Welcome Licorice Pizza (LP) lovers! Setup guides and Vinyl FAQ.
Given the diversity of opinions regarding setting the cartridge azimuth â€“ leveling the cartridge, equal output, channel separation - it makes sense to look at how they affect the performance.
This may be a long and somewhat tedious reading, but hopefully some findings will prove useful.
I presume general knowledge on part of reader regarding how the phono cartridge operates, and how the two separate signals are generated.
We shall consider several cases, all of them representative, with their particular implications. Your cartridge may indeed possess the characteristics of more than one case. Be also aware that this discussion concerns only the simplified model for just one parameter. But hopefully this discussion will simply expose the various interactions.
The importance of channel separation for stereo reproduction has been established (or at least letâ€™s presume so for this brief discussion). In that light usually the phono cartridge represents the weakest link in a typical system. It does, therefore, make sense to try to obtain the best possible performance in that category.
Case A A Perfect Cartridge.
We shall consider a representation of a cartridge whereby it looks like a perfect square positioned so its diagonal is vertical, with its sides at 45 degrees to the surface of the record. A perfect cartridge â€“ in addition to its perfect 90 degree angles â€“ also has equal outputs on both channels.
Letâ€™s presume for this and further discussion that we have a moving magnet cartridge (moving coil will work the same way) and that the full scale output is a perfect 1mV.
As we know, the signal intended for one channel should appear in one channel only, the magnet crossing that channelâ€™s magnetic flux lines at 90 degrees. As, at the same time, it will be moving along the flux lines of the other channel, it should induce 0V output in that channel.
The magnitude of signal induced in the proper channel will be 1mV times sin(a) where a is the angle between the magnet motion and the channel flux lines. Therefore for a=0 degrees you get 0V output, for 90 degrees â€“ 1mV.
Any error in that angle â€œaâ€ will cause the max output to drop from 1mV, and in addition some output appear on the wrong channel. Just for fun now, look at what happens when we have a small, 1 degree error. The max output will suffer little, it will still be .999mV, but the channel separation instead of a former perfect infinity will now be just â€“35dB. Such is power of Sin function.
But letâ€™s go back to our PERFECT cartridge. One can now easily see that by rotating the cartridge away from the perfect azimuth, we will not change the channel balance, as both channels will drop according to the same angular error sin(a). So attempting to set azimuth using EQUAL OUTPUT will result in no difference in reading over broad range of angles. 10 degree error will be invisible â€“ it will produce the same channel balance as 0 degree.
Channel separation will suffer greatly, however. With just 2 degrees of angular error we will reduce our channel separation to â€“29dB, and 5 degree error will drop it to a rather mediocre â€“21dB.
In essence, while the equal output function has ZERO sensitivity to azimuth adjustment, the channel separation is VERY sensitive, and we kill it before realizing it.
One can then conclude that in the case of perfect cartridge it is completely improper to use the equal output test.
Case B. Cartridge with Angular Error.
Now consider a cartridge that has slight manufacturing inaccuracy in the angle of its internal structure. Letâ€™s presume that the structure is no longer a square, but rather a diamond with its angles equal to 85 and 95 degrees.
It can be shown easily that the best achievable channel separation of such cartridge will be just â€“27dB, provided the diagonal is vertical. It is also obvious that this is where the two output will be equal as well.
But what happens when we tilt the cartridge just 2.5 degrees to one side because of our mounting error? The outputs of one channel that used to be just 1uV below their max of 1mV (the result of the initial 2.5 degree internal error), will now drop by 4mV. The other one will go up to a perfect 1mV. I would state that the difference between 1.000mV and .996mV is still VERY small, hard to measure. But the effect on the channel separation is far greater â€“ it now drops to just 21dB in one direction, while increasing to a perfect near infinity on the other.
As we see here, the channel separation is a much stronger function of the angular error, allowing us to spot the optimal setting far easier. While the max output changes by just .4% or so, the separation drops by about 100% - VERY easy to read.
One observation in that case: if we indeed set the cartridge for equal outputs, then we DO get the optimal azimuth as well. And vise versa.
Case C. Cartridge with Angular Error and Channel Imbalance.
Now letâ€™s move our cartridge that is even closer to reality by introducing some small output difference. Letâ€™s say, that one channel has 1mV, while the other â€“ 1.005. That cartridge still has 5 degrees angular error for maximum achievable channel separation of â€“27dB.
We now attempt to set the azimuth by balancing the channels for equal outputs. In order to achieve that we need to tilt the cartridge so its output drops on the stronger channel while increasing on the weak one. We can easily see that we obtain the max signal on the weak channel by making the magnet cross its flux lines at a perfect 90 degrees â€“ by rotating the cartridge by 2.5 degrees. The other channel then is at 5 degrees error, and its output is now reduced to .999 â€“ and we achieve an almost perfect â€œequal outputâ€.
We do, however, now significantly reduce the channel separation â€“ it is now down to just â€“21dB.
In other words, any output difference in the two channels will force us to sacrifice the channel separation in order to achieve equal output condition.
Case D. Phono Stage Gain Error
Let us now presume that we are setting the azimuth by trying to achieve equal signals at the OUTPUT of the phono stage. Letâ€™s further presume that that phono stage has a small gain error between the two channels â€“ something as little as .25dB or so.
It can be shown that by doing the equal output adjustment in this case we will aggravate the situation by including the phono stage error in our procedure. In order to compensate for it we will have to reduce the channel separation even further, by moving ever farther away from the point of lowest channel cross talk. It will be easier to understand by bringing in our â€œPerfectâ€ cartridge that really needs no rotation, but now WILL be rotated to â€œpre-dropâ€ one channel by .25dB.
One can continue building such cases by including other error sources in all possible combinations. However, it should be clear by now that the following conclusions can be reached:
1. Only in the case of a cartridge with equal voltage outputs but some built-in angular error will the points of equal output and best channel separation coincide.
2. The closer your cartridge is to a â€œperfectâ€ one, the harder it will be to set the azimuth correctly by â€œequal outputâ€ procedure.
3. Setting a cartridge with inherent channel imbalance for equal output is guaranteed to destroy the channel separation.
4. Any difference in channel output should be dealt with in the preamplifier. Such adjustments have no effect on channel separation.
Therefore, in terms of how to set the azimuth, we shall state the hierarchy of methods:
1.. The preferred method is by achieving the best channel separation.
2. In absence of proper tools for that method, a level cartridge is the second best.
3. Setting azimuth by achieving equal outputs is at best inaccurate and will only work with some cartridges. It is guaranteed to give bad results with most cartridges, however.
I tried to keep this to some reasonable length by not doing much math here â€“ all those calculations are rather simple and can be done quickly with your calculator â€“ once you visualize the basic phono cartridge geometry. You might want to repeat them because they are fun and will give you nice insight into what is really happening.
One big point here is that if you indeed accept the notion that channel separation is important for stereo reproduction, then there is only one proper way of setting the cartridge â€“ and that is by directly monitoring this parameter.
You may of course decide that the separation is NOT the most important issue and try other means. In that case this will at least let you understand better how things interact.
This post is made possible by the generous support of people like you and our sponsors:
Topic - Funny things happen on the way to azimuth adjustment - Victor Khomenko 08:40:29 07/09/01 (15)
- Where does anti-skating fit in this equation? - Lew 14:59:36 07/10/01 (1)
- Re: Where does anti-skating fit in this equation? - Victor Khomenko 15:11:22 07/10/01 (0)
- This is one for the archive. (nt) - Rob Thomas 02:20:57 07/10/01 (0)
- Both interesting and enlightening - Gannet 10:23:55 07/09/01 (1)
- Re: Both interesting and enlightening - cfraser 02:53:49 07/10/01 (0)
- Re: Funny things happen on the way to azimuth adjustment - jk 09:46:41 07/09/01 (1)
- Re: Funny things happen on the way to azimuth adjustment - Victor Khomenko 10:02:30 07/09/01 (0)
- Re: Funny things happen on the way to azimuth adjustment - Mr Vinyl 09:19:21 07/09/01 (7)
- Re: Funny things happen on the way to azimuth adjustment - Victor Khomenko 10:06:55 07/09/01 (2)
- Re: Funny things happen on the way to azimuth adjustment - Mr Vinyl 11:04:18 07/09/01 (1)
- Re: Funny things happen on the way to azimuth adjustment - Victor Khomenko 12:50:56 07/09/01 (0)
- HFNRR anit skate - jk 09:36:15 07/09/01 (3)
You can not post to an archived thread.