How accurate is RMAA distortion measurement?

[analog_sa]

I wonder if anyone has experience using RMAA to test amps. Last night i tested my latest IT PP creation and RMAA reported unbelievably low THD (around 0.02%) and absolutely disastrous IMD of around 5%.

I am using the loop-back mode and passing the amp output through a resitor attenuator to compensate for gain. Measurements were taken at around -2db which sounded moderately loud if connected to a speaker.

Similar measurement on a GC type amp yielded imd bellow the noise of my laptop soundcard.

Does my tube amp really suck?

It is very simple. A 5687 (Ea=180v, Ia=11mA) feeds a Lundahl LL1660 18mA connected as SE-PP. The output of the transformer feeds a pair of Sovtek 2A3 with cathode bias running at around 270v A-C and 58mA. And there is an input stage for some extra voltage gain.

This is my first attempt to use a transformer as a phase splitter and although my expectations were low the sound really surprised me with dynamics and speed - so much better than any active splitter i've tried. As a first iteration it's not perfect - it may sound a bit cluttered in complex passages and that's why i'm curious about IMD. Not to mention that the 1660 rings like a bell on square waves.

My longer-term (next week?)plan is to complete the front end following the Olson/Thorsten concept of splitting the phases from the input.

Any ideas re the dismal IMD score?

 

[Frank Berry]

Does the amplifier sound good to you?
If your answer is yes, forget the measurements.
Remember ..... It's all about sound.
Many times the measurements don't agree with what we hear.

 

[jan.didden]

..what is RMAA???

Jan Didden

 

[analog_sa]

Frank

I certainly agree. Still, there is this nagging suspicion that IMD of 5% cannot be good. Nor does it make any sense for an amp to develop a disproportionate amount of IMD compared to harmonic distortion. If the thd was 5% i would have no problem with that and would expect a certain amount of imd as well.

Most likely there is a problem with my measurement setup.

Does the amp sound good? Yes, it sounds spectacular in many respects but not quite 'relaxed'. I guess it's difficult to have 'speed/dynamics' and 'relaxed' at the same time And these square waves can give anyone nighmares.

 

[peranders]

..what is RMAA???

Jan Didden

RightMark Audio Analyzer

Russian freeware, excellent!

http://audio.rightmark.org/

 

[dimitri]

Jan Didden wrote ..what is RMAA???

RightMark Audio Analyzer
http://audio.rightmark.org/
the russian analog of Spectra(+)

 

[jan.didden]

Thanks P-A, Dimitri, I was unaware of RMAA

Jan Didden

 

[peranders]

Jan, if you don't have tested RMAA, do so, because it's a pleasant experience, easy to use , nice looking reports etc. Also the support is really good! And the most peculiar thing is that these russian dudes don't want any money for it! They have also a support forum.

 

[ChocoHolic]

Poor Sound Card???

WOW!!!
This soft ware seems to be exactly what I need!!!

BUT??!!! What's this? For first tests, I measured a simple cable
which should have acceptable low influence on the signal.
==> ...did this in order to check the performance of my sound card,
some older sound blaster....
The results do not differ much, no matter
if I choose 44kHz/16Bit or 192kHz/32Bit(float)...
They are much below what one could expect from 44kHz/16Bit
anyway!
The frequency response drops below -3db already slightly above
6kHz!!!!!!!!!!!!!!
Noise and dynamic around 85dB.
THD around 0.5 percent.
IMS around 0.3 percent.
Crosstalk around 82dB...

May be that's the reason, why I never felt lucky
with any music from this machine
...and always stated: For music you can kick that stuff!!!

1. Do you think my sound card can really be that poor?!
2. If yes, which sound card is fine?

Cheers
Markus

 

[ChocoHolic]

....will check this with my lap top soon....

 

[ChocoHolic]

Alright the Lap Top looks much better.

Frequency response is +/-0.5 db up to 16 kHz
and drops below -3db above 20kHz.
Noise and dynamic: about 84 dB
THD: below 0.01 %
IMD: around 0.02 %
Cross talk: around 80 dB

Altogether: May be helpfull for one or the other measurement.
Great! Thanks for the link to Right Mark!!!!!!

Bye
Markus

 

[arnoldc]

may i know how the connection from the amp to the pc input was made?

 

[TonyTecson]

amp out connects to the line in of your pc's sound card, you have to use an attenuator, like an l-pad, as the pc's soundcard inputs can only take line-level signals...

you will have to use the 1/8in baby plug....



analog_sa,
may we know what attenuator did you use to test your amp? a simple schematic of your test set-up would be nice...what frequencies did you use?

thanks,
tony

 

[arnoldc]

joan2, thanks for the confirmation of the attenuator. i know i have to have something between the speaker out and the PC but i don't know what it should be

analog_sa, may we know?

will the attenuator be a variable in the test?

at least i can now test my foreplay pre and see how bad, or good it really is

 

[peranders]

ChocoHolic, and others, Rightmark has listed many soundcards and their performance so you can compare.

http://audio.rightmark.org/results.shtml

 

[analog_sa]

i know i have to have something between the speaker out and the PC but i don't know what it should be

I just used a fixed attenuator so as to set the amp gain to unity. This will obviously lead to inferior S/N but should not affect the distortion tests.

 

[fmak]

ChocoHolic, and others, Rightmark has listed many soundcards and their performance so you can compare.

http://audio.rightmark.org/results.shtml

--------------------------------------------------------------------------------
Proves what I have found out measuring Creative cards - crap.

I am not sure of the absolute accuracy of the results as RMAA works perfectly on some cards/computers only. I have has 'infinitely flat' freq responses and no figures for some tests.

 

[fmak]

I wonder if anyone has experience using RMAA to test amps. Last night i tested my latest IT PP creation and RMAA reported unbelievably low THD (around 0.02%) and absolutely disastrous IMD of around 5%.
-------------------------------------------------------------------------------
The Rightmark seems to give much better numbers than my WinAudioMLS 64 bit FFT using ASIO input. I suspect that the software reports band limited numbers to 22kHz even on 24/96 setting. This is the 5.3 version.

 

[Gordon McGregor]

Re: Re: How accurate is RMAA distortion measurement?

I wonder if anyone has experience using RMAA to test amps. Last night i tested my latest IT PP creation and RMAA reported unbelievably low THD (around 0.02%) and absolutely disastrous IMD of around 5%.
-------------------------------------------------------------------------------
The Rightmark seems to give much better numbers than my WinAudioMLS 64 bit FFT using ASIO input. I suspect that the software reports band limited numbers to 22kHz even on 24/96 setting. This is the 5.3 version.

It depends of your sound card and measuring settings. You need to control the level of input signals as well, if your soundcard clips at some levels http://members.cox.net/alexhardware/images/imd.png , you may get unpredictable results.

I use RMAA for a long time since ver.4.3 (measuring amps and AV processors as well as sound cards) and have no problems with it at all.

 

[fmak]

Re: Re: Re: How accurate is RMAA distortion measurement?

I use RMAA for a long time since ver.4.3 (measuring amps and AV processors as well as sound cards) and have no problems with it at all. [/B][/QUOTE]
-----------------------------------------------------------------------
But have you compared it with an accurate measurement system. If you examine measuirements made on sound cards using iyt, you will find surprisingly good distortion and noise nos.

The software up to 5.0 was quite quirky.