Phono II and III 3.18 time constant resistor value.

[bob orbell]

Following on from my post 7, I have a second Phono III populated board with stock components including the 14K3 resistor at R21-22, using the same valves and power supply, the scope trace was almost identical until just after 5K cycles, then there is a rise to just under 4db all the way to 15K then another rise to 20K, no wonder it sounded too bright. I also expected more interest on this subject and I am very disappointed in the lack of positive communication, only Matthew expressed interest and I do hope he changes future sales of Phono R21-22 to the correct value, as for the bypass capacitors, why did WAD remove them ? I think I know the answer and 5751 valves will again change the RIAA equalization. Bob

[John Caswell]

Can I suggest you build an Hagerman inverse RIAA filter and stick a square wave into it, then into the phono stage, and look at the results. If as you say there I a lift in frequency response you will have a massive leading corner overshoot, which suggests to me you have problems elsewhere.


John

[bob orbell]

I do have a Hagerman inverse RIAA filter, you gave me the components some time ago. I am now giving up on getting sensible feed back about this, for the record, my Phono with 2K2 in R21-22 sound just right, it has taken almost 10 years to achieve the correct value for R21-22, the other with 14K3 does not sound correct and a scope trace confirms my calculations, I am convinced that 14K3 is the wrong value, sadly, no one can or will prove my calculations are correct or wrong, if anyone can, please come forward. Bob

[Richard]

Hi Bob,

How did you arrive at 2k2?

The value of 14k3 is from NealG who wrote the FAQ article and I’ve found and attached his original drawing to it.

Allen Wright’s cookbook gives 2 worked examples;
820pf with 3k87 ( = 50153 ) and
1000pF with 3k18 ( = 50049 )
Phono 3 would be 220pF with 14k4 = 50238

Allen does say to tweak to satisfaction so all I can think is that Peter’s 2k2 in the Operating Instructions is from his earlier testing.

[bob orbell]

Richard, thank you for your attention on this issue, if you read the first post this will explain how and who did the calculations. The figures you posted from Allan Wright are time constant calculations, the equation is t = CR, all calculation must be done in uF and ohms, so, the first is 0.00082 x 3870 = 3.1734, the second is 0.001 x 3018 = 3.018, your figures are 0.00022 x 14400 = 3.168, so we think what is this guy Bob on about ? well, it is NOT as simple as that, the time constant 3.18uS is shared with the 75uS, one now needs to find the total Thevnin resistance that each time constant sees, it is not simple, trust me. After many hours my friend Brian (who has more mathematical ability than all of us put together on this forum) ( no offence intended) came up with 2177.803 ohms, 2K2 was a value,( it is only 82 ohms too much,) I had and fitted two resistors in R21-22, it now sounds excellent, and a few days latter I found the build instructions from Peter C and the value 2K2. This value I am positive is correct. Bob

[Richard]

Thanks Bob, but I cannot see the working relating to 2k2.

Reading the circuit left to right, input to output,

the 2nd corner 318uS/500Hz is of similar construction to the 4th and is formed by 39k2 8n2 which is 495Hz. So is this wrong as well?

[bob orbell]

Richard, I would imagine that there will be hidden capacitance somewhere that will be in series with the 8.2nF capacitor, maybe as small as 50 or 100 pF that must be taken into account, please dont ask me to explain all the Math, it took Brian some time to complete the calculation of the 75 and 3.18 time constant, he tried to talk me through it, but my head still hurts. I hope you now see how complicated RIAA correction is, but the value 2177 ohms does make Phono sound correct and I think also that Peter C came to the same figure, after all, I am sure Peter would have access to very clever engineers before he released the series III. I think it may have been Dave Cawley that used the value in the FAQ when he was spouting about what he had done with Phono, which was nothing apart from changing to the incorrect value of R21-22. Bob

[Richard]

Hi Bob,

Yes stray cap and inductances too perhaps. I went through the Phono 3S calculations about 15 years ago and they agreed with the circuit and the built units checked fine using SW with a Hagerman 3.18uS iriaa.
These days there are calculators online for the valve and filter maths. I checked my old workings in about 10 mins(!) using them and was pleasantly surprised they mostly agreed with the manual ones from all those years ago (I found a minor mistake which made little difference to the corners).
For the filters I used the basic first order calc out of the valves and the RC characteristic calc into the following grids. I don’t know if that’s the correct way but it gave the correct corner figures. It does look strange as 2 corners share one capacitor. Any comment welcome.
I was asking about the use of 2k2 in particular though as if my calc is correct it would put your new “3.18” corner up around 320kHz. As you say, Peter did show that value in a photo and it was also in a circuit drawing, though mine came with 14k3 and matching circuit.
Reading Jim Hagerman the Lipshitz riaa also had a corner around that figure. It’s only a second guess but we know of Peter’s respect for classic designs and if I were to start listening to a new 4th corner why not start there.
Anyway, it doesn’t really matter so long as it sounds good to you. Yours would be a corner part way between Phono2/3 and Phono3S. We could use any resistor between 0R and 14k3 and listen. To trim it with a square wave though a mirror iriaa should be used.

https://www.ampbooks.com/mobile/ampl...ce/calculator/
https://www.omnicalculator.com/physics/rc-circuit
http://www.vacuumstate.com/fileupload/SP_15_Article.pdf
http://www.hagtech.com/pdf/iriaa.pdf
http://www.hagtech.com/pdf/riaa.pdf
https://www.aqvox.de/download/Aqvox-RIAA-50kHz.pdf

[Richard]

I see the Wiki page on RIAA which includes "The Mythical Neumann Pole" (their title) was updated the day after you started this thread Bob,

https://en.wikipedia.org/wiki/RIAA_equalization