronic volume i.c. and it doesn't shine.

vt4c wrote:Your concern on multiple connection and interconnect is valid.

I've tested some of these electronic volume i.c. and it doesn't shine.

For the benefit of other members here, why not share the names of the i.c.? Do you have any brand that you would recommend and what do you think of PGA2311?

FYI the TI PGA2311 2-channel analog volume control IC price at mouser.com is US$12 for 1 piece and the price is lower for quantity purchase.  Based on the data sheet, it does have internal op amp.

IC volume control is also used in high-end A/V preamp/processor like the US$9,500 Bryston SP3 which uses US$22 Cirrus Logic CS3308 8-channel analog volume control IC.

Given the cost of the volume control IC, the IC manufacturer does have constraints in selecting the quality of the internal components used.  That's why DIY enthusiasts choose DIY route in making preamp.

The op-amp is redundant if used within -95 to 0 dB. In most DAC, the volume control usually used to attenuate and not for gain.  So I wouldn't worry too much about the op-amp unless gain is involved.

You should stop worrying but enjoy music instead.  Please read this article and don't worry so much about analog volume control IC.

http://www.theabsolutesound.com/articles/pass-labs-xp-30-reference-line-preamplifier/

The flagship US$16,500 Pass Labs XP-30 preamp uses US$25 NJR Muses 72320 IC, a two-channel electronic volume control with an internal resistance-ladder, optimized for low noise and sound quality.

You are playing devil's advocate, aren't you?  How can one peacefully enjoy music if still reading TAS....  

I can't remember the exact part number but have evaluated volume i.c. from TI, Analog Devices and Wolfson. Most, if not all has some form of internal op-amp in the signal path within the i.c. chip

Based on the PGA2311 datasheet (pg. 7, fig. 1), there is always an op-amp in the signal chain irrespective of what level the part is set.

vt4c wrote:I can't remember the exact part number but have evaluated volume i.c. from TI, Analog Devices and Wolfson. Most, if not all has some form of internal op-amp in the signal path within the i.c. chip

Based on the PGA2311 datasheet (pg. 7, fig. 1), there is always an op-amp in the signal chain irrespective of what level the part is set.

 

Thanks. I always thought the op-amp only engaged above the unity gain. At least, that was what claimed by an employee of TI in one of the forums. 

I am happy with the digital volume and if I ever need any gain, I am sticking to software part of JRiver. Whatever floats my boat. )