It is an intersting discussion.
But, the test lab as usually we do are standard and allow us to understand if the project is fine or not compared also with the other brand and project, I mena under technical aspect.
The "A" filter allow, when you are measuring with a very low level signal, to kill the ac spurios coming from the interconnections around the test set, first of all.
I normally measure also without filter and the differences are around 10-15 dB but if I put the AP in linear and starting to select the low filter from <10 Hz then 100 Hz then 400 Hz the level reach the reference at 1 kHz; this is easy to see because the AP read the major components of the residual and it is easy to see 50Hz or 100Hz or multiple.
And this is valid for each riaa stage, less or more.
Of course in a riaa the high filter selection on AP, always in linear but selecting the filters at 22kHz, 30kHz and 80 Khz, doesn't make any differences.
When I test the frequency response of a riaa in some case I need to do the test more than one time because, due the sporadic interferences, in the region of the bass range sometime the curve has some "spike". So I need to check carefully any connection also for ground path.
I normally use the SUT for MC stuff; in this case the s/n is measured with a reference signal of 0,5 mV then the input shorted ( zero ohm) and in this case the s/n remain the same, sometime little bit better.
Walter
But, the test lab as usually we do are standard and allow us to understand if the project is fine or not compared also with the other brand and project, I mena under technical aspect.
The "A" filter allow, when you are measuring with a very low level signal, to kill the ac spurios coming from the interconnections around the test set, first of all.
I normally measure also without filter and the differences are around 10-15 dB but if I put the AP in linear and starting to select the low filter from <10 Hz then 100 Hz then 400 Hz the level reach the reference at 1 kHz; this is easy to see because the AP read the major components of the residual and it is easy to see 50Hz or 100Hz or multiple.
And this is valid for each riaa stage, less or more.
Of course in a riaa the high filter selection on AP, always in linear but selecting the filters at 22kHz, 30kHz and 80 Khz, doesn't make any differences.
When I test the frequency response of a riaa in some case I need to do the test more than one time because, due the sporadic interferences, in the region of the bass range sometime the curve has some "spike". So I need to check carefully any connection also for ground path.
I normally use the SUT for MC stuff; in this case the s/n is measured with a reference signal of 0,5 mV then the input shorted ( zero ohm) and in this case the s/n remain the same, sometime little bit better.
Walter
A couple of l things strike me about this post:
1. You have an AP! You lucky devil. I wish I could afford one. Fortunately I can do almost as well with a good USB sound interface and REW. THis has an excellent real time analyser (RTA) which allow you to see exactly what is going on.
2. I regularly measure the noise from my tube microphone preamp designs and I do not experience spurri from the interconnections around the test set up. I do see spurii, but they are all accounted for by power supply noise or external interference due to inadequate screening. Both of these sources are part of the design and need therefore to be included in measurements. The gain range of these preamps is 26dB to 70dB and noise is measured with 150 ohms across the input.
Cheers
Ian
Hi Ian
I have personally two Ap, SYS1 and SYS Two with a USB adapter. Plus other stuff.
In Audioreview lab we have three AP , one AP SYS1, one SYS Two and one AP5xxx
I agree about REW and a good sound card, they help a lot and always i sugesst a minimal but good setup
Leaving the simulation!!!! 🙂
Walter
I have personally two Ap, SYS1 and SYS Two with a USB adapter. Plus other stuff.
In Audioreview lab we have three AP , one AP SYS1, one SYS Two and one AP5xxx
I agree about REW and a good sound card, they help a lot and always i sugesst a minimal but good setup
Leaving the simulation!!!! 🙂
Walter
I'm glad you use an A-weighting filter, because if you would optimize the amplifier without A- or ITU-R 468-weighting, you would end up in another optimum, one where noise at frequencies that you can barely hear is reduced and noise at frequencies that you can hear quite well is increased compared to your present optimum.
I think the standard test is not very useful for comparing your amplifier with other amplifiers unless you know that they have about the same noise current as yours (that is, are also tube or FET input designs with a standard input termination resistor), see post #19. Otherwise an amplifier that is almost 20 dB worse than yours under practical conditions may seem slightly better in the standard test. I do understand that you need good figures in a standard test if you want to sell products, no matter how useful or useless the standard test may be.