distortion meter
Benny
Are there some who knows something about a homemade distortion meter?
Regards
Benny
Phil Allison
"Benny"
>
> Are there some who knows something about a homemade distortion meter?
** Yep.
Made my own ( using common ICs) some time back - 3 spot frequencies ( 66Hz,
1001 Hz and 5875 Hz) with 0.001% resolution.
Problem is, you also need a low THD audio generator with VERY low frequency
drift and fine frequency adjustment.
Had to make one of them too.
..... Phil
Patrick Turner
I designed and built my own distortion measurer.
There is a Wien Bridge oscillator using one opamp which produces a 1V
rms signal at 1kHz with THD = 0.05% approx.
The frequency can be varied very slightly above and below 1.0kHz.
The output of the oscillator is switched to adjust its level for 0dB,
-6, -12, and -18dB.
The switched output is then amplified by a discrete component solid
state voltage amp with a class A output stage and with a lot of NFB
but it also acts as a bandpass filter with a Q of about 12, so the THD
of the output is always below 0.01%.
The output levels are 8V, 4V, 2V and 1V, and lowest THD is on the 1V
output level and = 0.002%.
There is a 5k pot to adjust the level to be fed to a device to be
tested such as an audio amp input.
The signal from the device output is brought to to the input of an LCR
bridged T notch filter. The notch depth can be adjusted with broad
range pots and narrow range pots. So by means of adjusting both the
oscillator F and the notch filter depth a null of the 1kHz signal in
the signal is reduced over -80dB, leaving all the noise and harmonics
to be dispayed on a CRO and measured by a wide band millivoltmeter.
I have a high impedance input buffer for series connection in front of
the bridged T to allow testing of anode circuits of signal tubes
without loading down the tube under test.
I also have a switchable hum filter to remove everything below 300Hz
because it will often otherwise swamp what you are trying to see on a
CRO or meter.
There is also a 1.5kHz to 10kHz bandpass filter with gain of 10 to
amplifiy the recovered THD after the notch filter. This uses 2 oppamps
and a passive LCR circuit. It is very handy when THD signals fall
below 1mV.
I built the gear myself over a period of a month and gradually
eliminated all the faults and problems of noise and distortion
production one tends to get when one sets out to make a useful peice
of gear which I could not buy anywhere.
Its better than the old HP N&D meter which someone gave me after I
built my own. Mine has a remote PS, and all metal box, and very
careful location of coils in filters to prevent the case iron causing
3H in the signals being tested.
I did read a lot of old Wireless World magazine articles and some
other old books for some guidance.
For intermodulation distortion I made a special tunable bandpass
filter with Q = 50 and even level output between 1kHz and 11kHz, and
this allows analysis of recovered THD to be further analysed, or IMD
products detected when say an 80Hz signal is used with a 5kHz signal
and you can find the amount of amplitude modulation levels of the 5kHz
signals.
The filter has a remote PSU and uses 3 opamps with a few RC networks
used for combined positive and negative FB that nobody ordinary would
ever be able to work out for themselves. One pot is used to tune the
F.
I suggest that you study very hard like I did and spend many evenings
in your workshop to achieve what I did.
Or else you just download a Spectrolab program and do it all with an
old PC.
Patrick Turner.
Trevor Wilson
"Benny" <hojv...@gmail.com> wrote in message
news:433adc6c-6ac9-41b3...@21g2000yqj.googlegroups.com...
> Hello
>
> Are there some who knows something about a homemade distortion meter?
>
**As PA has intimated, it is best to build a spot frequency one and a very
stable oscillator. I built one back in 1974-ish, using state-of-the-art OP
amps (709). It was tunable from 10Hz to more than 20kHz, but it's resolution
is not that great, nor is it much fun to use. I still have it somewhere.
It's easy enough. You just need a high sensitivity millivoltmeter and a
tunable notch filter. Of course, nowadays, such things can be done with
sound cards and software for not too much money.
--
Trevor Wilson
www.rageaudio.com.au
Phil Allison
"Trevor Wilson"
>
> **As PA has intimated, it is best to build a spot frequency one and a very
> stable oscillator. I built one back in 1974-ish, using state-of-the-art OP
> amps (709).
** The uA709 was not SOTA in 1974 - not by a mile.
For making AF notch filters, the LM308H was THE one by far.
100+ dB CMRR and 1 nA input bias typical.
One is in my -100dB Wien bridge notch filter and out performs any substitute
I can find.
The RC4558 (dual) was out in '74 too- ate 741s for breakfast.
..... Phil
Bret L
lot of time and nothing better to do, I recommend just buying a good
audio measuring set off the shelf. The prices have come down big
time.
Neutrik has the little handhelds which are quite cheap.
Arny Krueger
>
> Are there some who knows something about a homemade
> distortion meter?
Back in the days before computer digital audio became mature and cheap, I
spent a ton of time building and modifying distortion analyzers and audio
signal generators.
I have modified IG18 and IG 5218 generators and modified a IM 5258 analyzer
to bottom out below 0.005% THD over the audio range. At 1 KHz that's
interesting, but doing the same at 20 or 50 Hz is something else again.
I just about died the first time I started mixing them up with a Turtle
Beach Fiji audio interface and various software packages. Compared to the
computer approach, all of my lovingly cultured analog gear was very
primitive stuff, relatively hard to use, relatively unstable, and
spectacularly uninformative.
Right now a good on-board audio interface and a copy of the Audio Rightmark
program will equal or better just about any analog gear that you can try to
build or buy for a a reasonble price. If you spend about $150 for one of
the better audio interfaces, then even the best stand-alone equipment on the
market today will be in your sights.
Probably the best construction articles about audio test equipment is a
series of articles by Robert Cordell that were published in Audio magazine.
Also check the diyaudio forum, where cordell posts.
Bret L
>
> I have modified IG18 and IG 5218 generators and modified a IM 5258 analyzer
> to bottom out below 0.005% THD over the audio range. At 1 KHz that's
> interesting, but doing the same at 20 or 50 Hz is something else again.
>
> I just about died the first time I started mixing them up with a Turtle
> Beach Fiji audio interface and various software packages. Compared to the
> computer approach, all of my lovingly cultured analog gear was very
> primitive stuff, relatively hard to use, relatively unstable, and
> spectacularly uninformative.
Yeah but your analog gear was hardly up to HP standards, was it?
>
> Right now a good on-board audio interface and a copy of the Audio Rightmark
> program will equal or better just about any analog gear that you can try to
> build or buy for a a reasonble price. If you spend about $150 for one of
> the better audio interfaces, then even the best stand-alone equipment on the
> market today will be in your sights.
A PC, IS NOT a piece of test equipment.
It is not benchworthy.
There are things like National Instruments' PXI that are but they are
still expensive as opposed to good standalone instruments.
Pete Millett has a good piece on using the HP 8903 (the box that's
cheap because it DIDN"T replace the HP 339) and on building a sound
card interface which when attached to a PC will give you some good
functionality. Interesting project but not a replacement for an AP
ATS-1 in my opinion.
Arny Krueger
news:478cbbb4-d98b-4b7f...@g4g2000pri.googlegroups.com
>> I have modified IG18 and IG 5218 generators and modified
>> a IM 5258 analyzer to bottom out below 0.005% THD over
>> the audio range. At 1 KHz that's interesting, but doing
>> the same at 20 or 50 Hz is something else again.
>>
>> I just about died the first time I started mixing them
>> up with a Turtle Beach Fiji audio interface and various
>> software packages. Compared to the computer approach,
>> all of my lovingly cultured analog gear was very
>> primitive stuff, relatively hard to use, relatively
>> unstable, and spectacularly uninformative.
>
> Yeah but your analog gear was hardly up to HP standards,
> was it?
Depends which HP. It was head and shoulders above 333 and 334, and didn't
miss the 339 by much.
>> Right now a good on-board audio interface and a copy of
>> the Audio Rightmark program will equal or better just
>> about any analog gear that you can try to build or buy
>> for a a reasonble price. If you spend about $150 for one
>> of the better audio interfaces, then even the best
>> stand-alone equipment on the market today will be in
>> your sights.
> A PC, IS NOT a piece of test equipment.
I'm not talking about a plain old PC. I'm talking about a PC with certain
add-ons.
> It is not benchworthy.
Tell that to all the people who use them as test equipment on the bench and
out in the field.
> There are things like National Instruments' PXI that are
> but they are still expensive as opposed to good
> standalone instruments.
Here's the point - there's a tremendous price performance advantage to using
PCs and professional-grade audio interfaces for signal analysis.
> Pete Millett has a good piece on using the HP 8903 (the
> box that's cheap because it DIDN"T replace the HP 339)
> and on building a sound card interface which when
> attached to a PC will give you some good functionality.
> Interesting project but not a replacement for an AP ATS-1
> in my opinion.
General replacements for AP equipment with PCs is not what I'm talking
about.
>> Probably the best construction articles about audio test
>> equipment is a series of articles by Robert Cordell that
>> were published in Audio magazine.
>> http://www.cordellaudio.com
>
>> Also check the diyaudio forum, where cordell posts.
I'm not sure that the Cordell equipment is that much more benchworthy than
approaches based on audio interfaces.