If you look at the Gain-Bandwidth Product of any amplifying device you will see that it has a linear roll-off with rising frequency. It is impossible for the "over-kill" low noise device to have a bandwidth approaching 150,000 Hz, and gain! However, as such phono stages are on offer, you may think that many designers are plainly ignorant. On the contrary, this is exactly where the fad for bandwidth limiting stems from.

As a musical instruments timbre (the way it sounds) is dependant on all its harmonics being present and ordered correctly in time. Bandwidth limiting therefore alters the sound. A fact that seems to be lost on some members of the Hi-Fi press.

At one time there were no one-box-does-all phono stages. Magnetic phonograph cartridges were generally all moving magnet types. Preamplifiers were inside amplifiers or "stereograms". Although moving coil existed in those days, they were rather bulky and gave a similar output to that of today's moving magnets. The moving coil fad started in earnest at a much later date.

Originally, to make the usual (low output) variety of moving coil cartridges work into the phono stage you had to use a step-up transformer. Not until recently had technology advanced sufficiently for semiconductors to replace step-up transformers.

It is now possible to make gain stages that are low noise and have the required real bandwidth to amplify the tiny outputs of moving coils, and thus dispense with the step-up transformer. But as yet, it is still impossible to optimise a circuit to perform both the functions of MM and MC amplification without sacrificing the performance of one or the other.

There is one more factor in the one-box compromise. It is mathematically impossible for the frequency equalisation to be identical for both the gain ratios required. Get it right for MM, and the bass will be a good decibel up for MC. You can get round this by making the EQ passive, placing it between "flat" amplifier stages, and some commercial designs do exactly that. But there's an even bigger drawback with this method called clipping of high frequencies. This is what is described as grain and by some extreme of imagination, detail! It is in fact distortion.

The one-box phono stage designer is in a no win situation, but such products are on sale, somehow get good reviews, and are very lucrative for their manufacturers. However, the compromises explained here will at least throw some light on why our no-compromise products score so highly in direct A-B comparison.

All our phono stages are fixed. Either as Moving Magnet or (soon to be introduced) Moving Coil stages. The only way to have both options is to have one of each, or to use a step-up transformer with your MM stage. Better still, the Elevator EXP electronic step-up transformer.

The layman may have difficulty understanding the comments below, and judging by their offerings, some audio designers would have difficulty following it too.

The most expensive type is in fact, a MC head amp and a MM phono stage in one box, with a switch to choose between the two. The switch, by necessity, is right in the path of the phono stage input, just where you don't need a switch contact! Also, the very sensitive MC amp has to share the same power supply rails with the less sensitive MM phono stage.

The other type uses the same gain stage for both MM and MC with switches to alter it's gain and impedance characteristics. It's impedance characteristics have to be optimized for the higher impedance MM cartridge (switches bring in smaller resistors to correctly load MCs), but then, the noise current is wrong for the smaller source impedance's of MCs and noise will become a problem ("you can't have your cake and eat it").

There are integrated circuits that do try to address these issues, but they are very slow (have narrow bandwidth) and sound that way. If you did design for MC noise gain you'd end up with a large bias current, which, although being safe for MCs, could fry a MM. The input offset would be such that the stage runs out of headroom with a MM leading to the typical hard sound, especially if it's a multistage passive EQ design (mentioning no names). There are bias current canceling techniques to minimize the problem of cartridge "fry-up", but to implement such a design discretely requires a prohibitive number of components, some of which have deleterious effects. Far better to use a chip solution, of which, just one exists. This chip's slew rate limits bandwidth (it sounds sluggish/boring).

Actually, there is one other chip that could do it, but it's difficult to implement - a bit of a "catch 22" case - because it is only stable above a particular noise gain, and in a conventional active RIAA stage would never be stable. A full book can be written explaining the virtues of different techniques. Until then, there's the no compromise two box solution: The Elevator and Era Gold/Jazz Club."