Yes. If your scope has an A-B setting, you can measure large amounts of distortion. It takes an attenuator and a fiddly hand to null out.
For a better measurement, use a precision instrumentation amp with high CMRR at the frequncies of interest to compare input to (scaled-down) output. Then you only need a single channel.
For a better measurement, use a precision instrumentation amp with high CMRR at the frequncies of interest to compare input to (scaled-down) output. Then you only need a single channel.
If by A-B, you mean a dual timebase, yes it has one.
It also has little calibration pots for the vertical sensibility
I was trying to play with the CH2 invert, in ADD mode, and adjunsting the sensibilities to cancel the signal, and see if sometimes there is a difference
but it's not easy to adjust the signals to cancel them
Is there a better method? And only with a dual scope, no other lab tools
It also has little calibration pots for the vertical sensibility
I was trying to play with the CH2 invert, in ADD mode, and adjunsting the sensibilities to cancel the signal, and see if sometimes there is a difference
but it's not easy to adjust the signals to cancel them
Is there a better method? And only with a dual scope, no other lab tools
In Nelson Pass' article on the Citation 12 (in his archive section I believe), he describes how to use the amp to "measure itself" which is a slick system. I haven't tried it and havent thought too hard about it (it's a bit of a self licking ice cream cone) but it's something to check out.
Sheldon
Sheldon
No, I meant a setting that would subtract B from A, while letting the voltage scale be different on the two channels. A + B with an invert is just as good. Try scaling down the output externally, with a voltage divider and a 10-turn pot for trim; that will make the null a little less twitchy. But only a little less.
Sy,
I gather you propose to oput the input signal on one channel, to put an attenuated version of the output on the other, then subtract and null for min level.
This will indeed let you large amounts of THD only. But the phase shift through the amp will also limit the minimum you can get, unless you also delay the input signal and equivalent amount. It gets messy quickly.
I know that this method has once be proposed by Linsley Hood it was (or maybe Peter baxandall), but he used a series circuit of two adjustable phase shifters for the input network. He then subtracted the scaled down output of one channel of his amp (because the amp was inverting, he just connected the two channels together IIRC). The result (ideally, the THD only) was fed into the other channel, allowing him to listen to the distortion. Slick, but not for the faint of heart.
Jan Didden
I gather you propose to oput the input signal on one channel, to put an attenuated version of the output on the other, then subtract and null for min level.
This will indeed let you large amounts of THD only. But the phase shift through the amp will also limit the minimum you can get, unless you also delay the input signal and equivalent amount. It gets messy quickly.
I know that this method has once be proposed by Linsley Hood it was (or maybe Peter baxandall), but he used a series circuit of two adjustable phase shifters for the input network. He then subtracted the scaled down output of one channel of his amp (because the amp was inverting, he just connected the two channels together IIRC). The result (ideally, the THD only) was fed into the other channel, allowing him to listen to the distortion. Slick, but not for the faint of heart.
Jan Didden
I once applied the subtractive method proposed by Baxandall and by QUAD to an amplifier having a second-order Butterworth-like roll-off below 1Hz and above 135kHz. With Tracy Chapman's "Matters of the heart" CD as a test signal, I could get about 60dB of suppression of the music, using an adjustable first-order correction circuit for the low-frequency phase shift and an adjustable first-order correction circuit for high-frequency phase shift...
When you measure with a single mid-band sine wave, an adjustable attenuator with one adjustable phase shifter should be enough. Without adjustable phase shifter, a few degrees of phase difference can obscure any distortion below 1% or so.
When you measure with a single mid-band sine wave, an adjustable attenuator with one adjustable phase shifter should be enough. Without adjustable phase shifter, a few degrees of phase difference can obscure any distortion below 1% or so.
yeah, found it...
...I'll have to take a look. I've thought about using an instrumentation amp (as mentioned) to compare L and R. I would make a tweak to one side and see how it changes. I wouldn't know which side is actually better, however, I guess that whatever tweak (to both L and R) made the two channels most alike (and with least visible distortion) would be the best.
Thanks for the Nelson Pass reference.
http://www.passdiy.com/pdf/citation.pdf
JF
stokessd said:
...I'll have to take a look. I've thought about using an instrumentation amp (as mentioned) to compare L and R. I would make a tweak to one side and see how it changes. I wouldn't know which side is actually better, however, I guess that whatever tweak (to both L and R) made the two channels most alike (and with least visible distortion) would be the best.
Thanks for the Nelson Pass reference.
http://www.passdiy.com/pdf/citation.pdf
JF
Rod Elliot had a "SIM" project... which, in a nutshell, gives you the difference between the input and the output of an amp; that's it, distortion, noise, artifacts, etc. That would be interesting to examine with a scope. It's on his website.
MarcelvdG said:
Marcel,
IIRC, the object of that was to listen to the distortion, the claim being that in a competently designed amp, at realistic listening levels at the normal distance to the speaker, the distortion would be inaudible. What was your experience here?
BTW, Excellent article in EW on the RIAA/low noise stuff. Keep it coming!
Jan Didden
analog_sa said:
Guido Tent said:
I'm a student, on a limited budget (having recently bought a scope has limited the budget even more)
And I've got no access to lab tools since I'm still on hollidays
But, if there are some good distortion analysers that can be found for cheap on ebay, why not
Jan, when I tried the subtractive method with first-order filters and "Matters of the heart" as a test signal, I ended up with a residue that sounded like very soft filtered music, rather than distortion. I got about 60dB of suppression of the music, but apparently, the residue of the suppressed music was still much stronger than the distortion products. I haven't tried to get a better phase and amplitude match by using higher order filters.
Therefore, the only conclusion I could draw was that the distortion products with "Matters of the heart" as a test signal were substantially smaller than -60dB. When the amplifier was momentarily driven into clipping, I could immediately hear a loud click. I had to use fairly high signal levels to get an audible residue.
Therefore, the only conclusion I could draw was that the distortion products with "Matters of the heart" as a test signal were substantially smaller than -60dB. When the amplifier was momentarily driven into clipping, I could immediately hear a loud click. I had to use fairly high signal levels to get an audible residue.
Bricolo said:
I'm a student, on a limited budget (having recently bought a scope has limited the budget even more)
And I've got no access to lab tools since I'm still on hollidays
But, if there are some good distortion analysers that can be found for cheap on ebay, why not
Hi
I suggest you use your CDP and a test disc as a low distortion generator, and build a double T filter as a notch, giving you the residual signal. This in turn can be analysed using your scope, or eventuall an FFT using PC and soundcard.
This is how I measured when I was a student
regards
Guido
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.