Upper limit of cartridge output?

[oguinn]

I’ve seen that the lower limit for the Eros is in the 1mv range, with some success down to 0.3 mv. Is there an upper limit? I’m looking at the Grado Platinum, for instance, that has a 4.8 mv output. Would that cause any issues? Any upper threshold to keep in mind?

[Paul Joppa]

The answer depends on a lot of assumptions, unless you do the experiment with a variety of cartridges and setups. That said, I have done the following theoretical calculation:

Peak output should be around 20vRMS before clipping (we haven't done this test, but I think we should some time).

A (very!) few cartridges can track 50cm/sec, at least at some frequencies. (That's 20dB headroom above the 5cm/sec reference, and very few recordings are cut with more than 14dB headroom.) But 20dB is a factor of 10, so a nominal 2v output allows for a 20v peak.

Eros gain at midband (1000Hz) is about 300, which means 6.7mV reference input would make 2v output.

Based on the above, my guess is that 6.7mV at 5cm/sec is an extremely conservative estimate of maximum input.  I can't say I've ever heard of someone clipping an Eros on a commercially-available recording with a commercially-available cartridge.

[oguinn]

That’s an extremely detailed and much appreciated reply, PJ. Thank you!

[grufti]

Is the gain in the EF86 about 110? The RIAA curve @ 1kHz is 20dBV down - factor 10. Gain in the 6922 is about 30. (110 - 10) * 30 = 300. Is that it? I've owned some version of the Eros for a decade - Eros 2 at this point, bit I've never measured just the first stage.

[Paul Joppa]

Yes, your analysis is correct. Pentode gain can be very high, though problematic at higher frequencies in the usual implementations. (It's even higher in the tape version ...)

Because of the somewhat unusual circuit arrangement in Eros, the gain of the EF86 is greatly reduced at higher audio frequencies. This lets the first stage be stable at high frequencies and insensitive to the various parasitic capacitances that limit pentode performance at high gain.

[grufti]

That makes the Eros circuit even more of an achievement than I suspected. I'll definitely study it in some more detail now. Thank you, Paul.