I was approached by a studio owner in Sao Paulo, Brazil, asking to rebuild an M1516 and add a few modifications to it. The owner wanted a great sounding analog console with API style tone. Since it will be used mainly for tracking, he wanted to add transformer based direct outputs to each channel. The M1516 has 4 program summing buses and 4 matrices. We decided to upgrade some of transformers while leaving some stock. I started by swapping all of the electrolytic capacitor in the signal path with Nichicon Muse Bipolar capacitors. The caps in the power supply were replaced with better versions from Nichicon also.
The M1516 came with a combination of NE80200 discrete opamps for the preamp and AN6522 and TA7322P monolithic opamps for EQ and housekeeping. I decided to swap the monolithic opamps for modern versions from Texas Instruments, the OPA2604 and OPA604. The AN6522 opamps in the EQ gyrator had 16v Zener diodes to drop the +/-24v rails down. I removed the Zeners and ran the OPA2604 at +/-24v. Now the stock EQs have better detail and more headroom. The summing opamps is the Program buses are also TA7322P, so I swapped those as well. This dropped the noise floor and improved dynamics. Since the new OPA604 opamps are DIP and the old TA7322P are SIP, I decided to use an adapter PCB from Brown Dog Adapters in Idaho to save time instead of building them from scratch. I was disappointed to find out that these came unassembled. I lost precious time building 40 little assemblies, only to find out that the right angle header pins were too big to fit in the holes of the circuit board on the M1516, by 0.004″. Good thing I have a collection of PCB drill bits. Couldn’t risk using a power drill, so they were all drilled out with my trusty Dremel.
After upgrading the monolithic opamps, I next turned to the transformers. The owner wanted to add 16 direct output transformers, and swap the Program 1 & 2 and Matrix 1 & 2 bus output transformers. After some research I decided on using the API copies made by Jeff Steiger from CAPI in Chicago. The output transformers are EA2503, based on the original design by Ed Anderson of API in the 1960’s. These are 6db hotter that the stock Tamura transformers. The direct output transformers are also API copies, the EA2623. To add the direct output circuit, I was able to hijack a pair of unused pins on an existing circuit board connector. Yamaha even left 2 empty pins in the holes. They were crimped, so I had to solder to them. But it made for a factory looking modification. I swapped the output transformers pretty easy. Same mounts and everything. The 16 direct output transformers needed to be mounted somewhere. The best spot seemed to be on the steel plate covering the stock transformers. A few holes and rivets later and they almost look stock. Those were sourced post fader and plumbed into an ELCO 16 channel panel mount connector, stolen from an ADAT.
The last step was to upgrade the Low Mid and High Mid EQ filters in the first eight channels. The stock gyrator circuit sounded better with the OPA2604 upgrade, but it doesn’t compare to the API style inductor based filters. It’s a common modification in PM2000 EQ circuits, so parts values for common frequencies were easy to come by. I chose shielded inductors, Poly film capacitors and high precision resistors to build the filter tanks on Radio Shack (RiP) perfboard. After removing the stock gyrator components, I cut apart the filter sets and mounted them on the input channel boards. The were easy to wire and fit in the space allotted with room to spare.
After 4 months of periodic work, I finally reached a point where I could apply power again. First I verified the output voltages on the power supply. installed an few channels and made sure nothing smoked. After some positive tests, I decided to install the remaining channels in groups of 4, test between each group. To my surprise, everything worked with the exception of one bad solder joint on a capacitor. I ran a battery of tests including bandwidth, distortion, phase and good old fashioned listening to music. The most impressive test was comparing the stock EQ to the LCR EQ. I used Smaart to measure the transfer functions of each frequency bank, with maximum boost applied. The LCR filters are wider, and distributed further apart. Most intriguing was the increase in signal level by switching to a passive filter. The first graph shows the stock gyrator filters, the second shows all the LCR filters, the third graph shows the High Mid Gyrator filter in red and the High Mid LCR filters in green.