#43 Focusrite Scarlett 18i6 teardown

As it turns out, i hadn't done (or at least documented) a teardown of this little guy, even though i've been using it for the last two years or so. Even though, as it turns out, i bought this a few months after starting this blog. Either way, i just put it up for sale (due to upgrading to the Scarlett 18i20 i inspected the other week). So i figured i might as well snap a few photos before it goes to its next owner (whomever that may be).

This is one model that apparently hadn't cut the mustard (enough) to make it to the Gen2 refresh / update of the Scarlett line. Although i kinda wonder if feedback from the Saffire product line prompted the shift of the input & output configuration.

That being said, even from the outside one can begin to see a certain amount of "family resemblance" yet again - the casing is virtually identical to its older cousin, the Saffire Pro 24.

But on to the good stuff - the innards. I hadn't recalled, but apparently i had indeed recapped the power supply section. I admit, i guess i have a thing for preventive maintenance, and at least i can pass it on with a clear conscience that it WILL last. The rest of the (far less stressed) electrolytics are a mix of Jamicon (the medium and bigger-sized ones) and CapXon (the little blue ones), plus the few "anonymous" SMD ones.

Speaking of the power supply - deja-vu, anyone? It's a bit of a hybrid - the layout's virtually identical to the one in the Saffire Pro 24, even down to the layout, but using the bill-of-materials of the Scarlett 18i8. Makes perfect sense, though - why reinvent the wheel? So we have the same set of two LM2672, LM2670 and L7805.

The Xmos core is supplied by the same OnSemi NCP1529.

Speaking of which... On a side-note, i hate these assinine AND well-obscured marking-part-number associations that Xmos use. Both of which they've changed at least once or twice already, in their relatively short existence. At least according to this document (for which you "even" have to accept their website's cookies before opening it actually works), the "US1032L2C5" marking translates to XS1-L16A-128-QF124-C10. Because logic...

In the vicinity there's the AT25DF041A SPI flash chip containing the firmware (which i took the liberty of reading out, before reassembly - just in case), as well as the SMSC (now Microchip) USB3318C interface chip, plus an OnSemi NCP1117 fixed 1.8V linear regulator.

Near the 13MHz quartz crystal, we have the already familiar Cirrus Logic CS2300 PLL chip, providing the main system clock. 

Yet more familiar sights - a quad of Cirrus Logic CS4272's handling the A-D and D-A conversion. Just too bad one of them doesn't have its DAC side used for anything. An extra analog stereo output would've been quite welcome.

A bit of a repeat on the analog side as well. Mostly NJM4565 opamps, with the exception of the NJM2122 for the mic preamp duties, the NJM4556 for the headphone output, and the TL072C's for the instrument inputs.

Speaking of family resemblance - we have a bit of a repeat of the 18i20 "incident", reusing front panel designs from the Saffire line. Well, i suppose there's nothing wrong with offering a level of familiarity for customers loyal to the brand, upgrading to the newer USB units.

Aaaaaand that's about that, really. Now, fingers crossed for finding a buyer ASAP - bills won't pay themselves, or go away...