Does component quality matter?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
In one of the articles in Electronics and Wireless World there was a preamp design -- basic, which nonetheless claims great figures and zero phase shift

The author constructs two prototypes -- one with off the shelf components and another with all those exotic components: audio-grade caps and gold-plated contacts and what not. Music was played to a professional group of listeners with the source shifting between the two prototypes imperceptibly. Most listerners did not notice any change in the quality of the music.

Now, I am about to construct an active crossover filter, a project that appeared in Elektor. Would it greatly matter if I used 10% polyester caps? While I could buy 1% MFRs where I live, 7.5mm pitch boxed polycarbonate caps were just unavailable in small quantities.

Please respond
 
AX tech editor
Joined 2002
Paid Member
component quality

You address two different issues here:

- component quality has to do with the materials used and construction of, say, a foil capacitor. Depending on your personal preference you can find authorative sources saying it doesn't make a difference, as in the article you quoted, or you can find authoritive sources claiming the opposite. You are on your own here, really.

- accuracy of values like 1% or 10%. In the xover you cited, the accuracy of the components determines the accuracy of the exact xover frequency and the accuracy with which the signal shifts from one section to the other. Generally one wants 5% or 2% values so that the design is nominal. That said, the drivers in your speakers have specs that are generally much more varying than 2% or 5%. Variations in Fs, inductance and Vas of more than 10% are quite common. Unless you measure the performance of the total system and correct as necessary, you might as well go with the 5% or even the 10% values.

Cheers, Jan Didden
 
Componet Quality

janneman,

Thanks. There, indeed, are two issues. What I wanted to know, specifically, is:

1. How perceptible/audible would the shift in the crossover point be if I were to use 10% polyester caps (green caps)?

2. Would you say that the audio quality would degrade?

Please share your experience.

Thanks.
 
AX tech editor
Joined 2002
Paid Member
component quality

dhar said:

How perceptible/audible would the shift in the crossover point be if I were to use 10% polyester caps (green caps)?

Let me try to clarify. Because of the uncertainty of the actual speaker driver characteristics, which will also shift with time, on which you would base calculations for the xover freq etc, a cap with a +10% tol might give better results than one that was right-on with 0% tolerance. My advice would be to built the xover with reasonable quality 5% or 10% caps, then do listening tests and play with the cap values to see if a larger or smaller value sounds better to you. Even if you use 1% caps, still do the tests and playing around. Not knowing your system, I have no idea how audible changes would be.

Cheers, Jan Didden
 
Componet Quality and Tolerance

janneman, thanks for your assuring reply.

I have TDA7294 amps, home made, and as for speakers, they are real el cheapos. Vifa and the likes are just unavailable where I live. I use 8" Philips (made in India) which resonate at 50Hz.

Anyway, "playing around" was just reassuring.

thanks again.
 
Does Component Quality Matter

Vivek,

excited to know that Peerless is available in India (must be very expensive?). Please could you let me know the retail outlet in Bangalore/any other place from where you buy them? Also, is there an outlet (SJP Road? I have tried all of them, unsuccessfully) from where I could buy 7.5mm pitch boxed capacitors for my active cross over project?

Thanks.
 
For a xover, go for the most precise you can have/afford. But there is no need to go to 0.005%. If talking about passive filters, 10-5% is OK, for low voltage 5-1% if possible.

All things acoustical ultimately depend on the velocity of sound in air. Take a long look at that in function of temperature/humidity/pressure. You will feel assured that 10-5% is OK.
 
r_s_dhar said:

Now, I am about to construct an active crossover filter, a project that appeared in Elektor. Would it greatly matter if I used 10% polyester caps? While I could buy 1% MFRs where I live, 7.5mm pitch boxed polycarbonate caps were just unavailable in small quantities.

The audibility of a difference in crossover frequency is greatly a function of the order of the filters you are trying to construct. The higher the order, the more likely a difference would become audible. High order filters are not possible to make without very accurate component values- if you build them with looser tolerance parts, you may find that you are building an entirely different filter that you thought.

Components that bear the same labeled value actually vary from part to part with a Gaussian distribution. A bunch of 10% parts will have the same average value as a bunch of 1% parts. So if you have a meter and a bunch of parts, you can select 10% tolerance parts that are actually within 1% of the desired value.
Most people have neither an accurate and precise meter nor a bunch of parts to sort.

Besides the frequency of the crossover being somewhat different than you were shooting for, you also have to contend with channel to channel matching. Even if the absolute frequency may not be particularly audible, a difference between crossover frequencies in the two stereo channels may be. Mismatches in the crossover frequency will cause unusual phase shifts which will be heard as the "position" of a sound source wandering back and forth depending upon its frequency. This will be clearly audible with test tones, but may not be with more complex music signals.

You can simulate the circuit and see what moving the capacitors around will do. Unfortunately, unless your simulator can do Monte Carlo analysis, this is a slow process. You need to randomly vary the capacitor and resistor values by whatever their tolerances allow to see what will happen. Be sure to look at the phase response as well as amplitude response. If you start to see phase shifts on the order of 20 degrees or more, you might expect some audible problems. The 20 degree number is based on listening to tone sweeps generated in SwCADIII and varying the phase in the two channels to see where the shift becomes audible.

In the end, it's best to use tighter tolerance parts, but if they aren't available, then screw it. Use what you can get now, and put in the tighter parts later, when and if they become available.

MR
 
Re: Componet Quality

r_s_dhar said:
janneman,

Thanks. There, indeed, are two issues. What I wanted to know, specifically, is:

1. How perceptible/audible would the shift in the crossover point be if I were to use 10% polyester caps (green caps)?

2. Would you say that the audio quality would degrade?

Please share your experience.

Thanks.

Usually, DIYer like to know that they used high quality components even if they can not tell the difference :) It is all about fun and a little of money.
 
AX tech editor
Joined 2002
Paid Member
Component quality

Francois,

OK, I accept that. I also know that sometimes it's about insecurity, and getting security by letting someone else take the decisions for you.
My philosophy is either to make sure you know how it works, or else find out how it works. In the end, I find it more satisfactory to know the ins and outs of what I'm doing. But maybe that's just MY insecurity....
Y M M V

Cheers, Jan Didden
 
Does Component Quality Matter

I was just wondering....

polyester caps are supposed to be good allrounders. But polystyrenes are supposed to be best for audio purposes. Now, if you did not mind their size, why dont we use them in active xover?

I am one of those with scattered and ecclectic reading in electronics without matching hands on experience. Hence questions of this kind. Can anyone please tell me?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.