Imitation interrupted to repeat a rebuild

 

When we last met, I was working on locating the inductors needed to finish the Low Shelf filters in my API style LCR EQ. That task was sidetracked when I was approached by an engineer/producer and asked to rebuild his Yamaha M1516 console. He wanted transformer direct outs, new capacitors, 8 channels of LCR EQ, modern monolithic opamps and upgraded output transformers plus a unique request that I will address later. This would be the perfect opportunity to refine the LCR filters even further. But first I needed to replace a few hundred capacitors and build 40 more OPA604 to TA7322 adapters. I wasn’t very impressed with the adapters from Brown Dog Audio in Idaho. Instead I built mine using SOIC to DIP adapter PCBs purchased from eBay. Cut one trace and add a couple of jumper wires and they sound great. I spent less time assembling these than the Brown Dog adapters and mine are smaller, therefore fitting easier in the tight locations.

Photo Nov 20, 11 40 45 AM.jpg

Photo Nov 25, 1 31 37 PM.jpgPhoto Nov 27, 2 11 29 PM.jpgPhoto Nov 27, 2 11 52 PM.jpgPhoto Nov 27, 11 03 26 AM.jpgPhoto Nov 25, 2 15 20 PM.jpg

Now back to the missing inductors. After spending a few hours doing parametric value searches on Mouser, Newark and Digikey, I was at a loss. Most large inductors made today are intended for use in switching power supplies and other “high efficiency” buck boosting topologies. Their “Q” values are far too small to use in a musical EQ circuit, or their DCR is far too low. After some deep research I located an article that Scott Dorsey wrote for Recording magazine in 2005, where he used custom inductors wound by Magnetic Circuit Elements in Salinas , CA. After a website visit and some calculations I picked a few values to experiment with and placed the order. The downside of custom wound inductors is the lead time. They quoted 5-6 weeks but delivered in 24 days. While waiting for the order I fine tuned the resistor values in the High Mid and Low Mid filters to closer match the filter gains to closer match the stock gain of +/-15db. You may remember the first batch I made had a gain of +/-18db. When the order arrived, I mocked up a test filter set to verify my point to point soldering.

 

Photo Dec 08, 4 44 16 PM.jpg Photo Dec 17, 8 49 01 AM.jpg

I found some neat double sided prototype circuit boards that held 5 of the larger inductors and their necessary resistors. Another matching PCB held both the Low Mid and High Mid filter sets. I must comment on the quality build of the custom inductors from MCE. Nice thick bobbins and neat, clean wire routing. They made for a good looking package. But it’s not about how the circuits look, it’s about how they sound. I’ll be there soon.

Low Shelf EQ.png

In this EQ opamp circuit, a Low Shelf filter consists of a relatively large value inductor and resistor. The inductors are wired in series so each lower filter uses an increasingly larger inductor stack. The new circuit boards fit nicely side by side where the old gyrator opamps were mounted. I installed the test channel and powered up the console. To my delight, the Low Shelf filters were very close to the expected values.

Photo Dec 19, 5 30 17 PM.jpg

Photo Dec 10, 5 13 45 PM.jpg

The predicted DCR value of the coils was spot on, allowing the calculated resistor values to match the gain of the five filters within 0.5db of each other. Pink noise tests confirmed a shallower slope than the stock gyrator filters, but not by much, maybe 0.75db/octave less. But no one listens to pink noise, so I played a few audio tracks and plugged in a few mics. The Low Shelf filters were broad thick brush strokes of bass. To paraphrase, when it comes to bass, the inductors giveth and the inductors taketh away. The inductor filters sounded richer than the stock gyrator filters. The inductors induce a slight amount of harmonic distortion around the filter slope that I find pleasing. The stock filters seem a little sterile by comparison. Here are some fancy graphs and charts.

 

First the measured stock Low Shelf EQ filters at maximum gain.

Low EQ Stock.png

Next the measured LCR Low Shelf EQ filters at maximum gain.Low EQ Inductors.png

And finally the comparison between stock and modified.Low EQ Comparison.png

With the inclusion of the new inductors from MCE, I am very satisfied with the Low Shelf filter sets used in this M1516. I will be able to also use them in the API312/553 clone from part 1 but for now I’m enjoying how they sound in this console. Since I have experienced success with the Low Shelf, I thought I should give the High Shelf a go. The M1516 has a single fixed frequency for the High Shelf. I chose a different set of RC values that dropped the frequency slightly, but also increased the possible gain available to +/-15db to match the rest of the LCR filters.

Photo Dec 19, 5 30 46 PM.jpg

With the stock gyrator filters, these are the maximum gain curves for all frequencies. Notice that the high frequency of the low mid band overlaps with the low frequency of the high mid band. And the High Shelf only provides +/-12.6db of gain.

Stock EQ max gain.png

After swapping all of the EQ components, these are the new resulting curves.

New EQ Filters.png

LCR filter parts value chart. Email me for a Mouser parts list.

NOTE: For the Low Shelf and Low Mid filters, the inductors are in series.
Low Shelf (F3) L C R
51 Hz (+/-15.5 dB) 3040mH n/a 0R
70 Hz (+/-15.5 dB) 2090mH n/a 150R
123 Hz (+/-15.5 dB) 1140mH n/a 360R
177 Hz (+/-15.5 dB) 760mH n/a 435R
363 Hz (+/-15.5 dB) 380mH n/a 530R
Low Mid Bandpass (Q=.75)
118 Hz (+/-15.8 dB) 600mH 3.3uF 221R
160 Hz (+/-15.8 dB) 480mH 2.2uF 300R
230 Hz (+/-15.8 dB) 360mH 1.5uF 374R
336 Hz (+/-15.8 dB) 240mH 1.0uF 402R
632 Hz (+/-15.8 dB) 120mH 0.56uF 487R
High Mid Bandpass (Q=.75)
1.6 kHz (+/-15.8 dB) 39mH 270nF 523R
3.0 kHz (+/-15.8 dB) 18mH 150nF 523R
5.0 kHz (+/-15.8 dB) 12mH 82nF 523R
8.72 kHz (+/-15.8 dB) 5.6mH 56nF 523R
14 kHz (+/-15.8 dB) 5.6mH 22nF 523R
Hi Shelf (F3)
5.12 kHz (+/-15.5 dB) n/a 47nF 470R

Here is a report I generated with more details.

EQ Mod Report

Now for the surprise. The owner asked if it was possible to add a pair of DI’s to the console to be used with guitars, basses, etc, plugged directly in, but with the correct high impedance loading. Since I still have a bunch of PM2000 input PCB’s left over from a previous project, I thought of a way to fashion a pair of 312 style preamps with 1 Meg ohm input impedance, 1/4″ input and a 10db pad, recycling the Tamura GA81720 transformers from the Pgm outputs. I found a spot under the master sections that held the rebuilt circuit board and transformers nicely. A few holes in the surface to mount the jacks, pots and switches. I plugged in a Gibson bass guitar and a Martin acoustic. Both sounded fantastic. I routed the outputs from transformers into channels 17 & 18 of a 28 channel ELCO connector mounted on the rear panel. Channels 1-16 are post fader direct outs using the EA2623 output transformer from Jeff Steiger of Classic Audio Products of Illinois.

Photo Jan 06, 11 17 20 AM.jpg

Photo Jan 06, 9 59 00 PM.jpg

Photo Jan 06, 10 00 28 PM.jpg

 

Photo Jan 08, 2 34 56 PM.jpg

Photo Jan 12, 6 28 40 PM.jpg

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s