Phono Transfer System Construction Information

[Gold]

I'm making three separate PTS's. Each will have a tonearm attached. I have the three working. Doing the frequency response tests in the construction information produces the stated results. My gain structure is a little different so I'll use my numbers instead of what's in the instructions. If there is a 1k tone at -12dBu at the input, the output will be 0du. I have 6db gain post EQ plus the 1646. At 20K I get +19.8dBu. So far so good. When I check 200Hz (actually 197.5Hz) I get +8.3dBu like in the instructions. When I test at 20Hz I'm only getting +15.25dBu. Shouldn't it be +20dBu?

This is using the AP P1. Generator has output impedance of 40Ω. I think in the past I only checked it at 200Hz as in the instructions.

[mediatechnology]

The rolloff at 20 Hz is intentionally introduced - to reduce warp - by the value of the film caps C33-C36.
You can make them much larger in value using electrolytics and it will follow the curve to 20 Hz.

[Gold]

Thanks Wayne. I’ll see what I have and what will fit. I rhink I’ll stick with a film cap.

[JR.]

IIRC the IEC proposed a real pole at 20Hz (7950 uSec) back in the 80s (?).

The RIAA dismissed that suggestion and left the sub 20Hz response pretty much undefined beyond the three standard RIAA time constants (75uSev, 318uSec, 3180uSec).

Back in the 80s when I was trying to escape from the audiophile market I made my last phono preamp (ever). Hoping to appeal to a more professional market, I incorporated the IEC real pole at 20Hz.

Without my knowledge, my preamp prototype was shared with a well respected reviewer. He observed that my low bass response was a little weaker than the $5k Mark Levinson he was comparing it to..... Duh

====
[TMI]
Back in the day when I was designing companding NR for tape recording, record warp and/or not centered record holes could wreak havoc with vinyl transfers to tape. The excessive very LF warp energy would get encoded into the compressed signal, but not reproduced by the tape path (especially cassette tape paths) to be available for accurate playback expansion. What resulted was a phantom LF signal modulation caused by the missing very LF content. I fixed that inside my NR compander kit with an adaptive HPF that increased the LF cutoff pole only during low input levels . [/TMI]

Sorry about the veer...

JR

[Gold]

I have some 4.7uF Cornell Dubilier PP film caps that I used in the Sontecish EQ. I still have some that are within a couple of precent of each other. This produces +19dBu @ 20 Hz.

I’ll leave the other ones with 1.0uF in place and will listen to see which I like best.

[mediatechnology]

On the latest run of RIAA boards I added more capacitor footprints for C33-C36.

We found that when used as a cutting amp, with feedback, a 47uF should be used to reduce phase shift at C33-C36 so I added an electrolytic footprint.

[JR.]

Note: back when I spent too much time on my bench messing with phono preamps (1970s) I discovered a phase shift issue caused by a 22uF aluminum electrolytic capacitor's internal inductance (in a gain leg in series with 360 ohms).

I expect modern electrolytic capacitors are much better but maybe keep the film caps in parallel with the electrolytic?

Or not.

JR

[mediatechnology]

These are working into a 14K3Ω load.
Wima now have some nice 5 mm lead spaced films in the MKS2 series that go up to 6.8µF so keeping them film is do-able.

[JR.]

These are working into a 14K3Ω load.
Wima now have some nice 5 mm lead spaced films in the MKS2 series that go up to 6.8µF so keeping them film is do-able.

Cool... I was measuring phase shift at 20kHz from a 22uF caps ESL working into a 360 ohm R. 14K should not cause an issue even with modern electrolytic capacitors.

Sorry for the veer...

JR

[Gold]

These are working into a 14K3Ω load.
Wima now have some nice 5 mm lead spaced films in the MKS2 series that go up to 6.8µF so keeping them film is do-able.

It looks like if you want polypropylene 4,7uF is the largest easily available. Mouser has offerings in a 22.5mm LS for the old boards from Panasonic and Kemet. The Cornell-Dublier I have aren't listed now. There is a 6.8uF Kemet PP 22,5mm LS listed but as non stock.

It doesn't look like there is enough space on the PCB to do a 27mm lead spacing. There are 6.8uF and 10uF PP with 27mm LS from multiple manufacturers including Wima.