Flat Phono Preamp Based on John's P10 and 2SK389

[mediatechnology]

A Fully-Balanced DIN Tonearm harness for the Balanced-Input Flat/RIAA Phono Preamplifier.

I received the Lumberg KV50/6 240 degree 5 pin DIN a day early from Allied Electronics and built a fully-balancedtone arm harness.

Allied has the Lumberg KV50/6 for about $9: http://www.alliedelec.com/lumberg-kv-50-6/70151666/
All Electronics currently have them surplus for around $3: http://www.allelectronics.com/make-a-st ... ing/1.html
Based on the prices of DIN females specifically made for tonearms the above seem very reasonable.

The connector pin block has to be modified slightly to be a DIN tonearm connector.
It takes a couple of minutes at most.


Lumberg KV50/6 5 Pin 240 Degree Female Locking DIN Connector

The pin block has to be removed from the outer metal shell which will be discarded.
Tabs extending from the back of the pin block interlock into an insulating plastic shell that is also not used.
The tabs interfere with insertion of the connector body into the tone arm.
Filing them off to make the end of the pin block round allows clearance.
Apparently this is what makes DIN tone arm connectors "special."


Lumberg KV50/6 5 Pin 240 Degree Female Locking DIN Connector Modified Pin Block

Once the pin block has had the plastic tabs filed down with the body completely circular it can be used to make balanced phono cables:


Lumberg KV50/6 5 Pin 240 Degree Female Locking DIN Connector Modified Pin Block With Leads

The assembly slips into the Audio Dynamics tonearm.

I realize that the Brand-Rex cable isn't made of unobtainium: I could have paid $229 if I wanted unobtainium.

This is a comparison of unbalanced coaxial RCA cables to fully-balanced shielded twisted pair cables using a Stanton 681 dummy cart mounted in a tonearm:

Unbalanced RCA connections with a grounded differential input:


Balanced Flat Phono Preamp with RIAA EQ Applied, Noise Floor, Unbalanced coaxial RCA Connections, Grounded Differential Input, Stanton 681

Balanced shielded twisted pair connections with a fully differential input:


Balanced Flat Phono Preamp with RIAA EQ Applied, Noise Floor, Balanced Shielded Twisted Pair Connections, Fully Differential Input, Stanton 681

The difference in "hash" between 1-10 kHz is significant.
The 60 Hz hum is about 6 dB less.
The performance with the grounded differential instrumentation amp input is good, but given the modest investment to make a balanced phono preamp cable I'd say the improvement from a fully-balanced connection is worth it.

Tomorrow I'll try the fully-balanced harness on the actual turntable.

[JR.]

Besides the hum spikes, that noise floor looks different in character. I would be curious subjectively about what the hiss (not hum) difference sounds like? Surely well below any surface noise so probably moot.

JR

[mediatechnology]

I used my balanced tonearm as a hum sniffer.
I haven't found the exact cause but I have found the area of the room where the hum spike occurs.
I can rotate the arm within the field and get deep nulls.
Unfortunately both channels do not null at the same time.
I think I have a saturated linear wall-wart powering something or a UPS with saturated magnetics directly below me creating a field with a strong odd-order.

Besides the hum spikes, that noise floor looks different in character. I would be curious subjectively about what the hiss (not hum) difference sounds like? Surely well below any surface noise so probably moot.

The area from 2-10 kHz seems to be high-order line harmonics from LED lighting or PC SMPS.
There also seems to be less RF so perhaps some of the difference may be RF rectification folded down into the audio band.
Or, it could be "all of the above."

With the unbalanced RCA coaxial connection I can see the midrange noise rise when I power up the LED overheads.
The workbench LEDs aren't a problem but if I probe the base of them with the "hum sniffer" I definitely see a rise in the midrange band.
(They happen to be the same brand as the overheads.)

The balanced connection seems to reject the very high-order hash well.
It's hard to A-B but the fully balanced noise floor sounds smoother.
Less buzz and more B-flat hum.
It may well be below record surface noise...

I did discover that my balanced link between remote preamp and EQ board were actually terminated SE.
Though at the moment they're less than a foot apart changing it also lowered the high-order hum spikes.

[mediatechnology]

This is the balanced noise floor with an AT-96E mounted on the SP-15.


Balanced Flat Phono Preamp with RIAA EQ Applied, Noise Floor, Balanced Shielded Twisted Pair Connections, Fully Differential Input, Audio Technica AT-96E

The RIAA Output gain was calibrated to +4 dBu with 3.54 cm/s rms with left/right-only modulation.
With both channels modulated the equivalent lateral modulation is 5 cm/s rms.
The AT-96 output is rated at 4.2 mV at 5 cm/s lateral.
The gain at 1 kHz is approximately 49 dB.
0 dBu in the FFT is roughly +8 dBu.

FWIW the surface noise on the CBS STR-100's extended lead-out is approximately 20 dB higher.

[Gold]

Looking good Wayne. This is something I've wanted for a long time. Seeing the measurements is enlightening. Pretty much confirms what I've suspected for a while. Reducing the hash that creeps into every setup will do far more to improve sound quality than another zero or two in the THD or noise floor specs.

[JR.]

While better is always better, if it's 20 dB below the surface noise any incremental improvements will probably be subtle.

JR

[Gold]

The noise spectrum of buzz and hum is very different from rumble, pops and ticks. I'd expect to easily hear improvements. A good record should have a noise floor of about -65dBu.

[brianroth]

The noise spectrum of buzz and hum is very different from rumble, pops and ticks. I'd expect to easily hear improvements. A good record should have a noise floor of about -65dBu.

I'm seeing that -65 dBu number and wondering how far above 0 dBu is the "overload" point. Or, IOW, what is the total dynamic range of a good record?

Bri

[Gold]

The noise spectrum of buzz and hum is very different from rumble, pops and ticks. I'd expect to easily hear improvements. A good record should have a noise floor of about -65dBu.

I'm seeing that -65 dBu number and wondering how far above 0 dBu is the "overload" point. Or, IOW, what is the total dynamic range of a good record?

Bri

Overload is dependent on playback. A Fischer-Price turntable will clip before a high end setup. The EBU broadcast spec was +12dBu as the clip point. The equipment was built to that spec but was of course way over built.

Edit: A more correct way to say it is the EBU standard is 12dB headroom above reference level.

[ricardo]

I'm seeing that -65 dBu number and wondering how far above 0 dBu is the "overload" point. Or, IOW, what is the total dynamic range of a good record?

Tomlinson Holman (da THX man) did an article for Audio magazine which looked at maximum theoretical & practical levels on vinyl using the SHURE and other data. It looked at the maximum levels cut on various records, what could be tracked and warps & other stuff too over the audio & sub/supersonic spectrums.

Wayne, that would be a good article to track down & archive.