Some more RIAA questions

[Sheldon]

I've been following with interest several threads on RIAA stages. I do this hobby to learn and come up with some competent gear. In that spirit, I've been looking at some designs and trying to understand them. The following circuit I took from Steve Bench's site. http://members.aol.com/sbench101/Preamps/RIAA5.gif

Redone here to simplify. Looking at the 75us section (R3,R5, C1), I calculate about 62us. At the operating conditions of the second ECC88, I estimated an anode resistance of about 23.4k. (Estimated ra = 4250. Adding the contribution of the unbypassed cathode R (562*34) gives about 23.4k total.) This in parallel with the 20k load resistor gives about 10.8k. Finally 10.8k*5700pf gives about 62us. I'm missing something.

Similarly, according to Morgan Jones, R7 and R12 should be in a ratio of 9:1 and R12*C5 ;should give 3120us. I get about 3320us. Again, I'm missing something. I don't think the stray capacitances are it, as they would make the value higher still (I think). In this case, the output impedence of the cascode is around 95k.

Anyone willing to set me straight?

Sheldon

edit: C5 = 6500pf, sorry about the small print

[Sheldon]

While I'm at it. How about adding a buffer like this?

[EC8010]

I think you may be neglecting the effect of C3. Unfortunately, C3 interacts with the RIAA so that the simple equations become useful to get you into the right ball park for a computer simulation, but no more.

[Sheldon]

Originally posted by EC8010
I think you may be neglecting the effect of C3. Unfortunately, C3 interacts with the RIAA so that the simple equations become useful to get you into the right ball park for a computer simulation, but no more.

The calculations at least come in the ball park, so that's good. How about C1, though? Any way to get it right without computer syms? It seems that the output cap shouldn't interact that much. But if it does, it means also that input Z of the next stage is critical, so maybe the buffer is necessary, or a dedicated set of values for the chosen next stage?

If the values are close, short of computer syms (which would rely on accurate models, and in which I'm not competent), how do builders tweak the curve? Is there any PC based measurement system that can test the output? Measure in RMAA and compare the -3dB points and levels?

Thanks,
Sheldon

[EC8010]

One way to do it is to use an accurate pre-RIAA network before the stage and adjust your equalisation until you get a flat response.

http://www.diyaudio.com/forums/showt...ht=#post835412

[Sheldon]

Thanks, I'll check it out.

Sheldon