Nice, but only has a sweep from 800Hz up. Also somewhat confusing as to what a 'constant velocity amplitude' is. They also state 'linear cut' which suggests unequalised whilst saying the test record is designed to be used with RIAA equalisation. Then to complete throw things they say its only good to +/1.5dB, which is a bit of a barn door if you are trying to measure system response.
Real life also.
I have seen a circuit like this powered by photo diodes to get a floating supply which was claimed removed the mains as a noise source. It ios a NPN/PNP common base input of 100R. I am sure most have seen it. I suspect BC337/327 could be used.
Moving Coil Cartridge Head Amps
Does it include a worked out example of incorporating 12dB/8ve subsonic filtering in a single OPA RIAAA stage as I suggested in
#266 ?
This would be novel and USEFUL so suitable for inclusion in your new book.
It certainly does, but it's not novel since Holman suggested an implementation.
What is novel is two versions of 3rd-order subsonic filtering (18dB/oct) in one RIAA stage.
Likewise there is also a 4th-order (24dB/oct) version but that needs another opamp
How about including an 8th order filter?
http://www.diyaudio.com/forums/analogue-source/300043-diy-18hz-hpf-vinyl-de-rumbler-filter.html
Got a link? Is it a fully worked out example?
I grovel at your feet O Guru Self 😱
While we are on the subject of stuff which will make things sound better as opposed to a numbers race ...Pyramid said:
How about variable damping on the single OPA 2nd or 3rd order subsonic filters?
Holman shows HUGE variation below 100Hz from different arm/cartridge combinations and variable damping of the 12dB or 18dB/8ve subsonic filter can make things much nicer.
Pyramid, nicely worked out 8th order filter .. but I think you need to recommend 1% or 2% caps for accuracy ... though the variation in arm/cartridge resonances will swamp this 🙂
Would Douglas attempt a digital solution? A friend who works in Pro Audio ( who probably made the majority of mixing desks of the 1990 we listen to ) thinks it is the answer to many problems. This would be for RIAA and subsonics. John's thinking is he turns up to do a job where nothing is known and must get the sound right 10 minutes later. This means getting the main problems fixed and then tweak it. He said the big surprise is how good it sounds and it is almost making the impossible happen. John plays with active crossover points and phase shift. Usually 4th order is used, to his surprise first order can be had.
If you ask yourself how come 10 minutes? This is often because people shop around for the best price and don't give it a second thought that they are asking for bad sound. On top of that they won't book him again if he gets it wrong.
At the CES Al Stewart did a mini concert for us ( About 50 of us ). He said he had almost cancelled because the sound was so bad and it had taken all day. The sound was excellent and so close to the LP's.
If you ask yourself how come 10 minutes? This is often because people shop around for the best price and don't give it a second thought that they are asking for bad sound. On top of that they won't book him again if he gets it wrong.
At the CES Al Stewart did a mini concert for us ( About 50 of us ). He said he had almost cancelled because the sound was so bad and it had taken all day. The sound was excellent and so close to the LP's.
http://www.diyaudio.com/forums/analog-line-level/122432-analogue-vs-digital-riaa.html#post1500611
This was good reading and showed one or two problems and neat solutions.
This was good reading and showed one or two problems and neat solutions.
We have done better recently in several threads. 1) Audacity is 16 bits on recording (on a PC) no matter what the software reports, 2) Audacity's FIR filters are linear phase only, the RIAA is not. 3) The difference in crest factor of real music pre/post equalization is FAR less than 20dB.
Scott. Your ideas were so easy to understand. Thank you. I think it gives an interesting insight into how digital works into the bargain.
Not a link, but its from his 'New Factors in Phonograph Preamplifier Design' paper. No component values are given, so I worked them out & verified it with sim.
Seems to me you are suggesting using a subsonic filter as a bass equaliser. Different functions altogether, surely?.
No. I decided to stop at 6th-order for all-pole filters, because of component-sensitivity issues. You get Butterworth, Linear-phase, and Bessel.
For faster rolloff than that, I go up to 6th-order elliptical filters.
Would you consider including some information on cartridge "tuning"? Lots of cartridges have a fall off in the high frequency response followed by a stylus resonance peak. The Handbook for Sound Engineers had several graphs by the people at Stanton Magnetics showing the effect of various terminations on cartridge frequency response. These were actual measurements using test records. With the proper termination you can really smooth out the frequency response. That was what I liked about the Shure V15VMR. Very smooth low frequency response and stylus resonance at 35kHz, well above the audio range.
I was able to very easily locate a PDF of "New factors in phonograph preamplifier design" by Tomlinson Holman 🙂 Something i wasn't aware of before. The 18dbOct HPF is indeed novel ! I simmed it with an Opamp, but without the RIAA components for -3db @ 15Hz. However, i couldn't reach the full -db figures quoted below 15Hz ?
Apart from the clever HPF, it seemed to have decent specs for the time too.
@ kgrlee et al
I used duckduckgo to search for it, & it was listed close to the top results 😉
Apart from the clever HPF, it seemed to have decent specs for the time too.
@ kgrlee et al
I used duckduckgo to search for it, & it was listed close to the top results 😉
Thanks for this Zero D.
I now remember seeing this in Jurassic times and feeling rather icky cos the 2nd order stage has bits between cartridge & input device.
My (probably totally irrational) prejudice is to run my EVIL input base currents through the cartridge by direct connection. 😱
But this isn't what I posted in #266
What I tried (in #266) IIRC, I first saw from J. Linsley-Hood. IM(totally prejudiced)O, this was a better implementation .. though only 12dB/8ve.
Of course the JLH filter is MUCH more difficult to calculate for a textbook response than Holman's 18dB/8ve
Doug, I hope you have at least a worked out example or preferably the full design process for the JLH ... the Holman design process (at least for Butterworth) is trivial and available in practically any textbook on filters.