The question was asked- why the focus on certain aspects of the design- namely extremely low high order harmonic distortion, extremely tight RIAA compliance, excellent many tone intermodulation distortion and low output impedance for the driving stages.
In my opinion, these are the areas that allow the phono stage/preamp to produce such an apparently high audio quality. The conclusions that these are the correct areas to focus on are largely based on psycho acoustic findings in the last 30 + years (for example the AES GedLee papers from 2003).
In essence low order harmonic distortion is masked by the human auditory system, whereas high order (i.e. 9th) distortion is not masked and can be audible even at levels as low as 0.01%. So the amplifier needs to keep this high order distortion lower than 0.001% to provide an audibility margin. This is particularly the case at relatively low signal levels, and for distortion products up to the limit of human hearing (i.e. 20kHz).
So, for example the 9th order harmonic distortion of a 2.2kHz input is a useful metric, especially versus input level, while THD is not. Hence the focus on having sufficient feedback at 20kHz, together with sufficiently low open loop distortion for the opamps chosen.
The RIAA compliance is a reflection of the fact that many people can detect broad frequency response variations of <0.25dB, whereas there is no evidence that I can find that people can detect <0.1dB variations. So, the goal was to have <0.1dB p-p RIAA error in order to produce an objectively/subjectively neutral frequency response.
This also led to the development of the "warp" filter which is intended to not disturb the RIAA characteristic, as embodied in the LP format, while allowing for the rejection of many unwanted LF artifacts.
Finally, the extremely good many tone IM performance, including ultrasonic frequencies up to 40kHz or so, recognizes that in the presence of complex (music) waveforms intermodulation creates a elevation in dynamic noise floor and an effective reduction in dynamic range, a degradation in attack characteristics, and a loss of clarity and imaging quality, so it has to be exceptionally low.
These are issues that are rarely even mentioned in preamp design, and even more rarely measured as they have been in this design.
The choice of gain structure, and opamps, and local gain stage frequency responses are all made to optimize these parameters and to maximize the wideband dynamic range before overload of each stage and ultimately the overall design.
Now as far as output impedance was concerned. The design was intended to be load agnostic as much as possible so that the dependence on intercomponent cabling characteristics was reduced. Small Rs have been added to each output to provide some stability protection with large values of load capacitance, but otherwise the focus has been on using opamps with low open loop output impedances and very low closed loop output impedances.
Opamp rolling is actively discouraged for the design. Each stage has had an optimal opamp chosen for it.
In my opinion, these are the areas that allow the phono stage/preamp to produce such an apparently high audio quality. The conclusions that these are the correct areas to focus on are largely based on psycho acoustic findings in the last 30 + years (for example the AES GedLee papers from 2003).
In essence low order harmonic distortion is masked by the human auditory system, whereas high order (i.e. 9th) distortion is not masked and can be audible even at levels as low as 0.01%. So the amplifier needs to keep this high order distortion lower than 0.001% to provide an audibility margin. This is particularly the case at relatively low signal levels, and for distortion products up to the limit of human hearing (i.e. 20kHz).
So, for example the 9th order harmonic distortion of a 2.2kHz input is a useful metric, especially versus input level, while THD is not. Hence the focus on having sufficient feedback at 20kHz, together with sufficiently low open loop distortion for the opamps chosen.
The RIAA compliance is a reflection of the fact that many people can detect broad frequency response variations of <0.25dB, whereas there is no evidence that I can find that people can detect <0.1dB variations. So, the goal was to have <0.1dB p-p RIAA error in order to produce an objectively/subjectively neutral frequency response.
This also led to the development of the "warp" filter which is intended to not disturb the RIAA characteristic, as embodied in the LP format, while allowing for the rejection of many unwanted LF artifacts.
Finally, the extremely good many tone IM performance, including ultrasonic frequencies up to 40kHz or so, recognizes that in the presence of complex (music) waveforms intermodulation creates a elevation in dynamic noise floor and an effective reduction in dynamic range, a degradation in attack characteristics, and a loss of clarity and imaging quality, so it has to be exceptionally low.
These are issues that are rarely even mentioned in preamp design, and even more rarely measured as they have been in this design.
The choice of gain structure, and opamps, and local gain stage frequency responses are all made to optimize these parameters and to maximize the wideband dynamic range before overload of each stage and ultimately the overall design.
Now as far as output impedance was concerned. The design was intended to be load agnostic as much as possible so that the dependence on intercomponent cabling characteristics was reduced. Small Rs have been added to each output to provide some stability protection with large values of load capacitance, but otherwise the focus has been on using opamps with low open loop output impedances and very low closed loop output impedances.
Opamp rolling is actively discouraged for the design. Each stage has had an optimal opamp chosen for it.
I think you might have misinterpreted Earl. You'd be doing really well to get the THD under 2% with vinyl replay. Nothing wrong with chasing ultra low disortion in the phono preamp but it's swamped by the source.
I also think you are missing the key point with the elliptical filter and what it actually does that is interesting. I'll try and write up my thoughts for discussion.
I also think you are missing the key point with the elliptical filter and what it actually does that is interesting. I'll try and write up my thoughts for discussion.
I understand perfectly well the harmonic distortion, and at least some of the mechanisms that produce them, associated with phono cartridges. I have characterized several extensively, and there's hardly a shortage of information out there. My understanding that the distortion is primarily 2nd and 3rd, with some fifth, with no evidence of higher order components above the noise floor. As the distortion is essentially restricted to low orders it does not violate the masking criteria that the high order constraint is targeting.
There are other details of the argument- concerning the action of negative feedback on the intrinsic distortion generated by the cartridge- but that will be left to another time.
in that sense your argument is a straw man one.
I believe that I have a fundamentally sound grasp of the function and the consequences of the warp/elliptic filter, both during the cutting process and the playback process. However, I am more than happy to be enlightened.
There are other details of the argument- concerning the action of negative feedback on the intrinsic distortion generated by the cartridge- but that will be left to another time.
in that sense your argument is a straw man one.
I believe that I have a fundamentally sound grasp of the function and the consequences of the warp/elliptic filter, both during the cutting process and the playback process. However, I am more than happy to be enlightened.
Indeed, there are some cartridges that do generate measurable amounts of high order (i.e. 9th) harmonic distortion, particularly in the vertical direction (the AT VM760SLC, for example). I cannot comment on their sound qualities.
I am aware of Earl Geddes metric- obviously. However, I don't find it a useful way to design amplifiers so I extracted what I consider to be relevant from the basis for the metric
The Gedlee analysis is well known and well perceived. However, there are several important observations that were made that can be used as general guidelines.
"
• The masking effect of the human ear will tend to make higher order nonlinearities more audible than lower order ones.
• Nonlinear by-products that increase with level can be completely masked if the order of the nonlinearity is low.
• Nonlinearities that occur at low signal levels will be more audible than those that occur at higher signal levels.
Again these may seem intuitively obvious.
"
These are used as the basis for the Gedlee metric, but in my opinion they lead to some conclusions concerning desirable circuit metrics that fall within the purview of classic distortion measurements/theory.
If you disagree, I would be happy to read your reasoning.
Amongst other references I also used the Cheever thesis.
https://next-tube.com/articles/Cheever/cheever.pdf
I would expect that you are familiar with that also, and I would welcome a discussion on the merits and interpretation, especially since he focusses more on the relative progression of the harmonic distortion components rather than the magnitude per se.
I am aware of Earl Geddes metric- obviously. However, I don't find it a useful way to design amplifiers so I extracted what I consider to be relevant from the basis for the metric
The Gedlee analysis is well known and well perceived. However, there are several important observations that were made that can be used as general guidelines.
"
• The masking effect of the human ear will tend to make higher order nonlinearities more audible than lower order ones.
• Nonlinear by-products that increase with level can be completely masked if the order of the nonlinearity is low.
• Nonlinearities that occur at low signal levels will be more audible than those that occur at higher signal levels.
Again these may seem intuitively obvious.
"
These are used as the basis for the Gedlee metric, but in my opinion they lead to some conclusions concerning desirable circuit metrics that fall within the purview of classic distortion measurements/theory.
If you disagree, I would be happy to read your reasoning.
Amongst other references I also used the Cheever thesis.
https://next-tube.com/articles/Cheever/cheever.pdf
I would expect that you are familiar with that also, and I would welcome a discussion on the merits and interpretation, especially since he focusses more on the relative progression of the harmonic distortion components rather than the magnitude per se.
I'd be interested in seeing your workings on the required 9th harmonic level compared with the vinyl noise floor as entirely possible I missed something, but given how hard it is to make an opamp gain stage that bad for high order harmonics I do wonder if it's the biggest worry in polishing the vinyl poop (I love vinyl for the experience but accept what a horribly flawed medium it is).
On the elliptic filter I personally think it's underselling calling it a 'warp' filter. Rumble filter is worse as the days of horribly noisy idlers is long gone. Whilst warp can be an issue on a badly matched or undamped setup, a filter isn't the cure as the audible damage is done, even if you stop the woofer flapping. The important part is that records are mixed to mono at LF in the mixing or mastering stage (drums nearly always in the middle after all) as there are hard limits on what you can cut and harder on what you can track in the vertical plane. Looking at the signal from a mid-side perspective (L+R and L-R, think Decca cartridge) the side channel is rolled off starting as high as 300Hz in some cases as the signal is blended to mono. Therefore anything coming off the record on the side channel at LF wasn't cut on there and can only be from the stylus flapping around and trying to hang on to the groove. People think about resonance in one plane as that is what many test records measure and at a single frequency, but reality is a lot less simple. So adding the blend to mono in playback is effectively removing stuff in the 20Hz to 200Hz band that shouldn't be there. This may to some be the alure of vinyl and some of it's 'sound' but purely personally I think having the ability to remove at least some of it is a good thing and people should be made aware of this (new to domestic replay) possiblity in enhanced sound.
If you just want to kill the woofer flap (and accept that the FM of the signal is still there) then there are better solutions for a really steep roll off below 50Hz IMO.
TL;DR I thinking you are underselling the potential benefits calling it a 'warp filter'.
On the elliptic filter I personally think it's underselling calling it a 'warp' filter. Rumble filter is worse as the days of horribly noisy idlers is long gone. Whilst warp can be an issue on a badly matched or undamped setup, a filter isn't the cure as the audible damage is done, even if you stop the woofer flapping. The important part is that records are mixed to mono at LF in the mixing or mastering stage (drums nearly always in the middle after all) as there are hard limits on what you can cut and harder on what you can track in the vertical plane. Looking at the signal from a mid-side perspective (L+R and L-R, think Decca cartridge) the side channel is rolled off starting as high as 300Hz in some cases as the signal is blended to mono. Therefore anything coming off the record on the side channel at LF wasn't cut on there and can only be from the stylus flapping around and trying to hang on to the groove. People think about resonance in one plane as that is what many test records measure and at a single frequency, but reality is a lot less simple. So adding the blend to mono in playback is effectively removing stuff in the 20Hz to 200Hz band that shouldn't be there. This may to some be the alure of vinyl and some of it's 'sound' but purely personally I think having the ability to remove at least some of it is a good thing and people should be made aware of this (new to domestic replay) possiblity in enhanced sound.
If you just want to kill the woofer flap (and accept that the FM of the signal is still there) then there are better solutions for a really steep roll off below 50Hz IMO.
TL;DR I thinking you are underselling the potential benefits calling it a 'warp filter'.
Your position on the "warp" filter is entirely correct. On the AK thread that is the actual source of the reference/design I go through the full set of implications, including the rather simple math associated with it.
The name was just a "catchy" one intended to allow people to relate to it, and besides, it does actually serve a useful purpose when there is a warp- it greatly reduces the "plump" sound that is generated by the untoward vertical movement of the arm/stylus structure while having a minimal effect on the horizontal signal which is the desired one. This has been demonstrated in "field tests" and the actual -3dB cutoff point (at 140Hz) was selected on the basis of audibility tests on warped records, as well as an understanding of the original action of the elliptic filter in the cutting/mastering process.
So, if you are in essence saying that the function is "damned by faint praise", I agree.
Now as far as the harmonic issue is concerned.
This is a tricky one, as essentially everything I say can be subject to criticism, and this format does not allow for correction/revision to permit more clarity.
in any case, I'll try to make my position as comprehensible as possible, and potentially extend the response over several posts as in essence I'm trying to congeal ideas extracted from many tens of pages of papers/theses etc. into a few concepts/parameters. However, whatever I say can only be considered to be an approximation, tailored to this limited format.
So, here goes.
The issue of high order harmonics.
The idea is that, in essence, there are two classes of audio systems- those with significant global negative feedback and those without.
Most transducers- such as phono cartridges, loudspeakers, microphones, are in the no global negative feedback class and have a different set of distortion rules applied when in comparison with those audio amplifiers which employ significant amounts of negative feedback.
Zero and very low amounts of negative feedback amplifiers are essentially in the same class as the transducers.
It needs to be accepted that the non-linear nature of the auditory system has been clearly demonstrated, and as a result the existence of Acoustic Harmonic Masking is a proven phenomenon.
That is, the "ear" generates/expects a regularly decaying harmonic structure when presented with a sine wave, for example. Any deviation from this harmonic structure will be audible, and the sensitivity to this deviation increases as the harmonic order increases. The rate of increase in this sensitivity is unclear, but there is evidence that several percent of 2nd harmonic distortion is acceptable, but 0.01% of 9th harmonic which is not accompanied by an appropriate amount of lower order harmonics is both audible and unacceptable. Hence, as long as the harmonic structure approximates the ears "desired" sequence the output is considered to be musical. If it does not, then it is considered to be discordant. This deviation in itself can be used as a metric for amplifier/transducer quality.
It is in this way that non-feedback systems and transducers can be acceptable even if they have high levels of intrinsic distortion, whereas systems with low levels of low order harmonics but relatively high levels (compared to the masking threshold) of high order harmonics may not be acceptable.
This is particularly true at low volume levels, due to the masking action, so amplifiers that demonstrate unacceptably high levels of high order distortion (i.e. sufficiently uncorrected crossover distortion) that exceed the sensitivity of the ear at low levels will be acoustically poor.
This tends to be the case with many relatively high feedback factor class AB amplifiers, for example, and justifies the poor reputation of amps that have good THD numbers versus those with worse THD numbers which have much better reputations.
However, there is an exception to this "rule". If all of the amplifier generated harmonics are at a level which is substantially below the audibility threshold (at about -110dBc) then the ear does not perceive them as being an add-on to the self generated harmonics and the acoustic "acceptability" will be equivalent to the ideally decaying harmonic case.
Most negative feedback systems have frequency dependent gain functions and hence feedback factors that are of an inverse frequency kind- i.e. the degree of gain/feedback factor falls as the frequency increases.
In addition the action of the input stage non linearities and the output stage non linearities in conjunction with negative feedback creates high order harmonic distortion products that have reduced amounts of feedback available to correct them.
The familiar distortion/feedback factor curves that have been around since these issues were first explored generally lack two cogent aspects- they do not capture the frequency dependent function of the feedback factor, and they also do not include the reduced level of intrinsic open loop distortion that can be caused by the presence of very high open loop gain at the frequencies of interest. i.e. if the input signal is very small then the open loop distortion created by the amplifier may also be small- so amplifiers with very high open loop gains and low open loop distortion alter the curves in beneficial ways.
Bruno Pudseys pointed this out in his inimitable style in an internet exchange.
So, what becomes important in all of this is, when considering the audible quality (or at least one aspect of it) of an amplifier with negative feedback, that it is vitally important to consider the amount of negative feedback available at least at 20kHz and hence the ability of the amplifier to correct the generated harmonics at this frequency to render them inaudible. It is also important to have amplifiers with low intrinsic distortion with the differential input signals actually present at the input of the amplifier.
This goal can be assisted by having adequate gain bandwidth products for the gain stages with applied negative feedback and if possible tailoring the frequency response of these stages to keep the feedback factor adequately high over the desired frequency range of operation.
The amplifiers/gain partitioning/location of the pole zero constellation in the RIAA deemphasis sections were chosen with these issues in mind. The goal was to keep all of the harmonic distortion products substantially below the noise floor (and < -110dBc) over the entire operating range of the preamp.
I hope that this makes at least some sense. I also hope that I haven't made too many typos etc. as I don't have time to make revisions at this moment and I have no idea how long I have to edit this post.
The name was just a "catchy" one intended to allow people to relate to it, and besides, it does actually serve a useful purpose when there is a warp- it greatly reduces the "plump" sound that is generated by the untoward vertical movement of the arm/stylus structure while having a minimal effect on the horizontal signal which is the desired one. This has been demonstrated in "field tests" and the actual -3dB cutoff point (at 140Hz) was selected on the basis of audibility tests on warped records, as well as an understanding of the original action of the elliptic filter in the cutting/mastering process.
So, if you are in essence saying that the function is "damned by faint praise", I agree.
Now as far as the harmonic issue is concerned.
This is a tricky one, as essentially everything I say can be subject to criticism, and this format does not allow for correction/revision to permit more clarity.
in any case, I'll try to make my position as comprehensible as possible, and potentially extend the response over several posts as in essence I'm trying to congeal ideas extracted from many tens of pages of papers/theses etc. into a few concepts/parameters. However, whatever I say can only be considered to be an approximation, tailored to this limited format.
So, here goes.
The issue of high order harmonics.
The idea is that, in essence, there are two classes of audio systems- those with significant global negative feedback and those without.
Most transducers- such as phono cartridges, loudspeakers, microphones, are in the no global negative feedback class and have a different set of distortion rules applied when in comparison with those audio amplifiers which employ significant amounts of negative feedback.
Zero and very low amounts of negative feedback amplifiers are essentially in the same class as the transducers.
It needs to be accepted that the non-linear nature of the auditory system has been clearly demonstrated, and as a result the existence of Acoustic Harmonic Masking is a proven phenomenon.
That is, the "ear" generates/expects a regularly decaying harmonic structure when presented with a sine wave, for example. Any deviation from this harmonic structure will be audible, and the sensitivity to this deviation increases as the harmonic order increases. The rate of increase in this sensitivity is unclear, but there is evidence that several percent of 2nd harmonic distortion is acceptable, but 0.01% of 9th harmonic which is not accompanied by an appropriate amount of lower order harmonics is both audible and unacceptable. Hence, as long as the harmonic structure approximates the ears "desired" sequence the output is considered to be musical. If it does not, then it is considered to be discordant. This deviation in itself can be used as a metric for amplifier/transducer quality.
It is in this way that non-feedback systems and transducers can be acceptable even if they have high levels of intrinsic distortion, whereas systems with low levels of low order harmonics but relatively high levels (compared to the masking threshold) of high order harmonics may not be acceptable.
This is particularly true at low volume levels, due to the masking action, so amplifiers that demonstrate unacceptably high levels of high order distortion (i.e. sufficiently uncorrected crossover distortion) that exceed the sensitivity of the ear at low levels will be acoustically poor.
This tends to be the case with many relatively high feedback factor class AB amplifiers, for example, and justifies the poor reputation of amps that have good THD numbers versus those with worse THD numbers which have much better reputations.
However, there is an exception to this "rule". If all of the amplifier generated harmonics are at a level which is substantially below the audibility threshold (at about -110dBc) then the ear does not perceive them as being an add-on to the self generated harmonics and the acoustic "acceptability" will be equivalent to the ideally decaying harmonic case.
Most negative feedback systems have frequency dependent gain functions and hence feedback factors that are of an inverse frequency kind- i.e. the degree of gain/feedback factor falls as the frequency increases.
In addition the action of the input stage non linearities and the output stage non linearities in conjunction with negative feedback creates high order harmonic distortion products that have reduced amounts of feedback available to correct them.
The familiar distortion/feedback factor curves that have been around since these issues were first explored generally lack two cogent aspects- they do not capture the frequency dependent function of the feedback factor, and they also do not include the reduced level of intrinsic open loop distortion that can be caused by the presence of very high open loop gain at the frequencies of interest. i.e. if the input signal is very small then the open loop distortion created by the amplifier may also be small- so amplifiers with very high open loop gains and low open loop distortion alter the curves in beneficial ways.
Bruno Pudseys pointed this out in his inimitable style in an internet exchange.
So, what becomes important in all of this is, when considering the audible quality (or at least one aspect of it) of an amplifier with negative feedback, that it is vitally important to consider the amount of negative feedback available at least at 20kHz and hence the ability of the amplifier to correct the generated harmonics at this frequency to render them inaudible. It is also important to have amplifiers with low intrinsic distortion with the differential input signals actually present at the input of the amplifier.
This goal can be assisted by having adequate gain bandwidth products for the gain stages with applied negative feedback and if possible tailoring the frequency response of these stages to keep the feedback factor adequately high over the desired frequency range of operation.
The amplifiers/gain partitioning/location of the pole zero constellation in the RIAA deemphasis sections were chosen with these issues in mind. The goal was to keep all of the harmonic distortion products substantially below the noise floor (and < -110dBc) over the entire operating range of the preamp.
I hope that this makes at least some sense. I also hope that I haven't made too many typos etc. as I don't have time to make revisions at this moment and I have no idea how long I have to edit this post.
Thank you Wyn,
That does make sense and any disagreements are at beer levels of discussion over vinyl noise floors, RIAA attenuation of that noise floor and gain balance. My personal interests these days are heading down the paths less trodden and mainly for fun, so I'm slowly evolving my MC system to having flat gain preamps and MM to have a current mode flat gain stage straight into an ADC then do RIAA correction in the digital domain. Partly because I can and partly because it annoys at least 2 tribes of vinyl lovers.
But it is always good to look at other's approaches to this ancient problem and see what inspiration pops out.
That does make sense and any disagreements are at beer levels of discussion over vinyl noise floors, RIAA attenuation of that noise floor and gain balance. My personal interests these days are heading down the paths less trodden and mainly for fun, so I'm slowly evolving my MC system to having flat gain preamps and MM to have a current mode flat gain stage straight into an ADC then do RIAA correction in the digital domain. Partly because I can and partly because it annoys at least 2 tribes of vinyl lovers.
But it is always good to look at other's approaches to this ancient problem and see what inspiration pops out.
So, newbie here interested in your design. As you note in the discussion, LOMC v. 2 has many options, and I've been looking to sort it out. BTW, I recommend your v.2 Build Notes to anyone wanting to get a clear picture of the different choices offered.
I'm curious about the op-amp choices -- i.e., one can use a single op-amp front-*‐end (with reduced parts count) per channel, or one dual op-amp (therefore parallel, in each channel) or two paralleled (single) op-amps (per channel). OK.
I assume the "stock" version uses the BOM-listed single-channel op-amps, with two in parallel for each channel, as shown in the diagrams posted.
I also assume the other variations I've just described are being tested by DIY'ers out there, so if one variation is a surprise winner, we'll find out.
So ... My basic question, I'm curious -- what is the intended (or measured) result of running the op-amps in parallel in this preamp design?
My apologies if this has already been addressed.
The intended (and measured) results of having dual opamps in parallel at the front end is to reduce the equivalent input noise.
When two low noise bipolar input opamps such as AD797s or LT1115s (or 1/2 OPA1612) are used in parallel the equivalent input voltage noise of the stage decreases by 3dB, but the current noise increases by 3dB. For LOMC cartridges with impedances of c. 10 ohms resistive, 10uH inductive the current noise is essentially irrelevant, so you get a c.3dB improvement in S/N ratio.
For the MM/FET input case the cartridge source impedance is much larger but the current noise is much, much smaller, so you get a 3dB improvement in the equivalent input noise when you double up. For example if OPA1656s are used the 2.9nv/rtHz noise spec at 1kHz of a single opamp would be reduced to c. 2.1nv/rtHz for the combination.
One minor benefit of having two identical stages in parallel is that the feedback components are also in parallel. This acts to reduce the standard deviation of the error of the 75us TC, which tightens the already exemplary deviation from the ideal RIAA characteristic even more.
All of the combinations have in fact been tried, with no surprises.
An add on board set has even been developed to allow for the retrofit of earlier single opamp designs to dual parallel opamps. This was not my doing. Some builders are quite enthusiastic about the improved noise and wanted the option so they "sponsored" the effort.
When two low noise bipolar input opamps such as AD797s or LT1115s (or 1/2 OPA1612) are used in parallel the equivalent input voltage noise of the stage decreases by 3dB, but the current noise increases by 3dB. For LOMC cartridges with impedances of c. 10 ohms resistive, 10uH inductive the current noise is essentially irrelevant, so you get a c.3dB improvement in S/N ratio.
For the MM/FET input case the cartridge source impedance is much larger but the current noise is much, much smaller, so you get a 3dB improvement in the equivalent input noise when you double up. For example if OPA1656s are used the 2.9nv/rtHz noise spec at 1kHz of a single opamp would be reduced to c. 2.1nv/rtHz for the combination.
One minor benefit of having two identical stages in parallel is that the feedback components are also in parallel. This acts to reduce the standard deviation of the error of the 75us TC, which tightens the already exemplary deviation from the ideal RIAA characteristic even more.
All of the combinations have in fact been tried, with no surprises.
An add on board set has even been developed to allow for the retrofit of earlier single opamp designs to dual parallel opamps. This was not my doing. Some builders are quite enthusiastic about the improved noise and wanted the option so they "sponsored" the effort.
Last edited:
Incidentally, by using the input stages of two boards in parallel or replacing the parallel device circuits with the dual input mini board add-ons, together with a minor capacitance change, a further 3 dB reduction in noise can be had for most LOMC cartridges, should one desire it. Reductions in noise beyond this point would be even more pointless and are not supported.
The noise improvements do not come with changes in distortion or PSR, but do come with a tiny reduction in overload margin due to increased IR drops in the stage decoupling networks.
The noise improvements do not come with changes in distortion or PSR, but do come with a tiny reduction in overload margin due to increased IR drops in the stage decoupling networks.
For those who just ordered boards.
The order for new units was placed, but the shipment to me will be delayed for several days due to the Chinese holidays.
As always I will provide tracking numbers when the boards are sent to their final destinations, which I anticipate being in about 2 weeks.
The order for new units was placed, but the shipment to me will be delayed for several days due to the Chinese holidays.
As always I will provide tracking numbers when the boards are sent to their final destinations, which I anticipate being in about 2 weeks.
Last edited:
I just completed the initial build of a v2.0 LOMC only, parallel LT1115, Canare StarQuad cables to Cardas RCAs. It fired up first time with no issues. The boards are not yet in a chassis.
Wyn is great to deal with in getting the PCB and very generous and helpful not only with build questions but also clear technical descriptions of design choices and tradeoffs.
I did a SQ comparison to my Elac Alchemy PPA-2, which is quite a fine phono preamp for $1,000. The rest of my system is a custom Lenco PTP turntable, Audiomods 6 tonearm, Paradox Pulse Guard R SMR cartridge loaded at 500 ohms, PS Audio BHK tube preamp and NHT xD 2.2 amps/speakers.
First impression after several hours of comparison listening is that it is in the ballpark with the PPA-2, which is quite an achievement. Peter Madnick has been doing great bang-for-the-buck products for a while. His VITB phono pre was Stereophile's budget component of the year way back in 1995 and his recent PPA-2 raises the bar at this price point.
I'd say the Wyn has similar but tighter bass with more precise leading edges. The PPA-2 has better timing and extension on the top end, noticable with more natural sounding acoustic bluegrass and space around the instruments. I bet the Wyn top end will improve once in a chassis with the loading tuned and some break in time.
Well done Wyn!
Wyn is great to deal with in getting the PCB and very generous and helpful not only with build questions but also clear technical descriptions of design choices and tradeoffs.
I did a SQ comparison to my Elac Alchemy PPA-2, which is quite a fine phono preamp for $1,000. The rest of my system is a custom Lenco PTP turntable, Audiomods 6 tonearm, Paradox Pulse Guard R SMR cartridge loaded at 500 ohms, PS Audio BHK tube preamp and NHT xD 2.2 amps/speakers.
First impression after several hours of comparison listening is that it is in the ballpark with the PPA-2, which is quite an achievement. Peter Madnick has been doing great bang-for-the-buck products for a while. His VITB phono pre was Stereophile's budget component of the year way back in 1995 and his recent PPA-2 raises the bar at this price point.
I'd say the Wyn has similar but tighter bass with more precise leading edges. The PPA-2 has better timing and extension on the top end, noticable with more natural sounding acoustic bluegrass and space around the instruments. I bet the Wyn top end will improve once in a chassis with the loading tuned and some break in time.
Well done Wyn!