The upcoming page* showcasing the USB turntables samples I've been posting put a renewed focus on ceramic cartridges. This caused me to update the case against ceramic carts in the turntable guide.

I added a new section on RIAA equalization and would appreciate a review for technically accuracy. In particular, I would like you to pick apart this declaration:



Am I over simplifying? Are there capacitors or other components down the line that also affect the equalization of the signal or is it truly all in the cart itself?

As always, I appreciate your insights.

-Jeff

* Sneak preview. This is the only link on the Internet leading to that page.

Ceramic cartrdiges are usually specified to work best with a specfic loading, which in some cases may include non-trivial RC networks.

The easy part of making a cartridge that has a mechanical RIAA deemphasis curve seems to be the overall -6 dB/octave slope. The turnover or flat spot between roughly 500 and 2100 Hz is more difficult to create mechanically.

Magnetic cartridges are constant-velocity devices: the signal is proportional to the velocity of the stylus. Since high frequencies cause the stylus to "wiggle" faster, you naturally get an increasing velocity (and hence signal level) as frequency rises. The cutting head used to make the master is also a constant-velocity device, so a magnetic cartridge "matches" the characteristics of the cutting head. The RIAA EQ is applied during cutting in order to restrict excessive groove excursion (bass cut) and to overcome surface noise (treble boost). When you play a record that was cut with a constant-velocity cutter using a constant-velocity cartridge, you therefore need to apply inverse EQ to correct the frequency response.

In contrast, ceramic cartridges are constant-amplitude devices. In other words, the signal level is proportional to the amplitude of the side-to-side motion of the stylus. There is no natural tendency for the signal level to increase at higher frequencies (because the amplitude doesn't get bigger), and hence RIAA EQ should not be applied. Of course, the frequency response characteristic you get out of an unequalised ceramic is only approximately OK-ish. But since its intrinsic sound quality is so low that's an academic point.

Thanks for the excellent responses.

I get that each cartridge has different characteristics and that, in a magnetic cartridge, the higher the frequency, the greater the signal.

What I'm still confused about is this: My understanding of applying equalization to an audio signal means raising or lowering the amplitude over a range of frequencies. Doesn't that mean something needs to be done on the "constant-amplitude" ceramic cartridge to render the signal flat(ish) again, as Arnold seems to be indicating?

Actually Clive and I agree. The ceramic cartridge inherently rolls of the treble with a -6 dB/octave roll-off because it responds to the amplitude of the music in the groove, not the velocity.

We also both agree that amplitude response gives you an accurate signal for about 80 percent of the frequencies from 20-20 KHz. Problem is, the RIAA curve is not a simple slope, but it has this shelf between 500 and 2150 Hz. nCeramic cartridges doin't have that shelf in their response curves. Unfortunately, this is right in the midrange where the ear has great sensitivity to such errors. So ceramic cartriges congenitally sound a little off in the midrange unless somehow some electrical circuit puts the shelf back in.

I'm still pretty fuzzy on the difference between constant-amplitude and constant-velocity (I even went as far as asking the EE forum on PhysicsForum.com to define and contrast the terms- no takers yet).

My fuzziness aside, does this mean my declaration stands up, particularly if we replace my uber-Californian adverb "organically" with "mechanically?"

Without resorting to calculus... think of the difference between a ruler and a speedometer. Ceramic carts are rulers, magnetic carts are speedometers.

A consequence of the math involved is that the "rulers" are less sensitive to high frequencies than the "speedometers". Ergo, 6db/octave drop.

Consider an LP groove with a frequency of X. In order for the stylus to trace this groove for a certain time period, it needs to travel a distance proportional to X. Now double the frequency to 2X. The groove wiggles back and forth twice as many times in a given period of time, and therefore the stylus must necessarily travel twice as far. Given that it must do so in the same period of time, it therfore follows that its speed is doubled. And if its output is proportional to velocity (a magnetic cartridge), the signal level also doubles. (Twice the signal level = 6.02dB, and twice the frequency = 1 octave, hence Arny's statements about 6dB/octave). But if its output is proportional to amplitude (a ceramic cartridge), the signal level depends only on the size of the sideways deflection.


The difference between constant-velocity and constant-amplitude is down to simple geometry. I suppose the analysis of geometrical changes with respect to time is the area of mathematics we call "mechanics", so you could use that term. But it's important not to confuse this term with what a lot of people think of when you say "mechanical" (like how machines work).

The reason why a magnetic cartridge is constant-velocity is due to the laws of electromagnetics. If you move a magnet next to a coil, you induce a current that is proportional to the speed at which the magnet moves (all other things such as strength of magnet, turns on the coil, etc. being equal, of course).

A ceramic cartridge is a constant-amplitude device because the nature of the piezo-electric effect is such that the generated voltage is proportional to the pressure applied.

Think about a sine wave plotted out the usual way. Think of the dot that plots the sine wave. As time progresses, the dot can be thought of as progressing across the page from left to right. The distance of the dot from the left of the plot is time, or if you will, the amplitude of the time.

The amplitude of signal represented by the dot can be thought of a number of ways, but by convention, we usually think of the amplitude of the dot as being the vertical distance from the hoizontal center line of the plot to wherever the dot is. Positive amplitude when the dot is above the center line, negative amplitude when the dot is below the center line.

As it goes across the page the dot has two velocities. One is the horizontal speed of the dot or its velocity in the time dimension from left to right, which we usually think of as being constant for the entire plot.

However, the dot also has a velocity going up and down the page which varies quite a bit. On the sloped parts of the wave the dot's vertical velocity is greatest, and on the flats on the top and bottom of the sine wave, the dot's vertical velocity is actually zero.

If the sine wave we plot represented the wiggles in a LP groove, a ceramic cartridge responds to the height of the dot, and a magnetic cartridge responds to the vertical speed of the dot.

Now, lets think about the plot of a sine wave whose amplitude remains the same but the frequency is rising as the dot goes across the page.

The waves are going to get more scrunched together on the right side of the plot, compared to the left side of the plot. Because the waves are scrunched on the right side of the plot, the sloped portions are going to be steeper. IOW, the vertical velocity of the dot will peak out at higher and higher velocites as the frequency rises.

So, if frequency is rising, and amplitude remains the same, then the (vertical) velocity increases. OTOH, if you wish to keep the vertical velocity the same, then you have do cut the amplitude as the dot goes across the page.

A constant amplitude wave would be a sine wave with the same height as the frequency rises. A constant velocity wave would be one whose amplitude shrinks as the frequency increases.



Hmm, you said: "You might say that ceramic cartridges equalize the audio signal organically: They rely on their chemical composition to implement the RIAA curve."

I would reword that to say:

"You might say that ceramic cartridges equalize the audio signal mechanically: They rely on the way they convert the groove into a signal to implement an approximation of the RIAA curve."

I'd put "Don't use a ceramic cartridge - they're crap in many ways."


Wave your hand backwards and forwards slowly.
Now wave your hand backwards and forwards quickly.


A ceramic cartridge responds to how far you moved your hand* - the same both times.
A magnetic cartridge responds to how quickly you moved your hand - much more the second time.

So fast movements (high frequencies) need reducing in the output from a magnetic cartridge, but not from a crystal cartridge.

Cheers,
David.

* = as in, how far is it from the left-most point your hand got to, to the right-most point your hand got to.

Thanks everyone helping me make sense of this. I think I have a paragraph that will work without having to get into too much detail.

The link goes back to this thread in case my readers want to learn more.

I also rewrote the next paragraph in the section based on what I've learned here.





It's hard to miss that sentiment in the USB turntable guide. It's everywhere. In particular, any turntable with a ceramic cartridge has a warning label that links to the "Avoid Ceramic Cartridges" section.

The section talks about poor frequency response and high VTF. What was missing was a discussion on how poorly ceramics perform RIAA equalization.

Thanks to everyone who participated in this thread, the guide offers one more reason to stay from these things.

I would say that the primary reason why ceramic cartridges don't need preamps is that throughout they put out a ton more signal voltage than magnetic cartridges. The eq issue is actually secondary if you hook both kinds of cartrdiges to an directly to the input of an amp and don't use a preamp.

Whoops! I meant to say "doesn't require a preamp for equalization." Thanks for catching that.

I covered the difference in output voltage earlier in the ceramic cartridge section.

Here it is again with the fix:

Works for me.

Knowzy, your smarts. energy and dilligence really impresses me and I wish you the best!

Thanks. That means a lot.

And please call me Jeff. There may come a time when Knowzy (the magazine) and Jeff (the person) are not one and the same.