Eros Phono kit is in transit - any capacitor upgrades?

[Paul Birkeland]

If you want to use .033uF cap, you'll have to change:

75K resistor to 72.2K
0.01uF cap to 0.0104uF

The 10.5K resistor can stay.

After making these changes, you'll want to measure RIAA accuracy to be 100% sure that the component values are agreeing with the calculations. 

To be fair, there are $10,000+ phono preamps on the market that use the exact same caps in the RIAA eq. 

[Paul Joppa]

By my calculations, the following changes are necessary (nearest standard 1% resistor in parentheses):

75K becomes 68.7K (use 68.1K)
0.030uF becomes 0.033uF
10.5K becomes 9636 ohms (use 9530 or 9760)
0.010uF becomes 0.0113uF

301 ohms on the C4S feeding the plate  becomes 288.5 (use 287)

[Paul Birkeland]

301 ohms on the C4S feeding the plate  becomes 288.5 (use 287)

Forgot about adjusting that one!

[mwaller]

By my calculations, the following changes are necessary (nearest standard 1% resistor in parentheses):

75K becomes 68.7K (use 68.1K)
0.030uF becomes 0.033uF
10.5K becomes 9636 ohms (use 9530 or 9760)
0.010uF becomes 0.0113uF

301 ohms on the C4S feeding the plate  becomes 288.5 (use 287)

I really appreciate the input. I think that is more work that I am interested in...
I have two v-caps that measure .009840 uF and .009843 uF.  Are these going to be sufficiently close to 0.010 to work in the "stock" RIAA network?
I seem to recall reading that the "ideal" value is actually just a bit above 0.010, which has me a bit concerned...
Thanks,
Mika

[Paul Joppa]

The theoretical ideal second cap is 0.010288uF, but the 6922 input Miller capacitance is around 50pF. the wiring probably adds another 10pF, so the target should be 0.010228. Subtracting the measured v-caps gives a residual of 388pF, the closest standard value is 390pF. That leaves you with a ~4% error - not big but possibly audible under some circumstances. Why not pick up a couple and bypass the V-caps?

Ceramic caps are crap for audio, with the exception of Type C0G or NPO which are said to be excellent; their main problem is that only very small values are available. 390pF is very small, and since it is dominated by the v-cap, you do not need precision in the bypass. For small values like this, mica and polystyrene are widely likes as well.

[mwaller]

The theoretical ideal second cap is 0.010288uF, but the 6922 input Miller capacitance is around 50pF. the wiring probably adds another 10pF, so the target should be 0.010228. Subtracting the measured v-caps gives a residual of 388pF, the closest standard value is 390pF. That leaves you with a ~4% error - not big but possibly audible under some circumstances. Why not pick up a couple and bypass the V-caps?

Ceramic caps are crap for audio, with the exception of Type C0G or NPO which are said to be excellent; their main problem is that only very small values are available. 390pF is very small, and since it is dominated by the v-cap, you do not need precision in the bypass. For small values like this, mica and polystyrene are widely likes as well.

Excellent info, thanks! I'll look for a suitable bypass to bring me closer to ideal. 
Can you tell me what the ideal value is for the 0.030 cap?  I've been offered a pair of V-caps around .03125, but I'm guessing they are too big. Thoughts?
Thanks again!
Mika

[Doc B.]

The flaw in this deal is that you guys are perplexing over picofarads, but you aren't talking about measuring the preamp's response when you are done changing everything. What matters is that the preamp faithfully reproduces the RIAA curve, even over the precision of the value of the cap based upon theory. The value that keeps you within, say, 1/2 or 1/4 dB of the established curve is the correct value, even if it deviates a bit from the theoretical value. If you look at the notion of improving the preamp with these mods that way, the best resistors for the eq circuit will then be trimmers.

That said, our preamps are designed to be pretty faithful to the RIAA curve with the stock components.

[mwaller]

My goal in pursuing these mods is not to improve RIAA accuracy over stock.  Rather, I'd like to squeeze the last bit of resolution out of this great circuit without harming the RIAA performance.  I would love to measure performance if I had the tools and knowhow, but I don't.  As best I can, I want to stick to the stock theoretical values since they sound great to my ears!  I'm going with 0.1% foil resistors, so uncertainty in those values will be reduced by an order of magnitude.  Thanks,Mika

[mwaller]

Hi Paul -
Is 0.03125uF too far out for the 0.030uF caps in the RIAA network?
Thanks,
Mika

[galyons]

Hi Paul -
Is 0.03125uF too far out for the 0.030uF caps in the RIAA network?
Thanks,
Mika

Is that the nominal labeled value or the measured value?  When dealing with RIAA EQ, I always measure the capacitance and use that value.

Cheers,
Geary

[mwaller]

Is that the nominal labeled value or the measured value?  When dealing with RIAA EQ, I always measure the capacitance and use that value.

Cheers,
Geary

0.03125uF is the measured value.  Thanks!
Mika

[Paul Joppa]

It's further out than I would like, but probably acceptable to most people.

[mwaller]

It's further out than I would like, but probably acceptable to most people.

What value is ideal per the model?
Thanks, Mika

[Paul Joppa]

What value is ideal per the model?
Thanks, Mika

0.030uF.

[mwaller]

So, I finally got around to installing some new parts in my Eros...
I replaced the output caps with V-Cap CuTF.  In the RIAA network, I replaced the 0.01uf with selected pair of V-Cap TFTF caps and the 0.03uf with a selected pair of AudioCap RT (polystyrene).   I also replaced the 47K input load and EF86 plate load resistors with Texas Components TX2575 "naked" foil resistors.  Sounds great! 
I'm noticing that the power transformer and aluminum chassis of the Eros get quite warm after a hour or so - almost too hot to touch. Is this normal?  It's been so long since I listened to vinyl that I can't remember how it was before the mods...
Thanks,
Mika