Phono preamp without confusing RIAA stuff

[sled108]

Just go Here..And use what you have..

The Phonoclone and VSPS PCB Help Desk

[PLB]

Hi PhantomBox,

Here in USA, we celebrated Presidents Day on Monday (didn't have to work that day) so I took the chance to check out the circuit of the "PAiA - RIAA phono preamp kit" and derived a mathematical model for it. This model defines its transfer function, which I then used to find the idealized component values, by equating its coefficients with those of the RIAA curve.
Are you still interested in using the "PAiA - RIAA phono preamp kit" if I were to provide you with the correct component values ?

Regards

Peter

[PhantomBox]

Sure. I'm always open for ideas. More so if you took the time to check and optimise the RIAA phone preamp (thanx a lot). Can you send the values through the forum so anyone could use the info or do you need an email address?

[PLB]

Hi PhantomBox,

No, I don’t mind sharing these results with you and any other DIYAUDIO member.
Here are the changes, which I've referenced to the schematic you linked to in your first post ie, PAiA - RIAA Phono Preamp Kit

Change C3/C8 from 47uF to 220uF
Change C4/C9 from 1nF to 2.136nF
Change C5/C10 from 4.7nF to 6.247nF
Change R4/R8 from 68K to 38.02K

With these value changes entered into a Spice simulation I was able to verify the compliance of the response to within +/-0.05dB compared to the response of the RIAA standard.
The value of C4/C9 can be constructed by paralleling a 1.8nF with a 330pF, C5/C10 can be constructed by paralleling a 5.6nF with a 680pF and for R4/R8, a 38.3K can be used.
I suggest that you use +/-2% tolerance or better for the capacitors (except C3/C8) and for resistors, use +/-1%.
The value of C3/C8 has been made large so as to minimize the low frequency droop that would otherwise result with the original value of 47uF.
However, with a 100uF capacitor, a loss of only 0.15dB at 20Hz results, but of greater benefit would be the increased attenuation of unwanted signals caused by record warp. In other words, it will provide a better rumble filter.
As is common with this circuit topology there is an extra zero in the transfer function that is located at just over 212KHz. At 20KHz, the effect of this extra zero is quite minimal and for the circuit values I’ve given you above, a rise in the response of about 0.03dB is observed together with a phase change of 5 degrees. Hope all this helps.

Have fun.

Peter

[PhantomBox]

Thanks Peter.
Since the capacitor values are a little hard to find (YV = a tough place) I will comment to you as soon as I have made out a supplier. Would C4/C9 work with 2 x 1nF in parallel and C5/C10 with 4.7nF + 1nF in parallel?

[mlloyd1]

i'd add 150pF to the 2x1nF

also, 4.7nF + 1.5nF is closer if you can find the 1.5nF

good luck,
mlloyd1

[PLB]

Hi PhantomBox,

I’m not sure I understand your question, so I will give you two scenarios.

With reference to the attached Spice simulations, the blue curve shows the response if all of the component values are changed to the ones I posted, except C4/C9 and C5/C10, which I have made 1nF+1nF and 4.7nF+1nF respectively (as in your post).
The red curve shows no components having been changed except C4/C9 and C5/C10, which I have made 1nF+1nF and 4.7nF+1nF respectively.

These responses are error responses relative to the RIAA curve.

Regards

Peter