John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
Jump to Latest
#87,562
jneutron said:
The signal which is the difference is a zero integral signal. I suspect it is not viewable when the measurement tool is one that calculates signal power. I discussed this issue a bit with a rather good measurement physicist.

My intent was to put together a pair of IA's to subtract tweeter signals of a bi and mono set. Scott sent me chips I intended to use in my build/test, however the last two years have not been conducive to any hands on work.

John
Click to expand...

What are you using now for biwiring that I/others might duplicate and do the test for you?

I have a couple good scope plug-in which might do the job; a TEK 7A22 differential amps and 7A13 diif comparitor.

??


THx-RNMarsh
 
#87,565
In attempts to use SIM and build a passive x-over, the same issue applies as with semiconductor SIMulated circuits. If the parts are not characterized well, you get bad results no matter what the SIM indicates.

An air core inductor, for example, is hard to get 90 degree ideal phase from it over a wide freq range or at least 2 octaves either side of the x-over freq.

Here is some data on a 0.5mH air core inductor pulled from a loudspeaker system. Very typical of all air core values sold for x-over use which I have measured. ---> at 100Hz, 1KHz, and 10KHz. L+Q on top istrument and graph of its phase vs freq below.


DSC02429.JPG


DSC02427.JPG


DSC02430.JPG


8.8 Ohm load.


THx-RNMarsh
 
Last edited:
#87,571
bear said:
Having just bought a DAC that permits the loading of three distinct FIR filters, I can say that IF a FIR filter is used, they impart a clear and distinctive sound. Period.
Click to expand...
Hmm. What would that be? I've built and modified a number of DACs where I could change the filter. I did hear some differences, but pretty subtle. Nothing I would rate as clear and distinctive. Maybe your chip has a broader range of control?
 
#87,572
davada said:
Unless I've completely misunderstood what I've read aren't the sigma delta dacs a current output from a precision current source and sink? So if the power supply gets through then they are not much of either.
Click to expand...

Not necessarily. Both Cirrus and AKM have voltage out DAC's using cap filter techniques that push the state of the art. The latest is 127 dB SNR THD+N of -115 dB. Theoretically you need a current out DAC to get more performance but at those levels its hard to see why you would make the effort.

You do need well filtered reference supplies to get the performance.
 
#87,574
1audio said:
Not necessarily. Both Cirrus and AKM have voltage out DAC's using cap filter techniques that push the state of the art. The latest is 127 dB SNR THD+N of -115 dB. Theoretically you need a current out DAC to get more performance but at those levels its hard to see why you would make the effort.

You do need well filtered reference supplies to get the performance.
Click to expand...

I thought the voltage output dacs simply have the integrator built into them rather than external. Maybe you have inside information into the architecture of these dacs which in not shown in the data sheets.
 
#87,575
RNMarsh said:
In attempts to use SIM and build a passive x-over, the same issue applies as with semiconductor SIMulated circuits. If the parts are not characterized well, you get bad results no matter what the SIM indicates.
Click to expand...

Yes, and I would think that the major hangup would be correct modelling of a speaker. As far as I know there are no sim models for speakers that come remotely close to the accuracy of the C and L and even active device models. In that sense, sims of passive xovers are pretty useless except for a very, very rough proof of concept.

Jan
 
#87,576
Markw4 said:
Quality and expense are relative things. If you want to tri-amp DSP crossovers using DAC-2s, it costs about $1k/output. So if you need 6 DAC outputs that would add up to about $6k. Cheap for some, expensive for others.

Also, if you want to do DSP crossovers, you can't avoid digital filters. Some sound different than others for various reasons, and not are all equally accurate, if accuracy if the goal.
Click to expand...

This is DIY audio, so I was referring to DAC chips. You don't need to pay $1000... Nowhere near, if you DIY it.

Digital filters sound like you design them too. Of course filters do have a "sound" - they are designed to alter the sound by.. filtering. You can design them for accuracy - or pretty much whatever you want. But the sound is imparted by the system design, nothing else.
 
#87,578
gpauk said:
This is DIY audio, so I was referring to DAC chips. You don't need to pay $1000... Nowhere near, if you DIY it.

Digital filters sound like you design them too. Of course filters do have a "sound" - they are designed to alter the sound by.. filtering. You can design them for accuracy - or pretty much whatever you want. But the sound is imparted by the system design, nothing else.
Click to expand...

I understand your points, but I don't think a DIYer could design and develop a top quality state of the art DAC for only the cost of a bag of components. As a practical matter, one would need a well equipped lab with state of the art test equipment. By the time all is said and done, the DIY cost of design and development and testing something equivalent to a DAC-2 would probably be much more than just going out and buying one.

If I am wrong and some DIYer can design a DAC of that caliber for cheap, then by all means publish it here and everyone will have good cause to be very happy.
 
Status
Not open for further replies.
Top Bottom