Announcement

Collapse
No announcement yet.

TC Electronic Studio Konnekt 48 revival

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    TC Electronic Studio Konnekt 48 revival

    Saw this baby on eBay a couple weeks back, on auction, faulty. Boots up, but no Firewire activity. Reasonably common issue - routinely, the TI TSB41AB2 PHY-chip's toast.

    While fully-working ones hover around 500eu (~550$-ish), i snatched this one for 184e shipped I had been drooling after one of these gems ever since i started considering getting an audio interface.

    Now, before getting on to the actual fault - this thing is CHOCK FULL of stuff, in addition to the abundant I/O options, "even" compared to my trusty Saffire Pro 40. Hell, it even has TWO friggin' transformers on the actual mainboard :p

    Cap-wise, it's a literal hodgepodge, not really sure what to make of it. Hermei in the actual mains power supply, with possibly a Teapo on the primary (it's heatshrunk, but with a Teapo vent design). On the mainboard, there's one Panasonic NHG, a number of Nichicon HD and a pair of PS on the "secondary" power supplies plus a few 7mm-high ones scattered about, one fat Elna on each mic channel, and an array of Hitano and Elite pretty much everywhere else (signal coupling, i'd reckon).

    The PCB layout / routing is indeed an engineering work of art. As the schematics indicate as well, there are test pads for literally EVERY friggin' signal. And i mean EVERY last ONE of them, i'm sure I've lost count how many different ground signals/planes there are; analog/digital grounds are isolated with a resistor in parallel with a few diodes. Damn near brings a tear to the eye

    On to the actual issue and whatnot - upon the removal of the case, as well as the digital daughterboard, while investigating the area with the PHY, i noticed what looked like (uncleaned) flux residue on a ~1in radius around the chip, as well as most-definitely-NOT-factory soldering on the pins of the chip. That, plus the pad missing from underneith pin 16 (N/C, so no worries there) was a pretty dead giveaway that the chip had actually been replaced already.

    Juuuuuust to be on the safe side, i extracted it, and replaced it with a brand-spankin' new one i had received but a couple days before the Konnekt itself. Soldered it back in, powered it up, plugged in the cable... (still) nothing

    Well, sod it...

    After first starting to search for some tech support contact data from TC, i noticed a code name/number on the main board, right underneith the silkscreened TC Electronic logo. What the hell, let's google it And lo and behold...

    [Edit: link broken, see attachment]

    Now you don't see THAT every day... When's the last time the manufacturer of such a complex piece of gear had ANY sort of service data available?

    Despite that, and pouring over the schematic for several hours, i took a couple days off, held a wee (late) birthday dinner and so on. Finally, tonight, i decided to squint at the schematic a bit more. Browsed to the sheet with the PHY... Saw the input supply voltage had an interesting designator, "5V_PHY", which then went into a 34063 buck-converter, which in turn, supplies the PHY with the 3.3v it needs. Ctrl+F, typed "5v_phy", first result was on the PHY page, then 3 on the block-diagram page, and then on the power supply sheet.

    Hey, this thing's supplied from a totally isolated winding on one of the two "secondary" transformers. Hmm... Multimeter, diode test: the rectifier schottky, forward-biased, checks out fine (0.16v, for a 2A 40v diode - seems legit); reverse-biased, not so fine - around about 1v. That can't be right.

    Unsoldered the diode, red lead on where its positive end was, black lead on the ground-lead of the associated cap - still 1v. Something's definitely fishy here...

    Just for the hell of it, i started desoldering the pins on one side of the 34063, starting with CMP, simply 'cause it was the left-most one, on the side facing me. Second pin lifted was the VCC. Measured again, 1.4v - HELLO... Lifted the next two pins as well (ISNS and DRC), measured again - voltage drop rose above 2v in less than a second, and then OL - GOTCHA!

    As a sanity-check, with the other 4 pins still untouched (including the GND pin), i measured the 4 pins i had lifted. VCC-GND shows roughly 0.96v, and ISNS-GND is around 1.4, neither of which, i'm quite certain, should be anywhere near that.

    It's 6am right now, and i just wanted to get this "off my chest", as it were. Some teardown (or, actually, more like "(already) torn down) pics to follow tomorrow; already placed an order for a few 34063's, so... Now we wait
    Attached Files
    Last edited by piernov; 09-21-2020, 09:32 AM.
    Khron's Cave - Electronics - Audio - Teardowns - Mods - Repairs - Projects - Music - Rants - Shenanigans

    #2
    Re: TC Electronic Studio Konnekt 48 revival

    And, as promised, "pics or it didn't happen"

    In the last two photos, you can see that small SOIC8 package in the top-left-ish area - that's the (most likely) culprit, the 34063.

    Either way, like i said - chock full of stuff, top AND bottom
    Attached Files
    Khron's Cave - Electronics - Audio - Teardowns - Mods - Repairs - Projects - Music - Rants - Shenanigans

    Comment


      #3
      Re: TC Electronic Studio Konnekt 48 revival

      those chips are known for failing,
      i used to change the bigger version in the atari jaguar consoles a lot.

      Comment


        #4
        Re: TC Electronic Studio Konnekt 48 revival

        Well that's annoying...

        While waiting for the replacement 34063, i decided to bodge in an LM1117-3.3 linear regulator. TSB chip only takes 60-70mA, so what the hell...

        Trouble is, still no firewire activity
        Khron's Cave - Electronics - Audio - Teardowns - Mods - Repairs - Projects - Music - Rants - Shenanigans

        Comment


          #5
          Re: TC Electronic Studio Konnekt 48 revival

          check the crystal and the reset signal.

          funny how they opto-isolated the reset signal!

          Comment


            #6
            Re: TC Electronic Studio Konnekt 48 revival

            Did some poking around with the multimeter before firing up the scope.

            For some reason, the two "pull-down" resistors on the B-pairs seem to show a far lower resistance than the 5.11k that's on the schematic. R570 (port 0, upper one on schematic) shows me 62.6ohm, and R571 (port 1, lower) measures as 57.1ohm, with probes both ways, and that's odd...

            Termination resistors check out fine. Crystal seems to be working ok as well.
            Khron's Cave - Electronics - Audio - Teardowns - Mods - Repairs - Projects - Music - Rants - Shenanigans

            Comment


              #7
              Re: TC Electronic Studio Konnekt 48 revival

              As the old saying goes, "even a kick in the a*s is a step forward...

              I ended up shoving in a(nother) new TSB chip, and now, after it finishes booting, the Firewire led blinks.

              Manual says "Flashing: Uploading firmware, hardware error or FireWire communication error."
              Khron's Cave - Electronics - Audio - Teardowns - Mods - Repairs - Projects - Music - Rants - Shenanigans

              Comment


                #8
                Re: TC Electronic Studio Konnekt 48 revival

                Y'know, on i-don't-know-how-manieth thought (definitely not second), i'm kinda-sorta starting to doubt the genuinity (and, thus, the proper functioning) of these two new TSB chips i had. Granted, they were from eBay, but from the UK (not the Far East).

                Just for the hell of it, i just placed an order for 10pcs (future-proofing) from Farnell - i seriously doubt fake stuff can get into their stocks, in this class of components (or at least i damn sure hope so).

                Just for illustration purposes, here's a snapshot from inside a Mackie Onyx Satellite. On the left there's its stock TSB, as installed from the factory, and on the right, it's one of the two chips i had. The markings look nothing alike. Yeah, i know, TI may well have more than one fab, but surely they would've at least "standardized" their marking layout, i'd reckon...
                Attached Files
                Khron's Cave - Electronics - Audio - Teardowns - Mods - Repairs - Projects - Music - Rants - Shenanigans

                Comment


                  #9
                  Re: TC Electronic Studio Konnekt 48 revival

                  tbh the one on the right looks more like a TI chip!!!

                  where did you get them exactly?

                  Comment


                    #10
                    Re: TC Electronic Studio Konnekt 48 revival

                    Does it? I got'em from a british seller on ebay.co.uk, "swle2000".

                    And for what it's worth, i've got a pci FW card (has some Compaq label on the back), and both TI chips have markings that resemble the chip on the left, which would lead me to think there are (far) greater odds of them being legit.

                    Either way, i just paid for that Farnell order, and the site said they had 124 chips in UK stock, so... Hopefully, by the weekend, or early next week, we'll see exactly what's what
                    Attached Files
                    Last edited by Khron; 01-27-2015, 04:49 AM.
                    Khron's Cave - Electronics - Audio - Teardowns - Mods - Repairs - Projects - Music - Rants - Shenanigans

                    Comment


                      #11
                      Re: TC Electronic Studio Konnekt 48 revival

                      the u.k. seller probably got them from aliexpress

                      Comment


                        #12
                        Re: TC Electronic Studio Konnekt 48 revival

                        Who knows... Either way, i just got the UPS tracking number from Farnell less than an hour ago, so they're on their way

                        I just had a "brain-fart" just now, though, looking at the schematic. Next time i pull the TSB off the board, i should re-measure all the resistors and ancillaries, around the FW ports - if some of those TVS's right on the data lines crapped out & shorted to ground, i can keep changing TSB's until the cows come home...
                        Khron's Cave - Electronics - Audio - Teardowns - Mods - Repairs - Projects - Music - Rants - Shenanigans

                        Comment


                          #13
                          Re: TC Electronic Studio Konnekt 48 revival

                          Just for kicks, since the Farnell package is only getting delivered tomorrow, I decided to probe the resistance to ground, straight from the FW connectors' pins.

                          Pin 1 = Vcc; Pin 2 = Gnd; Pin 3 = TPB-; Pin 4 = TPB+; Pin 5 = TPA-; Pin 6 = TPA+

                          Port 1, we have 40k, 0.1R, 117R, 7.5R, 2.3R, 113R
                          Port 2: 40k, 0.1R, 112R, 1.75R, 2R, 112R

                          Worst case tomorrow, i'll re-measure these after removing the current chip from the board.
                          Khron's Cave - Electronics - Audio - Teardowns - Mods - Repairs - Projects - Music - Rants - Shenanigans

                          Comment


                            #14
                            Re: TC Electronic Studio Konnekt 48 revival

                            LONG day, people... Gah...

                            Anyway, good news - both the 34063's and the TSB's from Farnell got delivered today

                            Had a whole bunch of work though, so i really didn't have time to tinker with this again until a couple hours ago (it's currently 3am here; i started around 1-something)

                            Step one: remove TSB chip
                            Step two: re-measure the stuff in the previous post, with the chip out

                            Surprise-surprise(?) - readings showed exactly the same, so there could be only one (further) culprit: the TVS's right on the data lines. Lifted the legs on the first one, re-measured - TA-DAAA, all OL on all four data pins Same story on the 2nd port.

                            Step three: replace 34063 - here's where things got a bit... ehrm... "hairy", let's say.

                            Due to apparently forgetting the old adage, "haste makes waste", i proceeded to use a small flathead screwdriver to pry off the old chip, off of the strip of epoxy(?) glue that was holding it to the board. And wouldn't you know it, good ol' Mr. Murphy never sleeps - as the chip suddenly popped off the board, the screwdriver slipped and pretty much "shaved" off two ceramics from underneith the main Dice II chip - wonderful

                            An 0603 100n (one end-cap came off) and a 1206 10u. The small one wasn't a big deal, the end-cap on the ground side just popped off, and the 3v3-side pad partially came off the board, but was still attached to the trace leading to it.

                            The bigger cap was more "interesting" - ground side ok, but the 3v3 side went straight to a via right next to the pad. The pad had totally come off the board, together with about half the circumference of the via. With some patience, and the careful maneuvering of the tip of a box-cutter knife, i managed to scrape off the silkscreen and soldermask off of the part of the via pad that was still left on the board, and by some "miracle" (and a bunch of flux) managed to stick some solder onto it. Set the big cap a bit crooked compared to its original position, but the solder blob on the end reached the via and seems to hold

                            After all this, "finally" got around to sticking on the new 34063. After scraping off some of the glue still left on the board, all went smoothly. A quick TSB-less power test showed it outputting 3.24v - i can live with that, and well within the spec for the TSB (3.0v min ).

                            Step four: install fresh TSB chip

                            Stubbornly, i chose to install it just like before, "manually", with buckets of flux and a 0.3mm conical tip on the soldering iron. A few wicked-off solder-bridges later, all looked good enough for a test-run.

                            Powered on, boot logo looked fine, no firewire led at all - fair enough, i hadn't plugged the cable in yet. Powered off, stuck FW cable in, powered on - new hardware found TC Near device etc.

                            Oh yeah... Oh yeah... OH YEAH!..

                            TC Near Control Panel works fine, "feels" the master volume control (animated slider)... Happy camper over here

                            Pics in the next post
                            Khron's Cave - Electronics - Audio - Teardowns - Mods - Repairs - Projects - Music - Rants - Shenanigans

                            Comment


                              #15
                              Re: TC Electronic Studio Konnekt 48 revival

                              Photo 1: Brand new chips Note the style of the markings on the TSB chip. Font is indeed more "blocky", but the style resembles the ones on the boards "demonstrated" a few posts earlier closer than the ones on the chips from eBay, imho...

                              Photo 2: The 1206 10u ceramic after re-bodge. Note the missing pad from the near end (as described in the previous post), as well as the "funnel" of solder leading down to the via it used to be connected to, via copper.

                              Photo 3: new 34063 in its new home.

                              Photo 4: the two (former) TVS's, with legs cut off and pads wicked.

                              Before i'll reinstall the board into the case, i WILL "drown" the two "hurt" caps in some cyanoacrylate, so they won't move or anything. That is, unless cyanoacrylate is known to react unfavourably with solder or something... Opinions?
                              Attached Files
                              Last edited by Khron; 01-29-2015, 07:54 PM.
                              Khron's Cave - Electronics - Audio - Teardowns - Mods - Repairs - Projects - Music - Rants - Shenanigans

                              Comment


                                #16
                                Re: TC Electronic Studio Konnekt 48 revival

                                so you doubled the workload by using the wrong tools!

                                hot air is king, and the resin - twist the heated part with longnose pliers to break the bond.

                                Comment


                                  #17
                                  Re: TC Electronic Studio Konnekt 48 revival

                                  Originally posted by stj View Post
                                  so you doubled the workload by using the wrong tools!

                                  hot air is king
                                  That was my first reaction when i heard about the booboo.
                                  Originally posted by PeteS in CA
                                  Remember that by the time consequences of a short-sighted decision are experienced, the idiot who made the bad decision may have already been promoted or moved on to a better job at another company.
                                  A working TV? How boring!

                                  Comment


                                    #18
                                    Re: TC Electronic Studio Konnekt 48 revival

                                    Further developments...

                                    Today, after a summary re-cap (mostly the power supply side, and the outputs), i reassembled the whole thing, and hooked it up to my laptop & monitors. Its control software saw it just fine, so i started playing some music, but... all silent.

                                    Took it apart again, checked some supply voltages, no issues there. But then i turned on my scope and started poking around on the digital side, only to discover that the LRCK signal going to the front pair of ADC's and the first two DACs was absent :shock: On Page 2 (top left) you'll notice it goes through R624 - i probed before it as well, and the signal's still nowhere to be found...

                                    In the schematic, it's the LRCLK_ADDA_G signal, coming straight out of the DICE chip. The other similar signal, LRCLK_ADDA_S is present (this one goes to the 8ch DAC for outputs 5-12, and the line input daughterboard).

                                    Now, since they're all supposed to work at the same sampling clock, i've got a good mind to solder in a wire jumper from the S signal over to the G. What's the worst thing that could happen?

                                    Well, ok, maaaaaaybe some fanout issue, but... Well, any reason why i wouldn't give it a shot?
                                    Khron's Cave - Electronics - Audio - Teardowns - Mods - Repairs - Projects - Music - Rants - Shenanigans

                                    Comment


                                      #19
                                      Re: TC Electronic Studio Konnekt 48 revival

                                      Well, i figured i'd go for it, what the hell...

                                      First try: cut a short piece of a resistor leg, bent it into shape (fortunately, the R624 for the lacking signal, and R619 with the working signal, only have one other 0603 resistor in-between), and soldered it on the chip side, between the two resistors. Powered the thing up - LRCK works, but had like 0.5Vpp.

                                      Second try: scraped the soldermask off the trace between R624 and the DICE chip, and then cut the trace. Just to make sure the cut was ok, i probed the two "ends" of the trace, showing about 55Kohms. R624 end to ground, same reading; chip end to ground, around 100R - that's WAY too low. Either way, reattached the jumper wire, powered the thing up, and...

                                      Ta-daaa! We can haz LRCK where before we had none, as well as visible (on the scope) music right on the output XLR legs
                                      Khron's Cave - Electronics - Audio - Teardowns - Mods - Repairs - Projects - Music - Rants - Shenanigans

                                      Comment


                                        #20
                                        Re: TC Electronic Studio Konnekt 48 revival

                                        you could buffer it with a cmos buffer.
                                        if you double-up, make sure there arent 2 sets of pullup/down resistorsetc on it.

                                        Comment

                                        Working...
                                        X