Looks really nice. It scares me a little that there is no heat flow thru the interior to outside, though.
Monitor the interior temp. esp electro's.
Bet it sounds wonderful
THx-RNMarsh
Last edited:
I use a dual op-amp solution in one case and it is flat. The example sim I showed earlier is using just a single resistor around the RIAA EQ network and gain setting resistor - so its a single resistor solution.
As I mentioned earlier, if you think 2.5 dB peaking between 100 Hz and 20 Hz is bad, then look at cart responses - that's the really nasty surprise.
Group delay - see the Holman paper (with apologies to the unconverted) for the audibility and the various standards. GD in the plot I showed earrlier is 4ms at 20 Hz and ~0 at 60 Hz vs ~0 at 40 Hz without the filter - i.e. its a non-issue.
If you are really worried about it, don't use any subsonic frequency shaping and enjoy the cone flap
Last edited:
I make space for the younger
My favorite ways are t-bones and rib eyes. They sound amazing on the grill, deep bass, sweet mids, with amazing details and sound stage.
I understand the point and therefore asked billshurv about the evidence needed to believe that a described event really had happened and personally I trust that people honestly report what they have done, but take it not for granted if specific informations are missing.
Of course, but usually wrt measured numbers it is possible to replicate if the details are published. What was measured, which method was used and which records were used for measurement.
Compare it to the cymbal crash from Holman´s article; i read first about this quite extended frequency range in one of Bob Cordell´s articles; Cordell cited Holman´s article as the source, but Holman mentioned another audio engineer (iirc David Hadaway) as the (final) source for the measured numbers and with some other informations (now surely more difficult to get then in those years) you could find the record to repeat the measurement.
Could be - seems to be another example where asking the author about an event could help - but it might have been that he just used the graphic from Shure.
According to the various articles about the so-called "Trackability" (by Kogen, Anderson and others) Shure asked experts about difficult to track records, analyzed these and found the reported numbers. The most extreme a ~100cm/s (considered to be unreasonable) example and others at still quite high 70cm/s . But with two exceptions none of the records examined were listed/mentioned in any of these articles.
On Shure´s website it was claimed that the "Tracking study" was presented in 1973 but I wasn´t able to find it.
The two "hot" records mentioned were:
-) Shure has measured 104 cm/sec at 7.25 kHz -- slightly distorted -- on side one, band two of Verve V-8558 (Woody Herman recording; the muted trumpet sound produces the high velocity)
-) A recent Sheffield record -- "Pressure Cooker" -- has an 11 kHz sibilence recorded without distortion at 40 cm/sec on the beginning of side one
Then just switch the sub-sonic filter out
I have a Manhattan Transfer LP my wife bought me 2nd hand that is in pristine condition (no pops or crackles) despite the fact that it was pressed in 1977 or 78 except for the fact that it has a serious warp. I can play it and it sounds perfect - with the subsonic filter switched in. Without, I have very serious cone excursions.
(all kudos to the arm and cart for tracking it)
I find his contributions far more interesting than much of what gets posted here.
Besides he fits right in, his latest audio power amplifier uses discontinued transistors!
Jakob, there's lots of papers written over the decades where it might be difficult to find all the references - that's just life. BuI see no reason to dismiss his findings anymore than I would dismiss Scott's digital RIAA findings.
Holman references 23 sources in his article and I have assumed he was as honest as his ability allowed him to be. If he subsequently discovered he was in error, we will never know, but somehow I think he was in the ball park.
(NB wiki says he became an employee of Apple in 2011)
Holman references 23 sources in his article and I have assumed he was as honest as his ability allowed him to be. If he subsequently discovered he was in error, we will never know, but somehow I think he was in the ball park.
(NB wiki says he became an employee of Apple in 2011)
I use a dual op-amp solution in one case and it is flat. The example sim I showed earlier is using just a single resistor around the RIAA EQ network and gain setting resistor - so its a single resistor solution.
As I mentioned earlier, if you thing 2.5 dB peaking between 100 Hz and 20 Hz is bad, then look at cart responses - that's the really nasty surprise.
Group delay - see the Holman paper (with apologies to the unconverted) for the audibility and the various standards. GD in the plot I showed earrlier is 4ms at 20 Hz and ~0 at 60 Hz vs ~0 at 40 Hz without the filter - i.e. its a non-issue.
If you are really worried about it, don't use any subsonic frequency shaping and enjoy the cone flap
You don't seem to know what you are talking about, just throwing out numbers without any sense won't cut it. Please show the peaking subsonic filter you are using, the way you determined the GD, there is a tight relationship between the filter type and order and the GD.
The correct solution is to keep the RIAA correction tight and add a high order high pass filter tuned to 12-14Hz (this value covers the majority of tonearm+cartridge combinations, but it can go down to 10Hz or 16Hz in extreme cases, like in high or low compliance tonearm+cartridge combinations). The error of a 15Hz 3rd order Sallen-Key high pass filter is under 0.3dB @20Hz.
Your beloved Holman says http://www.kallhovde.com/advent/phono-pre-research.pdf that the GD should be lower than 20mS, of course without any serious justification, other than "A listening test with worst case choices of test signal and listening conditions has shown that the 20 ms worst-case group delay of a three-pole complex filter is just perceptible". Anyway, a 3rd order filter as above will fit this criteria.
Your argument "the cartridge is not flat" is laughable; based on the garden variety of cartridge frequency responses, +/-3dB RIAA response would be acceptable. Based on the same logic, don't build amplifiers with better than 1% distortions, and better than +/-3dB 20-20k frequency response, the speakers are going to swamp the amplifier anyway.
Me? 
I don't have any audible rumble on mine (direct drive), neither, but , who knows where and for what a circuit will be used ?
(it was just to tease-you a little, Bonsai ;-)
I told you exactly what I did to get the earlier plot. I ran a 20k resistor from the output of an all active RIAA phono EQ amp to the junction of the bottom of the gain setting resistor (75 Ohms) and the top of the DC blocking cap (470uF). That's all - nothing else. See the LT1115 phono amp schematic Mark Johnson posted if you want more detail on the technique.
re the GD, LTspice is your friend. Click on the right vertical axis and select 'display Group Delay' - that's how I got the numbers I mentioned re the above. Additionaly, See fig 7 in the Holman paper where hearing sensitivity to GD is shown.
If you are fixated on the +2.5dB peaking between 100 Hz and 20 Hz on a phono EQ and not worried about things like speaker-room interaction or cart response then you indeed have a problem of priorities. But if its such a concern, then just don't do it or switch the filter out.
And no, I don't believe cart response anomalies are an excuse for sloppy design - the EQ amp is flat to within 0.2 dB before I connect the 20k resistor as described above.
Last edited:
+1
Dadod made an original design, very inspiring, with impressive results, shared his work for free with all the schematics etc., very helpfull, in the perfect good practice of DIY, and now, he is flamed for talking about ?
Someone is walking on his head.
Dadod, congratulations. And thank you for sharing.
- Status
- Not open for further replies.