John Curl's Blowtorch preamplifier part II

[PMA]

Pavel,

We are not arguing.

I know, Ed

I just intentionally wanted to remind these 16-20 years old Burr-Brown products. BTW, it was a shock for me when they became an acquisition of TI and their component list was suddenly restricted to some 15%. So it goes, only commercial products for masses survive - few cents S-D chips.

 

[1audio]

With that in mind I am starting to build a noise analyzer. I am aiming for .1 to 1,000,000 hz. range. I expect actually to only break that into octaves and use a peak detector. I might also consider some sort of slow decay detector. Due to the limited bandwidth I do not expect to consider RMS. The DC level would go to a log converter then into a multiplexer. The output would be a digital scope display.

I can use Biquad filters for all but the highest range. The version that starts with the LPF stage also adding gain is the preferred approach. I suspect I need to play with a biquad or LC for the top octave. I may need to use a discrete front end for the low frequency stages.

I haven't yet come up with a good decision on the input impedance. I expect to be able to get within 3 db of the noise floor.

ES

What is the intended use of the noise analyzer? At first glance it looks like a device noise analyzer. There are a lot of small details to getting useful output but once sorted they can be very useful. I would make an analog scaler/isolator interface to a sound card and use a PC myself. A simple heterodyne scheme could get the 1 MHz band if you really need it. Remember with noise you don't need ultra low THD. You do need headroom for crest factor. Are you planning to also provide for the standard weighting options?

 

[scott wurcer]

Why not indeed. I do happen to have a nice analog function generator in the lab and as it so happens one of the best prosumer recording/playback devices available (on measured performance anyway, which is what I use it for).

So I may do exactly that...


Ciao T

Yes, I doubt something like a Sound Devices 702 would produce results as bad.

 

[ThorstenL]

Scott,

Yes, I doubt something like a Sound Devices 702 would produce results as bad.

No idea on that unit, I will be using a EMU1616m, which uses AKM's more or less "best" converters, which means minimal actual levels in hardware (25 last time I looked) and only 128 OS at single speed (reduced at higher rates), so I think I'll do a 12KHz sawtooth at 192KHz sample rate, for fun...

Ciao T

 

[ThorstenL]

Hi,

Yes, I doubt something like a Sound Devices 702 would produce results as bad.

Looking at it, the converters seem rather "buget", going by the claimed SNR's etc. You may find yourself surprised. The converters seem to barely make 18Bit grade ENOB...

Sadly I have nothing of comparably (low) spec available to me...

Ciao T

 

[simon7000]

What is the intended use of the noise analyzer? At first glance it looks like a device noise analyzer. There are a lot of small details to getting useful output but once sorted they can be very useful. I would make an analog scaler/isolator interface to a sound card and use a PC myself. A simple heterodyne scheme could get the 1 MHz band if you really need it. Remember with noise you don't need ultra low THD. You do need headroom for crest factor. Are you planning to also provide for the standard weighting options?

I plan to inject noise into various stages from the power line to the loudspeaker connection and see where it shows up. Then how it affects sound quality. I suspect out of band noise degrades sound quality in a manner that can be quantified.

 

[RM]

Soundcardsoundscardsounds, doesn`t anybody use proper Audio Analyzers aymore?

Regards
Roar

 

[abraxalito]

With that in mind I am starting to build a noise analyzer.

Something like this one? I recall Glen talking about this project on the now defunct diy-audio-engineering.

http://users.picknowl.com.au/~glenk/LowNoiseAmplifierSchematic.pdf

 

[SY]

Soundcardsoundscardsounds, doesn`t anybody use proper Audio Analyzers aymore?

Regards
Roar

As backup. But soundcards work better and are much more versatile.

 

[simon7000]

Something like this one? I recall Glen talking about this project on the now defunct diy-audio-engineering.

http://users.picknowl.com.au/~glenk/LowNoiseAmplifierSchematic.pdf

No, I have enough amplification per se, it is dividing it into bands and looking at the peak and average levels.

 

[scott wurcer]

Hi,



Looking at it, the converters seem rather "buget", going by the claimed SNR's etc. You may find yourself surprised. The converters seem to barely make 18Bit grade ENOB...

Sadly I have nothing of comparably (low) spec available to me...

Ciao T

I was thinking portable not luggable (I assume this is a PCI unit), like a filmmaker might use in the field. I forgot you are a tough guy. No matter, just picking a highly respected unit off the top of my head.

Proceed, though I would expect your stock sound card is sufficiently low in spec.

 

[PMA]

But soundcards work better and are much more versatile.

They have issues, and one must be well aware of the issues.

 

[PMA]

so I think I'll do a 12KHz sawtooth at 192KHz sample rate, for fun...

Ciao T

The result will be horrible (regardless we forget corners/peaks). My example of 1kHz D/A - A/D triangle was with the AKMs, 48kHz/24bit mode.

I would like to mention that another great issues of the digital audio are necessary brickwall filters - time smearing and pre-ringing is unavoidable. All the natural transients are distorted, as most instruments work above 20kHz. It would not be the case if the digital audio sample rate was 500kHz - 1MHz, but 44.1kHz is simply unusable. I understand that the industry guys are not happy with my comments, but this is just the fact.

 

[SY]

They have issues, and one must be well aware of the issues.

IOW, like everything else on Earth.

I understand that the industry guys are not happy with my comments, but this is just the fact.

You'd think, then, that after 30 years of complaining, someone might have actually shown evidence of sonic degradation for 16/44.1... Absence of evidence is not evidence of absence, but it's certainly suggestive.

 

[simon7000]

You'd think, then, that after 30 years of complaining, someone might have actually shown evidence of sonic degradation for 16/44.1... Absence of evidence is not evidence of absence, but it's certainly suggestive.

SY,

I don't know a single pro who doesn't prefer higher and faster. The difference in a decent studio is about equal to the difference between incandescent lights and daylight. If you want to quote any test that shows 20/192 or better can't be distinguished from 16/44.1, it just ain't believable. There has to be other ignored issues such as jitter, monitor equipment, acoustics, etc.

ES

Also to my earlier quiz certainly you know what a #47 is?

 

[ThorstenL]

Pavel,

The result will be horrible (regardless we forget corners/peaks).

Well, if my claims are right, it will be.

If the orthodox position is right and there are no problems with low bit DS ADC's/DAC's all we will see is a lovely 12KHz sawtooth with the added effects of bandwidth limiting...

Ciao T

 

[ThorstenL]

Scott,

I was thinking portable not luggable (I assume this is a PCI unit), like a filmmaker might use in the field. I forgot you are a tough guy. No matter, just picking a highly respected unit off the top of my head.

Proceed, though I would expect your stock sound card is sufficiently low in spec.

The EMU 1616m is a fairly small external box sound system. PC-Card interfaces are available for use with laptops, so it is eminently portable. The converters are basically "120dB DR Class" with competent, if boring analogue circuitry.

I can even fit the 1616m and a pair of semi-decent (modified Behringer) mikes, XLR Cables, an external 1 TB hard drive AND my laptop into a laptop bag. I cannot fit the mike stands in there sadly.

It has quite a few outputs and inputs, more than I really need, but it allows some interesting possibilities.

I use it with a custom made breakout box to supplement the AP2 and other test gear, not normally for recording, the subjective sound quality is quite poor.

A full set of RMAA tests is here:

RightMark Audio Analyzer test: EMU 1616m

Note these are full analogue loop-back, that is DA looped back into the AD, the limit is actually the DA. It also includes most of the HF rolloff in the analogue stage design. My own unit(s) are modified to correct for the FR.

Ciao T

 

[Chris Hornbeck]

On another thread there is a fellow who seemed to improperly identify an IC. So for fun I thought I would list partial part numbers and see who can correctly identify the part. For example 741 would be a uA741 the first commercial op amp with built in compensation introduced by Fairchild.

47 (More than one answer!), 072, 83, 107, 300, 555, 703, 709, 722, 923, 7400

Any takers?

I'll play. 83 has Mercury in it and 107 has Germanium. 300 is a fetish item in the Pacific Rim. All the rest are modern - bah, humbug!

Thanks,
Chris

 

[ThorstenL]

Ed,

I don't know a single pro who doesn't prefer higher and faster.

But Ed, the tests are not double blind ABX and hence even with a 10,000 in 10,000 rate of return does not constitute evidence, just anecdotes...

TBH, exploited to the max CD can be quite competent as release media after careful mastering.

Also to my earlier quiz certainly you know what a #47 is?

A #47 to me is a directly heated tetrode of low power output.

I am sure since RCA used that number in the 1930's there have been many more electronic items marked "47", ruddy copycats.

Ciao T

 

[SY]

Ed,
But Ed, the tests are not double blind ABX

Or any other kind of double blind. You know, the stuff that's done routinely in any sort of DSP. There may be 10,000 anecdotes, but that's all they are- no more evidentiary value than reports of alien abductions and anal probing.

Now, before my meaning is wrenched out of context, there's a good reason to have a greater bit depth during recording (Ed has already started to do this), but as a playback medium, 16 bits is more than enough.