Alternative Acoustics preamp

[PRR]

> feeding cathode

Actually: feeding 2K||2K to ground.

The cathode hangs-on near there, but its loading is reduced by high NFB.

The network is, by inspection, classic RIAA. The midband gain (~~1KHz) is about 99.8K/1K or 100, a typical 1KHz phono gain. The highs roll-off to unity, and will be around 10 at 20KHz. The lows rise toward 1.56Meg/1K or 1,560. Which is higher than the RIAA 20dB rise over 1KHz, but the forward gain of two degenerated 12AX7 stages is not a lot over 1,000, so we have to trim the <50Hz gain to hit RIAA.

V1B wants a grid resistor.

 

[CJ]

thanks PRR!

yes it seems that there are many circuits kind of like this one>

 

[CJ]

and these, some circuits  use RC networks shorted to gnd, these use the NFB scheme,

why two networks, can  not you get the same rolloff with one?

 

[Tubetec]

Years ago I used run the Quad 22 preamp ,not the most accurate Riaa ever ,just one stage of Ef86 with feedback for the equalisation.
I still miss the vocal sound it made from vinyl though .
Subsequently converted a couple of these to mic pre's, in conjunction with stepup transformers,  they work ok on dynamic mics with high level sources,the mid cut control tames the edge on a 58 or 57 nicely . Of course hi-z outputs mean you need an extra output stage to drive line levels .

 

[Tubetec]

Heres the pick adapter component list ,they catered for all the different equalisation curves in use up to then.

 

[CJ]

well you need a test record to check for crossover points, harmonic tracing distortion,  playback loss, wow, flutter, warpage, stylus wear, noise modulation, acoustical radiation, recorded distortion, high frequency attenuation,  volume expansion, groove width, bottom radius and groove diameter.

 

[Gene Pink]

well you need a test record to check for crossover points, harmonic tracing distortion,  playback loss, wow, flutter, warpage, stylus wear, noise modulation, acoustical radiation, recorded distortion, high frequency attenuation,  volume expansion, groove width, bottom radius and groove diameter.

Let us not forget testing the ultimate tracking ability. I have a master for this. ;D

 

[Tubetec]

I used have a record with test tones on it ,excellent for fine tuning the performance of the stylus/tone arm. It had pure tones cut into it at various levels and positions on the disk, any mistracking could be heard as distortion of the sine ,you gradually moved from lower to higher frequencies ,and higher amplitudes ,tweeking things each time . Unfortunately I lent it to a DJ/promoter type ,never saw it again ,

 

[PRR]

> why two networks, can  not you get the same rolloff with one?

From historic dithering, the RIAA curve has a kink around 1KHz.

You need an R to set the <50Hz gain. A C to set the slope 50Hz-500Hz. An R to set kink gain 500Hz-2KHz. A C to set gain 2KHz up.

So 2 R 2 C.

Your first plan in #21 lacks an explicit 50Hz corner. They just set the 1KHz gain so high that the thing run out of gain somewhere near 50Hz.

Second plan in #21 seems to do the same thing a different way. As it only has one stage inside the loop it will suffer less with tube variations.

"Spiced" in #22 has the full 2R 2C network. If the amp had infinite gain the 2.2Meg would be nearer 870K. SPICE was run to estimate the error due to non-infinite gain and find a value for RIA2 which hit the goal.

2nd plan in #22 has the 2R 2C but again the big resistor has to be bigger than ideal to cover the non-infinite gain.

Turner's plan is 2R 2C and the big R is only a hair big because he did try to get max gain in 2nd stage, and buffered that.