#95 Presonus StudioLive Series III 32R teardown

I know, i know, i'm breaking my own vow from a while ago, to forever(?) boycott Presonus hardware, but... Well, on a technicality, they're Fender nowadays, so there's that. Either way, more or less on a whim, and at the available price, i really couldn't say no to this deal. With the introduction of the updated "SE" line, Thomann had some severe discounts on the previous-generation SKU's, and were seemingly eager to get rid of their existing stock. 

I had been daydreaming of one or several MOTU Stage B16's, but those seem to be about as rare as hens' teeth, and fetch 4-figure prices "even" on the used market. Then i looked for Presonus gear - also on the pricey end of the scale, but just for kicks, i took a quick peek at Thomann and when i saw this previous-gen 32-input, 18-output gem at 47%(!!!) off, i had to have it. Discounted to a bit under 1000e; ex.VAT came out to a (relatively) measly 730e. I dare you to argue this isn't a steal...

 

 

But let's pop the hood on this one, shall we?

 

Not super-mega cramped, but that's never a bad thing. Quite a bunch of test points scattered around, and a nice touch - all very clearly labeled.

The mains power supply is an actual MeanWell unit - LRS-75-12, rated for 12V 6A. Likely way overkill for this use (the marking near the IEC inlet "only" mentions 35W), but that only means this'll run that much cooler (or, well, less hot). I can just about see a Rubycon and a Nichicon near the output, and the primary side electrolytic is also a Nichicon (or possibly a Lelon, but even that's not the end of the world, on the primary), so there's some peace of mind there. But otherwise...

 

Ooooooh dear... CapXon, Fujicon, Aishi and... Hyecap? And also Nichicon VR's... *sigh...* For longevity, at least a part of me would prefer to fully re-cap this whole thing, but... Right now, i don't quite have the inclination (plus i'd need to restock on some cap values), so this'll live as it is for now, until it starts showing signs of fatigue. Full disclosure, i got this with the main intention of renting it out, in a rack case, as a compact and portable solution for on-location recording and mixing - even more flexible than my MOTU 8M rack, but useable as an extension to it, if ever needed. Hooray for standards like AVB, am i right?

Actually nevermind; had to order a bunch of other stuff from Digikey anyway, so i figured i might as well fill my stash of electrolytics for this too. You can't really put a price on peace of mind though, can you, at the end of the day? Especially if i want to rent this out; wouldn't want it kicking the bucket on a gig... Even if the datecodes on the stock caps point to late-2024 / early 2025 manufacture.

 

 

Before removing the "top layer" of boards, let's take a closer look at them. Starting at the front with the inputs 1-16, we can see the touted "Xmax preamps". Likely very similar to the ones in the VSL1818, albeit not using analog potentiometers for the gain adjustment, obviously. Pretty usual Cohen topology (two PNP's followed by opamps).

 

The transistors are 2N4403's, while the opamps are ST Microelectronics MC33079 low-noise quads. In addition to those, there are also three Analog Devices AD5204 4-channel 10k linear digitally-controlled potentiometers.

 

More of the same on the top board on the back, "just" eight more inputs like the twelve on the front board.

 

 With those two out of the way, we can have a better view of the quite sprawling main board.

 

One detail that immediately caught my eye was this empty footprint. The "WiFi SPI" marking near the non-populated header near it is quite telling - an option for a WiFi module. 

 

 Ok, diving into the big silicon - the main processor is an NXP LPC4350FET256 ARM Cortex-M4 microcontroller; that's assisted by an Alliance AS4C16M16SA-7TCN 256Mbit / 32Mbyte SDRAM, while the firmware is contained in a Spansion / Infineon S25FL128SA 128Mbit / 16Mbyte SPI flash chip.

The AVB connection is handled by an Analog Devices ADIN1300 Gigabit Ethernet PHY, while the "control" Ethernet connection goes through a Microchip KSZ8081 PHY.

 The heavy lifting is provided by a pair of Analog Devices Sharc ADSP21489 (biggest / fastest / most capable in their respective range). 

 The ADC's are four Asahi Kasei Microsystems AK5558VN's - 8-channels, (up to) 32-bit 768kHz(!!!) sampling. Three of them are along the front panel.

And the last one is in the rear corner. Next to it are the three AKM AK4458VN DAC's.

Moving on to the power supplies - a Richtek RT8293 handles the 3.3V rail (thanks to the generously marked test points).

Two more RT8293's are in charge of the 1V supply to the Xmos processor, and the 1.1V supply to the two DSP's.

Next up, the phantom power is provided by a Texas Instruments LM3478 boost controller, while the +15V rail is handled by a Richtek RT8509 boost converter.

Over to the side, an Analog Devices LT8609 is in charge of the -15V rail.

   

Three OnSemi LM317's provide a clean 5V to the analog stage of each set of converters.

And that about wraps this one up too. Already started the recapping process (which is a pain in the rear end, especially for the power supply caps with one grounded leg). "Only" takes over a hundred (if not several) capacitors, but all in the name of longevity and reliability, am i right?