Flat Balanced Input Moving Coil Phono Preamp Construction Information

Construction information for DIY projects, including the MS Mid Side Matrix, Elliptic Equalizer, Mastering Console, Phono Transfer System, Insert Switcher and the Dual Class-A Amplifier. You can post your baby pictures here.
Post Reply
User avatar
mediatechnology
Posts: 3910
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Flat Balanced Input Moving Coil Phono Preamp Construction Information

Post by mediatechnology » Tue Oct 01, 2019 7:07 am

This thread provides construction information for the "Low Noise Balanced Input Moving Coil Preamp Using the ZTX851" discussed here: https://proaudiodesignforum.com/forum/p ... ?f=6&t=783

Circuit boards and assembled and tested Flat Balanced Input Moving Coil Phono Preamps are available now.

Product links:

Assembled and Tested: https://ka-electronics.com/shop/index.p ... uct_id=104
Bare PC Board: https://ka-electronics.com/shop/index.p ... uct_id=105

Flat Balanced Input Moving Coil Phono Preamp assembly and test instructions:

https://proaudiodesignforum.com/images/ ... 2719_1.pdf

Image

Image

Large schematic: https://proaudiodesignforum.com/images/ ... ematic.png

Circuit Description: Introduction

This is a circuit description for the Flat Balanced Input Moving Coil preamp. The low noise perfromance of this preamp is made possible by the exceptional ZTX851 transistor. A special thanks goes out to Winfield Hill and Paul Horowitz who "discovered" the ZTX851 and also to the late Brad Wood aka "bcarso." A few days after Christmas 2015 Brad Wood wrote in an email discussing MC preamps:
"The parts that Horowitz and Hill (I finally got the 3rd edition) now recommend for low source Z are the Diodes Inc./Zetex ZTX951 and ZTX851, which is down where the discontinued parts are pretty much (but the noise is not specified). They sketch an example of a direct ribbon mic preamp using paralleled devices, with a claimed 70pV/sq rt Hz performance."
I emailed Brad back the following January 8th to let him know I'd built a prototype using the ZTX851 and that they were as good as Horowitz and Hill had measured. This is one of the many examples of "gifts" Brad bestowed upon me. If it had not been for Brad's suggestion I would have never made this preamp.

Preamp Overview

The Flat Moving Coil preamp, like its moving magnet counterpart, features a balanced input. The obvious advantage of a fully-balanced input is reduced noise pickup from common mode interference in the interconnecting cables between the cartridge and preamp. A second advantage to using a floating balanced input is the ability to direct couple the cartridge to the preamp inputs with reduced cartridge current.

An input transistor with medium gain operating at high collector current will have a high bias current in the tens of µA that, if direct coupled, flows into a grounded cart. A differential input with a floating cart connection has greatly reduced current flowing in the cartridge because the input offset current, Ios, is many times lower than the single-ended bias current, Ib. Ios is the difference in bias currents. Differential connection to the cartridge eliminates a very large coupling capacitor.

The instrumentation amp topology with the input transistors biased by the op amps was first shown by Demrow in 1968 as an instrumentation "data" amplifier. Motchenbacher and Fitchen in 1973 show the "op amp around the transistor" bias scheme. Cohen, in his paper published in 1984, adapted this same topology to a low noise microphone preamp. Cohen's design kept circuit impedances low and like Demrow et al used the op amp stages to sink emitter current eliminating a potential noise source. Cohen followed the front end with a cross-coupled common mode rejection stage.

The Flat Balanced Input Moving Coil Preamp follows Cohen's design philosophy closely with the exception of the necessity to use matched thermally-coupled ZTX851 transistors instead of the highly-matched, but somewhat noisier, LM394. A fully-differential servo is also used to control output offset.

In the following sections the left channel is described.

Front End

Input is applied to three pin Phoenix connectors. Capacitor C1 provides termination for the moving coil cartridges that require it. Resistors R1-R3 along with jumpers J1-J3 set the termination resistance. The value of R1 should be left at 10Ω for final test but the values of R2 and R3 can be whatever value is needed. R1-R3 are in parallel with R4 and R5. With J1-J3 open, the maximum differential termination is 998Ω formed by bias resistors R4 and R5. When calculating termination values be sure to include the 998Ω in parallel.

The bias current through R4 and R5 pull the inputs negative. To provide bias current "assistance," a small +15 mV compensating offset is introduced across R33. (This circuit trick was also used in the Audio Precision System 1 front-end.)

Q1 and Q2 are a hand-matched differential pair selected to have Vbe within 1 mV. R6 and R7 are the collector loads. Emitter current is "sunk" by IC1A and IC1B through R14 and R13 to provide DC feedback and set the operating point. The nominal operating current for the front end is 5.5 mA per device. The bias control points are the noninverting inputs of IC1A and IC1B. The voltage read from TP2, referred to ground, will be equal to the collector voltage of Q1. IC1A servos collector current by sinking a nearly-identical emitter current through R14. The output voltage at IC1A and IC1B is about -2.3 VDC. Because it supplies emitter current IC1 operates in heavy class-A.

The bias circuit will be discussed later in the servo description.

Differential "current" feedback is applied by a "U-pad" formed by the loop consisting of R13, J4, VR1/J5, R12/J6, R11/J7, R10 and R14. To maintain low noise the "Rg" gain resistors (R10-R12 and VR1) are kept to very low values. The gain of the first stage is 1 + [(R13+R14)/Rg]. For high gains the jumpers should be used since adjustment of VR1 below about 3Ω becomes difficult due to it's internal "hop-off" resistance.

C2 and C3 provide RFI protection. R8, R9, C4 and C5 provide AC stability. C4 and C5 provide local HF AC feedback around IC1A and IC1B. R8 and R9 isolate the high output capacitance of Q1 and Q2 from the inverting inputs of IC1 to allow C4 and C5 to be lower values. (If the builder wishes to lower R8 and R9 or change op amps be sure to look at the transient response using a square wave.)

Due to low feedback impedances IC1 should have high current outputs and be specified for driving "600Ω" loads. The NJM5532 is the best all-around op amp for this high-current application. The PC board has also been tested with the NJM2068, NJM2114 and an OPA1612 mounted on a DIP adapter. I'm not fond of the LME49720's EMI problems so I haven't tested it.

Common Mode Rejection Stage

IC3 and IC4 form a cross-coupled common mode rejection stage. Common mode rejection has many benefits. One benefit is to reduce hum and other common mode interference in interconnecting leads. A second benefit of common mode rejection is the elimination of coupling capacitors. Common mode rejection "burns off" the -2.3V DC component present at the outputs of IC1. Note that the ground "reference" connection for IC3 and IC4 telescopes all the way back to the input connector. An additional benefit of realizing common mode rejection is even-order distortion cancellation. "Double-balanced," an RF term of art and one recognized by Cohen, provides further even-order distortion reduction.

IC3 and IC4, because they are cross-coupled, also provide 6 dB gain. The additional gain provided by IC3 and IC4 reduce the peak current demands of IC1. The outputs of IC3 and IC4, when measured differentially, clip at +27 dBu. The input differential clipping point of IC3 and IC4 is +21 dBu. Each IC1 output has has to supply +15 dBu signal current into 316Ω in addition to 5.5 mA emitter current at clipping. If not for the voltage gain of the cross-coupled stage the peak signal current demands placed on IC1 would be double.

The outputs of IC3 and IC4 feed both the output Phoenix connector through R21 and R22 as well as the input to the differential DeBoo correction servo.

Differential Servo and Bias Network

The Flat Balanced Input Moving Coil Preamp is direct-coupled from input to output. There are no electrolytic capacitors in the signal path.

A servo is used to eliminate the output offset developed by Q1 and Q2's Vbe mismatch. To provide servo correction to the front-end, the emitter currents of Q1 and Q2 are varied by comparing the voltage control points at the non inverting inputs of IC1A and IC1B to the collector voltages at the inverting inputs. When IC1A is held in DC null, the voltage at TP2 will equal the collector voltage at Q1.

In a traditional "Cohen" topology the non-inverting inputs of IC1 share a common reference divider. To provide differential servo correction the voltage divider is split into two arms. R15, R16, R19 and R20 form a bridge circuit. If the servo is not providing correction, the DC voltage at the lower ends of the bridge arms are at 0V. A voltage divider is formed by R15/R19 and R16/R20 to provide 5.5 VDC at TP2 and TP3. The DC feedback loops around Q1 and Q2 set their collector voltages to 5.5V which, due to the collector loads being 1K, results in a 5.5 mA Ic.

IC2, an OPA2277, along with C7 and R23-R29 form a differential version of the DeBoo integrator. A DeBoo integrator has the advantage of having a passive input pole to allow a low bandwidth, DC precise, op amp to be used. The 0V common mode reference for the servo is provided by IC3 and IC4. Differential offset at the outputs of IC3 and IC4 are integrated by C7, R23-R26 and the instrumentation amp formed by IC2 and R27-R29. J8 defeats servo operation for testing. The outputs of IC2 feed R19 and R20 to inject a differential correction voltage back into the lower arms of the bias bridge.

The servo time constant and the differential gain inside the servo control loop (Av=4) sets the low pass response. When servo feedback is injected back into the input stage the result is a derived, subtractive, single-order high pass filter. The servo control point is not affected by front-end gain. The -3 dB point of the high pass response is approximately 13 Hz and, owing to the servo control point location, does not vary with front-end gain.

Reference Shunt Regulator and Supply Bypass

An 11V supply is used for the collector loads of Q1 and Q2 to keep signal swings at the inverting input within the common mode range of IC1.* A TL431 shunt regulator is used to subregulate the +15V supply to +11V. R15 and R16 also reference to the 11V supply. R32 is also tied to the 11V supply in order to regulate bias current assistance. (* Cohen used two red LEDs.)

R30 and R31 set the TL431's output at 11V. C8-C10 bypass the TL431.

R36, R37, D1 and D2 offer reverse-polarity protection for the preamp. Distrubuted bypass capacitors include C11-C14, C26 and C27.
https://ka-electronics.com

Free Hong Kong.
"We’ve Seen This Movie Before–How Communist China Is Following Nazi Germany’s Playbook Against America" https://www.breitbart.com/politics/2019 ... t-america/
"China Warns Brands to Accept Communism, Freedom of Speech ‘Ridiculous’" https://www.breitbart.com/asia/2019/10/ ... idiculous/
"There's A War On For Your Mind" Banned.Video https://banned.video/

User avatar
mediatechnology
Posts: 3910
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Flat Balanced Input Moving Coil Phono Preamp Construction Information

Post by mediatechnology » Wed Oct 02, 2019 12:56 pm

Added circuit description.
https://ka-electronics.com

Free Hong Kong.
"We’ve Seen This Movie Before–How Communist China Is Following Nazi Germany’s Playbook Against America" https://www.breitbart.com/politics/2019 ... t-america/
"China Warns Brands to Accept Communism, Freedom of Speech ‘Ridiculous’" https://www.breitbart.com/asia/2019/10/ ... idiculous/
"There's A War On For Your Mind" Banned.Video https://banned.video/

User avatar
mediatechnology
Posts: 3910
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Flat Balanced Input Moving Coil Phono Preamp Construction Information

Post by mediatechnology » Mon Oct 07, 2019 4:15 pm

Flat Moving Coil Phono Preamp PC boards are available now.

https://ka-electronics.com/shop/index.p ... uct_id=105
https://ka-electronics.com

Free Hong Kong.
"We’ve Seen This Movie Before–How Communist China Is Following Nazi Germany’s Playbook Against America" https://www.breitbart.com/politics/2019 ... t-america/
"China Warns Brands to Accept Communism, Freedom of Speech ‘Ridiculous’" https://www.breitbart.com/asia/2019/10/ ... idiculous/
"There's A War On For Your Mind" Banned.Video https://banned.video/

billshurv
Posts: 369
Joined: Thu Feb 01, 2018 7:07 am

Re: Flat Balanced Input Moving Coil Phono Preamp Construction Information

Post by billshurv » Tue Oct 08, 2019 4:21 am

My build on this has finally started!

User avatar
mediatechnology
Posts: 3910
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Flat Balanced Input Moving Coil Phono Preamp Construction Information

Post by mediatechnology » Wed Oct 09, 2019 8:31 pm

billshurv wrote:
Tue Oct 08, 2019 4:21 am
My build on this has finally started!
You're moving along quickly.
https://ka-electronics.com

Free Hong Kong.
"We’ve Seen This Movie Before–How Communist China Is Following Nazi Germany’s Playbook Against America" https://www.breitbart.com/politics/2019 ... t-america/
"China Warns Brands to Accept Communism, Freedom of Speech ‘Ridiculous’" https://www.breitbart.com/asia/2019/10/ ... idiculous/
"There's A War On For Your Mind" Banned.Video https://banned.video/

billshurv
Posts: 369
Joined: Thu Feb 01, 2018 7:07 am

Re: Flat Balanced Input Moving Coil Phono Preamp Construction Information

Post by billshurv » Thu Oct 10, 2019 5:12 am

Slowly but surely.

User avatar
mediatechnology
Posts: 3910
Joined: Sat Aug 11, 2007 2:34 pm
Location: Oak Cliff, Texas
Contact:

Re: Flat Balanced Input Moving Coil Phono Preamp Construction Information

Post by mediatechnology » Thu Oct 10, 2019 5:20 am

I have three PTS boards to build today.
An MM pre, a MC pre and an RIAA-EQ.
I had sold out of my assembled ones.

I've been putting it off because the weather has been beautiful and just gotten cooler.

I think I'm going to build all three at one time doing all the sockets in one pass, resistors in a second pass etc. per instructions.
That way I only have to rifle through the resistor bins once.

Should probably mention a "mistake" that both myself and another customer have made and that is to hook it up to a DL-103 and forget to remove the 10Ω test termination.
They do sound very dark and constrained at 10Ω.
It was one of those "Duh!" moments....
https://ka-electronics.com

Free Hong Kong.
"We’ve Seen This Movie Before–How Communist China Is Following Nazi Germany’s Playbook Against America" https://www.breitbart.com/politics/2019 ... t-america/
"China Warns Brands to Accept Communism, Freedom of Speech ‘Ridiculous’" https://www.breitbart.com/asia/2019/10/ ... idiculous/
"There's A War On For Your Mind" Banned.Video https://banned.video/

billshurv
Posts: 369
Joined: Thu Feb 01, 2018 7:07 am

Re: Flat Balanced Input Moving Coil Phono Preamp Construction Information

Post by billshurv » Thu Oct 10, 2019 7:15 am

I've been doing it that way with the 6 boards in the current build. For once I sorted the resistors into piles so 1-1k in one 1k1-100k in a second and 100k+ in the 3rd. Making things a lot smoother, although I know I'll have a slight dealy sorting the capacitors.

Post Reply