Hi! I'm 'still' working with a turntable -> PC + RIAA Filter project I started ~year ago. Filter parts are OK (thanks to R. Orban) but, now (as not good in electrical engineering), I'm wondering if there is a DSP solution (if needed) for to match the cartridge specs w/ sound card input specs. As for an example, a cartridge which is recommended to be connected through 47kΩ /< 200pF path is connected into a soundcard with specs: Type: E-MU XTC™ combo mic preamplifier and Hi-Z/line input w/ Soft Limiter A/D converter: AK5385A Gain Range: +60dB Frequency Response (min gain, 20Hz-20kHz): +0.0/-0.16dB Stereo Crosstalk (1kHz min gain, -1dBFS): < -110dB Hi-Z Line Input: - Input Impedance: 1Mohm - Max Level: +12dBV (14.2dBu) - Dynamic Range (A-weighted, 1kHz, min gain): 113dB - Signal-to-Noise Ratio (A-weighted, min gain): 113dB - THD+N (1kHz at -1dBFS, min gain): -101dB (.0009%) .. does this non-matching connection have effect in frequency response or is it just some voltage level matter in question? If there are some effects in frequency response then Q1: Can the effect been calculated somehow and if it can be then, which values are needed to be known and which formulas? Q2: Could the effect be canceled by using some FIR implementation? Juha
Do I need to fix the non-matching analog connection in A/D process?
Started by ●June 7, 2008
Reply by ●June 7, 20082008-06-07
On Sat, 07 Jun 2008 02:57:12 -0500, "jtp_1960" <jtp_1960@hotmail.com> wrote:>As for >an example, a cartridge which is recommended to be connected through 47kΩ >/< 200pF path is connected into a soundcard with specs:...>- Input Impedance: 1Mohm...>.. does this non-matching connection have effect in frequency responseAbsolutely. The whole thing is an RLC circuit.>Can the effect been calculated somehow and if it can be then, which values >are needed to be known and which formulas?You could figure it out if you knew the impedance of the cartridge, but it would be much easier and more effective just to connect a parallel 49.9K Ohm (that's a standard 1% value) resistor at the input to the sound card. Unfortunately the input capacitance of the sound card is not specified, but you might experiment with parallel capacitance to find what works best (which might be "none at all").>Could the effect be canceled by using some FIR implementation?Yes, but why not do it the easy way, by loading the cartridge properly? Greg
Reply by ●June 7, 20082008-06-07
Thank you, I'll try that parallel resistor 'trick'.>>Can the effect been calculated somehow and if it can be then, whichvalues>>are needed to be known and which formulas?>You could figure it out if you knew the impedance of the cartridge, > >>Could the effect be canceled by using some FIR implementation? > >Yes, but why not do it the easy way, by loading the cartridge >properly? > >Greg >Sorry, since I forgot to mention that the FIR filter would be done as software implementation (a VST plug-in) and also, my target would be a plug-in which anyone could be able to use in their system if those needed 'parameters' from their hardware are known (cartridge recommendations / soundcard input specs) ... something like: user adjusts/enters those parameters on plug-in interface for calculation for FIR coefficients and the filter would then be processed real time ... so the filter wouldn't be 'hard coded' for certain system but could be used w/ any system which does not use matched connection between cartridge and soundcard (this is because of software based RIAA EQ filter I mentioned, which expects a flat connection instead of (matched/)pre-amped RIAA stage). So, how to calculate the effect in frequency response? I've seen various calculators for DSP design on many sites but, never seen one I could use for this task. Juha
Reply by ●June 7, 20082008-06-07
Greg Berchin wrote:> On Sat, 07 Jun 2008 02:57:12 -0500, "jtp_1960" <jtp_1960@hotmail.com> > wrote: > >> As for >> an example, a cartridge which is recommended to be connected through 47k٠>> /< 200pF path is connected into a soundcard with specs: > > ... > >> - Input Impedance: 1Mohm > > ... > >> .. does this non-matching connection have effect in frequency response > > Absolutely. The whole thing is an RLC circuit. > >> Can the effect been calculated somehow and if it can be then, which values >> are needed to be known and which formulas? > > You could figure it out if you knew the impedance of the cartridge, > but it would be much easier and more effective just to connect a > parallel 49.9K Ohm (that's a standard 1% value) resistor at the input > to the sound card. > > Unfortunately the input capacitance of the sound card is not > specified, but you might experiment with parallel capacitance to find > what works best (which might be "none at all"). > >> Could the effect be canceled by using some FIR implementation? > > Yes, but why not do it the easy way, by loading the cartridge > properly?Are you sure about that yes? The R-L filter's phase is included in the RIAA spec. BTW: why not the standard 47K? Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●June 7, 20082008-06-07
jtp_1960 wrote:> I'm 'still' working with a turntable -> PC + RIAA Filter project I started > ~year ago. Filter parts are OK (thanks to R. Orban) but, now (as not good > in electrical engineering), I'm wondering if there is a DSP solution (if > needed) for to match the cartridge specs w/ sound card input specs. As for > an example, a cartridge which is recommended to be connected through 47kΩ > /< 200pF path is connected into a soundcard with specs:(snip) It used to be easy to find phono preamps to use magnetic cartridges with line level inputs. Used ones might still be easy to find. Otherwise, you run into S/N, but if that isn't a problem it should work. -- glen
Reply by ●June 7, 20082008-06-07
On Sat, 07 Jun 2008 09:08:57 -0500, "jtp_1960" <jtp_1960@hotmail.com> wrote:> ... so the filter wouldn't be >'hard coded' for certain system but could be used w/ any system which does >not use matched connection between cartridge and soundcardI can only express my opinion, and my opinion is that this is a bad idea. The cartridge is part of a resonant system, and trying to "cancel" the effects of a mismatch in such a system is far more difficult than just providing it with the correct load in the first place. It's somewhat like trying to correct for poor room acoustics with DSP -- it's a lot easier and much more effective to treat the room first, in order to make it as good as possible, and then do the fine-tuning in DSP. Greg
Reply by ●June 7, 20082008-06-07
On Sat, 07 Jun 2008 15:50:50 -0400, Jerry Avins <jya@ieee.org> wrote:>> Yes, but why not do it the easy way, by loading the cartridge >> properly? > >Are you sure about that yes? The R-L filter's phase is included in the >RIAA spec. BTW: why not the standard 47K?He's already abandoned the phase issue by using Robert Orban's RIAA equivalent filters, which match magnitude but not phase. See http://groups.google.com/group/comp.dsp/browse_thread/thread/59232b93211c041a The cartridge wants to see a 47K resistive load. The sound card provides 1M. 1M || 49.3K = 47K, 49.9K is the nearest 1% standard value. Greg
Reply by ●June 7, 20082008-06-07
Greg Berchin wrote:> On Sat, 07 Jun 2008 15:50:50 -0400, Jerry Avins <jya@ieee.org> wrote: > >>> Yes, but why not do it the easy way, by loading the cartridge >>> properly? >> Are you sure about that yes? The R-L filter's phase is included in the >> RIAA spec. BTW: why not the standard 47K? > > He's already abandoned the phase issue by using Robert Orban's RIAA > equivalent filters, which match magnitude but not phase. See > http://groups.google.com/group/comp.dsp/browse_thread/thread/59232b93211c041aThat is not a good idea if he's a critical listener.> The cartridge wants to see a 47K resistive load. The sound card > provides 1M. 1M || 49.3K = 47K, 49.9K is the nearest 1% standard > value.Got it. I'm not used to thinking in terms of 1%. (The cartridges impedances certainly aren't that accurate.) Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●June 7, 20082008-06-07
glen herrmannsfeldt wrote:> jtp_1960 wrote: > >> I'm 'still' working with a turntable -> PC + RIAA Filter project I >> started >> ~year ago. Filter parts are OK (thanks to R. Orban) but, now (as not good >> in electrical engineering), I'm wondering if there is a DSP solution (if >> needed) for to match the cartridge specs w/ sound card input specs. As >> for >> an example, a cartridge which is recommended to be connected through >> 47k٠>> /< 200pF path is connected into a soundcard with specs: > > (snip) > > It used to be easy to find phono preamps to use magnetic > cartridges with line level inputs. Used ones might still be > easy to find. Otherwise, you run into S/N, but if that isn't > a problem it should work.Used to be easy to buy and still easy to build. I'll post a transistor circuit if he doesn't want to use op-amps. Jerry P.S. A run from cartridge to amplifier of even three feet needs special low-capacitance shielded cable. The easy way is putting the pre-amp on the underside of the turntable board or using a follower there. Then you use a 47K resistor at the follower and ignore impedances thereafter. -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●June 7, 20082008-06-07
glen herrmannsfeldt wrote:> jtp_1960 wrote: > >> I'm 'still' working with a turntable -> PC + RIAA Filter project I >> started >> ~year ago. Filter parts are OK (thanks to R. Orban) but, now (as not good >> in electrical engineering), I'm wondering if there is a DSP solution (if >> needed) for to match the cartridge specs w/ sound card input specs. As >> for >> an example, a cartridge which is recommended to be connected through >> 47k٠>> /< 200pF path is connected into a soundcard with specs: > > (snip) > > It used to be easy to find phono preamps to use magnetic > cartridges with line level inputs. Used ones might still be > easy to find. Otherwise, you run into S/N, but if that isn't > a problem it should work.Used to be easy to buy and still easy to build. I'll post a transistor circuit if he doesn't want to use op-amps. Jerry P.S. A run from cartridge to amplifier of even three feet needs special low-capacitance shielded cable. The easy way is putting the pre-amp on the underside of the turntable board or using a follower there. Then you use a 47K resistor at the follower and ignore impedances thereafter. -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
