On Step Ups and the Importance of Managing Gain


The past several weeks have been some of the most eye opening in my entire audio journey. Despite over 30 years listening to LPs across a range of setups it seems I've only just stumbled across one of the secrets of analog reproduction -- the step up transformer AKA the importance of managing gain across the analog reproduction chain.

To understand what I mean you need to start with an understanding of my gain cascade prior to my most recent changes. My cartridge is an Acoustical Systems Palladian which is a relatively low output MC at 0.33mV. I've been running it into an ARC Ref 2SE Phono and then an ARC 40th pre. Both are connected balanced so the gain is 51dB (for the phono stage) plus 12dB (for the pre) = 63dB at the LO gain setting -- I tried listening to the HI gain (74dB) setting on the phono but found the added FET amplification added an edge to the sound, amplified hum and restricted dynamics. I'd therefore learned to live with having the Ref 40 volume control on the high side (about 70-80% of range) and "thought" I was getting good sound - great clarity, sound-staging and frequency delineation.

However with a pending upgrade in my phono stage to the VOSS phono, a pure 40dB only single ended affair, I was forced to add a step up to my setup. I've always sworn off the added complication and additional switching and cables so had assumed it would not do anything to improve things. How wrong I was. The model I have is the Music First Audio Classic V2 in a custom configuration built for me with two inputs (one for my mono arm and cartridge) and fixed 20dB gain (1:10 ratio). It also has a three way ground lift switch so any hum issues can be addressed at source (which works btw). At about $3K it's not cheap (and MFA have cheaper options) but in the scheme of things actually much less than the list price of the interconnect I needed to add to connect it to my phono stage (so in the context of my system a stone cold bargain). 

My gain cascade is now 20dB (step up) plus 45dB (phono, now single ended) plus 12dB = 77dB and the volume control is at 40-50% of range. Once I'd realized that running a step up into 200 ohms is not going to work at all and switched the ARC Ref2SE to 47K I was off -- and was I surprised! All of the sudden the soundstage opened up by 50% side to side, front to back and up and down. Dynamics on individual instruments suddenly snapped into focus and the overall presentation was transformed in a way I would not believe possible. One instance can serve as a sense of what I mean -- the two opening cuts on Joni's Hejira, "Coyote" and "Amelia", are mostly similar sounds layered over one another and can be pretty hard to sort out. With the new step up in place it's as if everything now makes sense and each instrument is positioned in space perfectly in location and scale with all the others.

So what's my take away and potential advice to any reading this?
  1. High gain (i.e. 60dB plus) phono stages are, as we all know, problematic so if you can avoid going that way consider looking into it, BUT
  2. Don't "push your gain" i.e. try to under drive a stage, you may not know it (as I didn't) but trying to amplify .33mV with "only" 51dB is probably losing a lot of detail into the noise floor of your amplification stage and/or giving you problems down stream if gain is set too high
  3. So given 1 and 2 try a step up -- you may be surprised as I was!
Not sure if this experience matches with others and would love some perspective from amplifier designers as to why I may have had such a bad (in retrospect) experience with my prior gain cascade.
128x128folkfreak

Showing 6 responses by almarg

Is it true that during loud passages the cart output can increase 10x and run into a ceiling in the phono which also causes distortion?
Yes, brief dynamic peaks on some recordings can significantly exceed the standard 5 cm/second test condition. My understanding is that a phono stage overload margin of 10x (which corresponds to 20 db), relative to a typical LOMC output rating of 0.5 mv, is an ample margin, but that some well regarded phono stages fall at least several db short of meeting that number, at least at some frequencies, and still do fine with all or nearly all recordings.

As you may be aware, the measurements section of Stereophile’s phono stage reviews usually present data on overload margins at various frequencies, and at various gain settings if gain is adjustable.

An additional concern that can arise in some cases, though, would be the possibility of overloading the input stage within the line stage, if the line stage is designed such that its volume control is not "ahead" of the active circuitry in that stage.

Regards,
-- Al

Folkfreak, yes, that is correct.  Everything else being equal, reducing the amount of feedback will reduce damping factor, increase gain, and make the amp more sensitive.  Meaning that the sensitivity number, i.e., the input voltage required to drive the amp to its maximum power capability, will become lower.

Regards,
-- Al
 
Robd2, the KAB gain calculator addresses only the gain provided by the phono stage. Also, it can be very important to note the following statement that is provided in its introductory text:
The optimum gain is based on acheiving [sic] 325mV rms output at 5 cm/s. For the current crop of CD recorders, 300mV is required for 0dB recording level with the recorder’s level control set at max. Aiming for 325mV gives a little margin.
So the calculator will not necessarily provide valid results in a situation where the phono stage is driving a line stage preamp which in turn is driving a power amp. For example, a reasonable line stage gain of say 10 db would boost that 325 mv to about 1 volt if the volume control is set at max. More often than not, 1 volt will not be enough to drive a power amp to full power. And that's with the volume control at max!  And of course there are many line stages which provide less than 10 db of gain.

So take the results provided by this calculator with several grains of salt, and throw in some additional db as well :-)

Regards,
-- Al

Raul & Michael (Swampwalker), you’re both right. Regarding the multiplication effect Michael mentioned, the transformer will of course step down the resistive component of the input impedance of the phono stage as it is seen by the cartridge, but at the same time it will step up (multiply) the input capacitance of the phono stage and the capacitance of the cable connecting the SUT to the phono stage, again as seen by the cartridge. That will tend to work in the direction of lowering the frequency and increasing the amplitude of the resonant peak in frequency response resulting from the interaction of cartridge inductance and the load capacitance applied to the cartridge. In the case of LOMCs that peak occurs at RF frequencies, but may affect sonics to the extent that the design of the phono stage allows audible frequencies to be affected by RF energy that is fed into it, and to the extent that the bandwidth of the SUT allows frequencies in the vicinity of that peak to be conveyed to the phono stage.


Also, differences in loading can affect the behavior of the transformer itself, including its susceptibility to effects such as ringing. The concept of an “ideal transformer” is often used in electrical engineering, because modeling a transformer as a theoretically ideal device can be useful for purposes of circuit analysis. But an “ideal transformer” is something that in the real world can only be approximated. So the noise-free gain provided by a SUT, that was referred to earlier, does not necessarily come without tradeoffs.

Atmasphere has made similar points in a number of past threads.

Regards,
-- Al


Assuming I’m getting the same voltage level output from the preamp with or without the step up why would the performance with the step up be so different? Is it inherent to the step up or is it also a problem of trying to amplify too small a signal at the pre amp (or for that matter at the phono stage).
I wouldn’t want to speculate as to the specific technical factors that may be involved, at least without being intimately familiar with the design of the phono stage. I would just note, again, that the two gain choices provided by the phono stage are very atypical, while the output level of your cartridge is well within the bounds of being typical for an LOMC.

I’ll add, also, that the gain of your VTL monoblocks is several db greater than the gain of most power amps, to a greater or lesser extent depending on the mode and feedback settings you are using. So with many other amps the low upstream gain you had prior to the addition of the SUT would probably have resulted in your 70% to 80% settings of the volume control being in the vicinity of 100%.

Best regards,
-- Al


Interesting post; thanks! And congratulations on the improvement you’ve realized to what was certainly an already exceptional system.

My feeling is that the root cause of the issue in this case may very well have been ARC’s choices of the gains provided by the Ref2SE, which seem a little strange. 51 db being too low for most LOMC applications (and too high for most MM applications), and 74 db being too high for most LOMC applications.

51 db represents a voltage multiplication of 355 times, which would raise the output of a 0.33 mv cartridge under the standard test conditions to only 117 mv. The very reasonable 12 db gain of your line stage would raise that to about 466 mv if the volume control were turned up to max, which would be much too low to drive most power amps to full power. While 76 db phono stage gain is typically best suited to cartridges having outputs that are considerably lower than 0.33 mv.

More typical for use with a 0.33 mv cartridge would be a phono stage gain of a bit more than 60 db, close to the 20 + 45 db you are now using.

Of course, you’ve changed several things in this process in addition to modifying the gain structure, i.e., inserting a transformer into the path; changing the load presented to the cartridge from 200 ohms to 470 ohms; and changing the interface between the phono stage and the line stage from balanced to single ended; and I suppose all of these things may have contributed to the resulting sonic differences to some degree.

BTW and FWIW, my 64 db Herron VTPH-2 seems to be an excellent match for my 0.5 mv cartridge. And during my discussions with Keith Herron prior to the purchase he indicated that the 64 db configuration of the VTPH-2 would have been a better choice for use with a 0.3 mv cartridge I was considering at the time than the 69 db configuration he also offers.

Best regards,
-- Al