[Previous Message]
[Next Message]
|
| Date: |
March 12, 2007 at 08:01:09 |
| From: |
m.x., [213.164.91.2] |
| Subject: |
Re: measuring trouble |
|
|
|
|
The TS measurement is really an impedance
measurement. If the amp voltage varies
with frequency (all amps have some
kind of high pass inside) this is no
problem. You simply measure amp voltage
and voltage across the speaker.
A much more serious problem is that most
voltmeters can not properly measure low
frequency AC signals!! The reading will start
to oscillate if you go much down below 50/60
Hz (they are aligned to mains frequency, because
a dual slope will then insensitive to mains
stray signals).
So I built an precision rectifier with very
low filtering frequency (about 0.3 Hz).
It takes longer time to settle, but 10 Hz
are still manageable with it. Still there
is some error, so an even lower frequency
would make sense.
Then I use a DC voltmeter.
The other thing was to build a small
1W class A/B amplifier (741 + two
BJT). It has no capacitors but
DC operation point is controlled with
an active autozero circuit, this
makes a lower corner of 0.1 Hz.
I think that most woofer measurements are
seriously flawed because of AC voltmeters or
sound cards which can not go low enough.
Note that a lower corner of 5 Hz means -3 dB
there, or roughly 30 % error.
It is a complete waist of time to have
than a 4 digit meter, most of the digits
are just fantasy.
I also noticed that it is good practice
to measure many frequency points.
The fitting of parameter will be much
better and you can see systematic
(see above) an random errors.
This way I noticed that some speakers
tend to vary their parameters a lot, depending
on previous mechanical stress (pressing
the cone) and also orientation towards
gravity.
m.c.
|
|
|
| View the previous message in this thread
Go to the top of this thread
View entire thread |
Posted with TalkShop version
2.76 BETA |
[Previous Message]
[Next Message]
|
|
|
Follow Ups: |
|
|
|
[
The Subwoofer DIY Page - Discussion Forum ] [ FAQ ] |
|
