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

Where we discuss new analog design ideas for Pro Audio and modern spins on vintage ones.
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

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

Post by mediatechnology »

I wanted to share with everyone a flow diagram of the Flat Phono Preamp for transfer and DSP-based declicking.

The design has RIAA EQ available for live monitoring of the transfer - or playback of the transfer- prior to final RIAA EQ being applied in software.
Based on doing a few flat transfers and doing both declicking/RIAA-EQ in software, this approach is easier on the ears and improves workflow.

While recording, the monitor output can be non-destructively switched between flat and RIAA EQ without affecting the record path.
On playback of the transfer, while previewing de-clicking, the monitor can be switched between flat and RIAA EQ.
This prevents the need to render the file with both de-clicking and RIAA EQ applied in software to hear the final result.
De-clicking decisions can be made with minimal commitment.
This limits "un-do's."

When the record path is switched to Flat recording, the monitor can be switched to playback without a affecting record.
In this mode it's possible to both record and playback different files simultaneously.
Workflow is increased because files can be processed as recordings are being transferred.

For less demanding recordings which do not require declicking, transfers can be made with RIAA EQ in the record path.

Image
A Flat Phono Preamp For Transfer and DSP-based Declicking with RIAA Monitoring

I'm re-building my prototype and hope to post schematics soon.
ricardo
Posts: 262
Joined: Sun Apr 10, 2011 7:24 am

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

Post by ricardo »

Though I'll probably never re-surrect the vinyl player in my sister's shed on the other side of the continent with all my records, at times like this, I do wish there was a LIKE button on this forum. :D
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

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

Post by mediatechnology »

A couple of weeks ago I rebuilt the Protoboard to see if the proposed flow diagram sped workflow.
I flat-transfered and software declicked the SQ version of Santana's "Abraxis."


The Protoboard and workbench are a fairly long distance from the turntable.
This is a good test of the balanced instrumentation amplifier phono input.
In addition to the 3 foot cables attached to the tonearm there was an additional 10 feet of spiral-wrapped RCA extension cables.
The extension cable alone added about 320 pF loading to the AT-120 MM cartridge.
Despite these obstacles the circuit was relatively quiet.

The flow diagram seems to provide what I need to provide flat transfer and capture while monitoring recording and DSP de-clicking with RIAA EQ.

I was reminded once gain how little gain is needed when recording flat.
I ended up reducing the first (flat) stage gain to about +18 dB for an overall flat gain of +24 dB.
The sound card input overloads around +8 dBu rms.
+24 dB of preamp gain showed un-equalized peak levels ranging around -12 to -9 dBfs with the handful of discs I tested.
This provides some headroom for really hot clicks though less than you might think.
Though it may not make sense to prevent clipping of peaks I found that while doing flat transfer headroom is your friend.

The only question I have about the overall transfer process is when to normalize the file using DSP-based RIAA de-emphasis.
I found that a good compromise in the preamp was to have about 9 dB of gain in the RIAA equalizer but the DSP RIAA EQ tables I'm using have unity gain at mid-band.
I think I need to experiment with adding gain to the EQ definition file or normalize the peak level before EQ.
Fortunately this does not affect the flat preamp or its gain structure.

Image
A Flat Phono Preamp For Transfer and DSP-based Declicking with RIAA Monitoring

Image
Protoboard of A Flat Phono Preamp For Transfer and DSP-based Declicking with RIAA Monitoring

One thing I added was an additional AC-coupled stage after the RIAA EQ.
Owing to the fact that the RIAA EQ stage has about +20 dB DC gain and high-value feedback resistors a DC offset can develop.
In order to permit the use of high bias current bipolar op amps a second film coupling cap was added.
The same high pass stage also provides post-EQ gain and buffers the THAt1646 input.

The only electrolytics (except decoupling) are outside the signal path and are for the THAT1646 OutSmarts feedback loop.
For fully-balanced outputs they are effectively out of the signal path.

In terms of quietness the front-end the choices (using bipolar input op amps) wound up being:

JRC NJM2114
NE5532
LME49720

Though I have not determined why, the LME49720 had popcorn noise.
I tried a handful of samples all of which were from the same lot.
I want to think that this is related to a specific lot of LME49720s but I don't know.

The NJM2114 has more bias current which provides a large amount of DC offset when the input is unterminated and it sees a 47K Rbias.
The same is true, to a lesser extent, for the 5532.
Both however are significantly quieter than the bias-current compensated LME49720.

The LME49720 in the RIAA EQ stage is an excellent performer and my first choice due to it's low Ib.

I hope to post a schematic of the preamp soon.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

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

Post by JR. »

Printing the peaks at -10dBfs before RIAA should be adequate for vinyl's dynamic range. RIAA is effectively 40 dB of HF de-emphaisis. Normalized for 1kHz, 20 Hz will be 20 dB hotter after RIAA, 20kHz 20dB cooler.

I really don't know if this is a limitation or not... running the pre-RIAA right up to 0dBfs may be OK if the convertor clips cleanly, keep in mind the RIAA looks like a LPF, and any overload events will probably be very narrow spikes. So I wouldn't expect much splatter or artifacts from clip events.

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

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

Post by mediatechnology »

Thanks John.

I think I'm going to put some limited gain trim in it anyway so people can calibrate their cartridge levels.

Also just found a Stanton doc where they suggest removing the ground strap bonding L-CH ground to the shell for certain metal head shells.
I think that may be the key to fully-balanced cartridge/preamp interfacing since the tonearm wiring is otherwise not bonded to the ground lead or turntable frame in my SP-15 and ADC arm.
User avatar
JR.
Posts: 3700
Joined: Sat May 24, 2008 7:21 pm
Contact:

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

Post by JR. »

Yup, the grounded cart can form a mutual ground path with turntable ground.

Tone arm should be grounded, just maybe not through the low voltage audio cable.

JR
Cancel the "cancel culture", do not support mob hatred.
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

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

Post by mediatechnology »

The Flat Phono Preamplifier Front End.

One of the problems many of us have is that our turntables are not conveniently located near our workstations or consoles.
After reading posts from Paul Gold and others - as well as having a turntable in my setup that's across the room from my computer - I decided to split the preamp from the RIAA EQ Monitor Switcher.
A remote Flat Phono Preamp allows the cartridge leads to be short and provides a quasi-line level balanced interface to the RIAA EQ and monitor switcher.
The added advantage to this approach is that someone wanting only a Flat Preamp has a less-complicated board to assemble.


The preamp that I originally had in mind is identical to this one.
I added a second THAT1240 to provide an optional balanced output.
The preamp's DC connections are supplied on a separate cable.
Three shielded twisted pairs are required to remote the preamp: Left, Right and DC.

Image
A Flat Phono Preamp with Balanced Inputs and Outputs for RAW, un EQ'd, Recording.

Description

The fully-balanced input is an instrumentation amp having a gain of +13 to +38 dB.
The preamp input is direct-coupled to the cartridge.
Though op amp bias current flows in the cartridge most of the measured data I've seen shows no harmful effects.

For moving magnet MM cartridges having the typical sensitivity of 5 mV at 5 cm/sec (1kHz) 18-20 dB of gain in this stage is plenty.
Though I do not own a MC cartridge I anticipate that 38 dB gain (44 dB overall) will be sufficient for most RAW sessions.

Common mode rejection is realized in the Flat Preamp by a THAT1240 double-balanced line receiver.
For applications where connection lengths are short and balanced outputs are not wanted the second THAT1240 can be omitted.

For use with the companion RIAA EQ Monitor switcher board the output coupling capacitors may be eliminated.
That board will have a film interstage coupling cap.
In free-standing applications particularly those with high gain output electrolytics may be required.

At low gains however there is not significant offset provided a cartridge is connected.
When the input is open circuit a significant bias current induced offset can develop so the choice of using an electrolytic is up to the builder.

Coming up next: The RIAA EQ and switching schematic.
Gold
Posts: 678
Joined: Sat Dec 01, 2007 5:20 pm

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

Post by Gold »

I have to say this is pretty awesome. I've had what this is rolling around in my head for a long time. Some features that I think would be very useful in the base station would be isolated balanced +4dbu outs and consumer RCA outputs. A meter buffer and adjustable output gain.

This may be outside the scope of this project but I think an EQ to help correct common interface problems and cartridge response would put it over the top.
Gold
Posts: 678
Joined: Sat Dec 01, 2007 5:20 pm

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

Post by Gold »

I have a couple of questions about the schematic. What is the advantage of using the instrumentation amp over the THAT microphone preamp IC? (My guess is source impedance). Why use the 1240's as line drivers and not a 1646?
User avatar
mediatechnology
Posts: 5444
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

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

Post by mediatechnology »

I have to say this is pretty awesome. I've had what this is rolling around in my head for a long time. Some features that I think would be very useful in the base station would be isolated balanced +4dbu outs and consumer RCA outputs. A meter buffer and adjustable output gain.


Thank you Paul.

The goal is to provide enough gain to produce +4 dBu on the equalized outputs.
I hadn't thought of a meter buffer but it makes sense to provide the ability to hang mechanical meters off the outputs.
That's a great idea.

I've put gain adjustment into the flat front-end to calibrate cartridge levels and channel balance but that might not be the best place to provide other level calibrations.
I'm thinking that for playback calibration purposes most people would want to meter post-RIAA EQ.
Since the post-EQ gain stage needs to have some gain anyway that might be a good place to put an additional calibrated trim.
What is the advantage of using the instrumentation amp over the THAT microphone preamp IC? (My guess is source impedance). Why use the 1240's as line drivers and not a 1646?
A THAT mic preamp IC would be an excellent moving coil front end. The noise current (and also bias current) is a little too high for moving magnet preamps.
Though the DCR of a MM is modestly low, say 1K Ohm, the inductance of the coil combined with the high noise current of the IC work against each other.
So your guess of source impedance and inductance which is source impedance vs. frequency is spot on.

The 5532 is significantly quieter than most mic preamp ICs in this application.
The THAT1510 has a current noise of about 2 pA/sqrtHz; the 5532 0.7 pA; the NJM2114 around 0.4 pA/sqrtHz.
The SSM2105 is about the same as the THAT1510, INA217 etc.
The LME49860 around 1.6 pA/sqrtHz
A BiFET op amp would have very low current noise at the expense of voltage noise.
A discrete FET solution like JR's P10 - what I tried first - will have the lowest current noise but getting good affordable FETs in through hole is difficult.

I used THAT1240s (one originally) to realize common mode rejection in the flat front-end.
The 1646 can't realize common mode rejection at its input, though its sibling the THAT1606 can.
I'm also not sure that the 1606s common mode rejection is actually trimmed.
The 1606's input CMR is 40dB minimum, 60 dB typical.
And, unfortunately the 1606 is only available in SMT.

Given the difficulties of SMT and the superior common mode rejection of the 1240 (90 dB typ) I decided to use the 1240 to realize common mode rejection and live with its lower current outputs.
To provide a fully-balanced output I used a second 1240 connected in a cross-coupled double-balanced topology.
The 1240 however has two limitations the 1646 does not.
It does not have OutSmarts and it cannot drive 600 Ohms to +21 dBu.
My thinking is that the 2K Ohm +21 dBu output of the 1240 will be sufficient to drive the EQ/Monitor Switcher since it presents a bridging load and that there's quite a bit of EQ gain after the preamp.
Secondly the flat preamp used standalone driving an A/D for RAW transfers will probably never see anything higher than +15 dBu.
I considered adding a final 1646 stage but thought that it might introduce additional noise, current and require more space.
The EQ and final RAW outputs will be buffered by 1646s.

I'm open to suggestions about added features or changes.
My application is pretty much one thing but in a mastering context there's a lot I haven't thought about.
Since you do that every day you can provide a lot of input.
It was the work of all of you posting here that made the MS so versatile and successful.
Post Reply