#3 MOTU 8Pre teardown & gain adjustment pot replacement

On the latest drum tracking session i surprisingly discovered that channel 8 on the MOTU 8pre in my recording rig was noisy and useful-signal-less. This same unit had a similar issue with channel 6 when i bought it (as faulty, with a shot Firewire chip).

Fortunately, when i was about to take care of ch6, i got two replacement pots. When investigating the issue, i had read it's a relatively common occurrence, so i figured that as long as i'm contacting the local MOTU distributor, might as well get an extra / spare one, just in case (especially since i had two 8Pres at the time).

I've always liked MOTU cases - nice, sturdy, reasonably thick cast aluminium. I also appreciate the cleverness - the top and bottom halves are the exact same part. This particular unit is a slightly newer revision of the Firewire version, with a switchmode power supply. I had another one, which had two PCB-mounted linear (steel, chunky) transformers. Also, of note is the cutout in the PCB on the left edge (in the region of ch1), which lines up with the slots in the casing. Another ingeniously simple solution, in order to offer some passive cooling to the insides, the "exhaust" being on the top right of the enclosure, above the power supply.

Since i've had this for a while, "of course" i had already re-capped it soon after acquiring it. Somehow, the stock "Meritek" caps didn't inspire all that much confidence, y'know?

Power supply got a Rubycon WXA 22u/400v and a YXF 47u/50v on the primary, and a pair of Panasonic FC 1200u/6.3v and FK 2200u/16v respectively. The thing puts out five power rails, +/-8v, 5.5v, 3.8v and 3.3v. The mains input filtering is comprehensive (fuse, NTC thermistor, two X2 caps, common-mode choke, two Y2 caps, all mains-rated with requisite markings). Even the snubber cap across the transformer primary is a Y2 unit.

The rest of the unit got an assortment of Panasonic FR and Rubycon ZL caps, plus a few Chemi-Con SMG and a lone Sanyo ME-WX. Nope, couldn't be bothered to mess with the surface-mount electrolytics though.

As long as i have this open, might as well do a little run-down of the silicon as well:

• The brains of the operation is a Philips LPC2104B. I'm guessing this has more of a supervisory / peripheral-control role, than anything else (driving the LEDs on the front panel mostly).
• The heavy lifting (protocol / format conversion and audio / data streaming) is, i suspect, handled by the Xilinx Spartan XC3S200 FPGA.
• The Firewire interfacing is done via a "classic" Texas Instruments TSB41AB2.
• Two Analog Devices ADM709 voltage supervisor chips keep the processors in reset until the power rails stabilize. As an added bonus, we have a nice little factory bodge wire in the area. 
• An SST (now Microchip) SST25VF040 4Mbit SPI flash chip holds the contents of the firmware.
• The clocking seems to be provided mainly by a TI TLC2933A VCO / PLL.

• All(?) the opamps seem to be JRC NJM4580 duals.
• Taiwan Semiconductor DB107G bridge rectifiers, used as clamping diodes (to the opamp power rails), for input protection. Clever space-saving board layout though, "piggybacking" these on top of the preamp-opamps.

• Four Asahi Kasei Semiconductor AK5385AVS stereo ADCs handle the eight inputs, and AK4382AT stereo DACs drive the two analog outputs and the headphone outs, respectively.

• Apart from those already mentioned, there are a few linear voltage regulators for the logic chips, as well as a little boost converter for creating the +48v phantom-power for condenser microphones.
• The main component in the mains power supply is a Fairchild Semiconductor (now part of On Semiconductor) FSCM0765R all-in-one flyback controller / switcher IC.
• Rectification on the secondary is via four SR360 3A 60v Schottky diodes.

Now that that's done with, the ch8 gain pot "finally" got replaced, and all is well, good as new (or even better, in fact) and all that.