Strain Gauge Preamp/ Phono. A new standard?

Great topic, Folke

Soundsmith claims that their cartridge does not require RIAA equalization. Since that equalization is incorporated into the recording, I can't understand how in can be ignored on playback, regardless of the operating principle of the cartridge.

Can any one explain this?

Check out the CES coverage here on the Gon.Soundsmiths' video covers the operating theory and its' implementation.Would love to hear one!

I agree, it sounds like an interesting product. The bit about the RIAA doesn't make sense at all to me either, though.

All the Strain-gauge cartridges from the old days had the RIAA EQ built-in too.

I heard the Strain Gauge cartridge at the RMAF last October. It was one of the best sounding cartridges I have ever heard.

The strain gage cartridge requires its own special phono stage, since "signal" is a varying resistance, not a voltage that is put out. Perhaps they build the RIAA into their pre-amp? Just a guess.

Here's an interesting thread from an AA user:

http://www.audioasylum.com/audio/vinyl/messages/70/704410.html

If you view the video from Mr. Lederman, he states that because this strain gauge cartridge is a displacement device and not a velocity-sensitive device, no RIAA equalization is required. Here's more info from their website:

http://www.sound-smith.com/cartridges/sg.html

I very much appreciate Audiogon's great coverage of the Soundsmith Strain Gauge Phono system, and for having the courtesy of posting this hopefully helpful and informative response:

The Strain Gauge is a DISPLACEMNET device, not a velocity sensitive device like all magnetic cartridges.

Velocity sensitive devices have a FLAT playback response, so they require RIAA EQ to invert the RIAA recording characteristic used when making masters for record production. RIAA has an interesting evolution; Mr. Galo has written a wonderful explanation of how, and why it evolved and changed over the years:
http://www.smartdev.com/RIAA.html

Displacement devices have an INHERENT 6dB/octave curve, very similar to the RIAA (not an accident, BYW)....they DO deviate slightly from the RIAA at two points, BUT, we do NOT correct for this, for doing so would ruin the beauty and simplicity of the device; the one super clean gain stage between you and the music we designed is almost one too many. More than one (to "fix" small anomalies) would be death to the magic. We do compensate passively for the low end roll, as required.

For those who may feel that absolute compliance with RIAA is mandatory, I would suggest that they take a microphone and laptop and make a response measurement for flatness at their respective wonderful listening positions. They may find that not only does it deviate far more than "less than one dB" as we do, but that "correcting it" with additional circuitry puts so much in the signal path that they join the ranks of anti-magicians.

Peter Ledermann/Chief Engineer/Soundsmith Corp.

I had strain gauge cartridges long ago, including the Win and the Panasonic. I saw them as quite good, but not reaching the realism of my Decca Londons. There are many benefits in the various grades of the Soundsmith phono/linestages, but it does take a commitment to this idea that I am not willing to make.

Hi Peter, the RIAA curve is flat from 318 uS to 3180uS. So there are about 4 octaves that will be spectrally incorrect. That is a lot bigger error than any room I've tested. BTW I've run plenty of room curves over the years- especially at shows. I agree that you see big variations, but not like this (at least not so far anyway).

I've also produced a variety of LPs over the years and I've found that only 1 db variations in the test pressing can be audible (this is easy to hear over a frequency spectrum as opposed to a single frequency, BTW).

So- do you offer an equalizer?

Thank you for your comments: they are greatly appreciated.

The subject of human response to amplitude and phase is an interesting one; the standpoint often determines the course of design choices for those of us in the audio design and manufacturing field. I, for one, strongly believe that the human ear is much more forgiving in the amplitude domain than it is in the time domain. After all, many amplitude deviations are the direct result of phase interaction due to resonance, both summation and cancellation, making "amplitude" perturbations audible and potentially objectionable in speaker drivers and cartridge/tone arm systems.

From that standpoint, the cutter heads models also vary considerably; many require one or more frequencies to be fed back with control of both amplitude and phase to stabilize the cutter head. The phase shifts introduced by the RIAA recording curve are further modified by the varied individual characteristics of the equipment used to cut the master. So, where does that leave us? The Strain Gauge is without much phase shift at all. If one leaves out the phase shifts normally introduced by use of a playback RIAA curve, is one better or worse off??


Please also refer to the graph near the top of Mr. Galo’s excellent article called Figure 1;

http://www.smartdev.com/RIAA.html

It refers to this graph as the “typical” RIAA response; in reality, it is far from the “ideal” response that one gets from the theoretical filters due to interaction of those filters. This, not the “ideal”, is what we have all been listening to for years with magnetic cartridges! Superimpose a 6dB/octave curve (Strain Gauge) and you will find that it deviates less than plus or minus one dB.

My experience is that severe phase perturbations over a narrow region are very audible, while those that are spread out are much less objectionable. It may well be that this is one of the positive differences one experiences when listening to the Strain Gauge. That, in combination with the general standpoint of multiple stages being inferior to a single stage, makes one think that possibly that too, is a motive for using a "simplified" strain gauge system. There are also numerous other contributors to good sound - extremely low stylus jitter being one.

Although some Strain Gauge preamps in the past have been made to more closely follow the RIAA, some were not. As far as I am concerned it is an open issue that will be addressed from varied standpoints, some of which I have mentioned above. My consideration in the Soundsmith Preamp is to make it mostly flat; the decision to mainly use the inherent 6dB curve of the Strain Gauge elements is based on 35 years of listening and talking with other designers and SG owners. It also provides a standard from which one is not precluded from adding what they believe to be proper EQ to our system, or Soundsmith offering such in the future to those who want it and do not feel that it detracts from the listenability.

The proof is truly in the listening; many who have listened in our rooms or who have owned our Strain Gauge or Strain Gauge systems in the past that employ flat response have enjoyed the result. If the system as I produced it was very, very wrong, I believe that there would have been many more negative opinions than positive. Again, thank you for your insightful comments -

Peter Ledermann/Soundsmith

So no EQ unit then. If you refer to Stanley Lipshitz's paper, presented to the AES back in the late 70's (arguably the definitive paper on RIAA equalization), it is possible to design an EQ network that will follow the mandated curve correctly, easily better than 1/2 db. He presented very easy formulae to work out practical and accurate values.

It seems that at the upper end of the 'flat area' of the RIAA curve (318uS), that a simple 6db/octave EQ will be off by nearly 24db.

Don't get me wrong- as a manufacturer I think I can safely say that all phono section designers struggle against the issue of noise that is possible when you are trying to amplify a signal from a low output MC cartridge. There is much allure to the idea of a cartridge that can put out a volt or two. But the EQ being off by that much... Wouldn't a proper EQ make the cartridge more attractive?

Again, thank you for your comments, as it allows me to try to illuminate issues about the Strain Gauge cartridge that have been long misunderstood.....

My comments again, are not to present a position that an EQ absolutely cannot be used; I am however, indicating that according to many, it is not required, and to do so would result in less performance for a variety of reasons I have tried to illustrate.

I am also concerned that your figure of 24dB is very far off from the reality of the device; it would make it unlistenable! Please observe the "real" RIAA curve in Galo's figure 1, and superimpose a 6dB/octave line on this "real" world RIAA graph, referencing both at 1Khz. The greatly simplified result is that in the response below 1KHz, the SG would be slightly boosted, and above 1K, slightly attenuated. For a cartridge with 70Khz bandwidth and immediate speed of the Strain Gauge, this slight upper end attenuation is not missed at all as there is a definitive acoustic signature that many notice, which more than makes up for this slight "loss".
I have never heard anyone complain that the Strain Gauge was "dull" sounding at the top end. Quite the opposite, if anything. So, with that in mind, a 24dB error (gigantic!) would make it unlistenable, which it certainly is not. By superimposing the curves, you will readily see that any "correction" required is very small indeed. Not 24dB!!

I am aware that I am sounding like a "broken record", but "fixing" the small errors would be equivalent to a successful operation wherein the patient dies, in the opinion of some. I am in that camp. My audio philosophy has been ~ Ikebana ~ "Less is more" if done right; this philosophy is hopefully balanced with the other I hold dear: nothing is perfect - all designs are compromises formed by real world constraints and lots of experience, or lack thereof. I have been, and will continue to be, guilty of both.

Noise issues:
When you say, "....allure to the idea of a cartridge that can put out a volt or two." are your referring to the Strain Gauge? If so, that would be an incorrect attribute for the Strain gauge. While ceramic cartridges (displacement devices as well) DO output high voltage, they are also high source impedance devices. The Strain gauge is absolutely neither of those, although I run into many who do think it is. The source impedance is a fairly low, purely resistive 1000 Ohms, and the current supplied is 4 milliamps. The amount that the element varies is less than .1%; this means that the change in voltage is in the range of 3 - 4 millivolts, requiring a gain in the high thirties.

Noise?? In fact, the RIAA EQ inherently reduces high frequency noise and distortion in "normal" preamp circuitry; making RIAA a Dolby of sorts. The absolute requirement for a fairly flat (EQ or not) preamp for the Strain gauge was a bit more problematic 35 years ago, as we had to create low noise circuitry for this "flat" preamp to avoid excessive noise compared with a standard RIAA preamp! Creating a circuit in this day and age with "ultra low noise and immeasurable distortion" is a joy, and very do-able. I would rather listen to this one gain stage and an ultra fast cartridge than anything else.

Now, .....although I have broken my oath to not spend too much time on the blogs.... I must now get back to work retipping all sorts of carts, listening, designing, making mistakes, and building Strain Gauges for the rarified few who have heard the new one........and now are spoiled. What have I done??? !!

Peter Ledermann/Soundsmith (in New York state, where a PhD. in Rudeness is readily available as a college Major)