You know me, i can't pass up a silly-good deal when i see one. That's how i ended up grabbing the other week, two AKG small-diaphragm condensers (the black P170 and P17) and a Behringer BA19A boundary-microphone. Yeah, i know, i know, but an equivalent total retail cost of ~250eu got quickly whittled down to 120eu (shipped!).
The Perception 150 i've had for a good few years now, and i figured i'd include it here, to see if and how the insides of AKG's lower end of SDC's have evolved over the years.
Starting off with the AKG's - apart from the obvious colour difference, the main changes between the Perception 150 and the later Perception 170 (similar pale blue housing) / P170 (black, pictured) / P17 (ditto) are the attenuator pad being -20dB instead of -10dB on the 150, and the lowest cutout in the side vents being square as opposed to rounded-off. That's something noticeable in the large-diaphragm offerings in the same range (Perception 100 / 200 / 400 versus Perception 120 / 220 / 420 and P120 / P220 / P420).
Moving on to the insides, not THAT drastic of a difference either, main points being the move to (even closer to) full-surface-mount components on the newer ones, the 150 still having the JFET and 1G resistors as through-hole components. Speaking of the insides though, looks to be a pretty standard Schoeps affair, albeit with a couple of quirks. There's a common-mode choke on the XLR output, to help with common-mode and RFI noise rejection. Next, the JFET seems to be powered from the maximum available voltage, without any zener regulation. And lastly, from the power rail supplying the JFET, there's an added filtering inductor before the zener-regulated input to the DC-DC oscillator (that produces the voltage for biasing the condenser capsule.
Nothing particularly noteworthy on the back of the boards, although the JFET bias adjustment trimpots are indeed a welcome touch. The datecode labels are quite telling as well - it would seem this Perception 150 was manufactured in the 14th week of 2007, the P170 in the 50th week of 2015, while the P17 is still newer, having been put together in the 46th week of 2013.
Speaking of labels... The barcode one inside the P170, coincidence or not, would seem to give credence to the rumor of, at the time, AKG having collaborated with Takstar, when designing this line of products. For all we know, they may well have outsourced / subcontracted the entire production to Takstar. That would explain the "TA" prefix there, but i might just be reading too much into it. Coincidence or not though, as i understand, Lewitt Microphones is in fact a joint-venture between some ex-AKG staff and the son of Takstar's CEO, allegedly. This is all of course more-or-less hearsay, but i'll let you decide on your own.
Moving on, there's Behringer's version of the much-better-known Shure Beta 91A boundary-microphone. I snapped this up simply because i had no other boundary-mic, and it's a pretty popular option for using inside of a kick-drum. Might not be very visible in the photo, but it's about palm-sized, but with a reasurring heft to it - spec sheet says 450-odd grams (around-about a pound or so).
The top/front protective grille comes off after removing the one visible screw, and we're greeted by...
Not a helluva lot, really... Not that it's not at least a bit intriguing, though. ~9-10mm electret capsule, cardioid, cute little casing for it, with shielding mesh, and nicely shockmounted in a rubber support; three-wire connection, though. Linkwitz mod, anyone?
And lookie there, even an opamp! May or may not be a genuine Texas Instruments NE5532, even...
Of course, i was completely compelled to reverse engineer the circuit. Overall decent-looking, but not without some concerning design decisions. This IS a Behringer product, after all...
Filtered power rail goes straight to the pad called "MIC+" (white wire), a second wire comes back to "MIC-" (red wire) and a resistor to ground, and there's a ground connection. Since these little electret capsules contain at least a JFET inside, the arrangement of wires / signals is a dead giveaway for this being a source-follower. That then feeds the switchable mid-notch part of the circuit, some more fixed tone-shaping, and then the first half of the opamp gives a gain-boost of about 5x (~14dB) to at least the low frequencies (the highs are rolled back down to unity-gain by C10 & C11), and then goes on to drive the XLR pin 2. The second half of the opamp inverts the previously-mentioned signal, and drives the XLR pin 3.
I also kinda-sorta recreated the board layout, just to help with sanity-checking the connections. Yes, that's a "C0" right next to the switch. My best guess is, someone typo'd that 0 instead of a 9, when naming the components, but oh well...
That being said though... That's a pair of 470 ohm(!!!) resistors, tapping off the phantom power from the XLR signal lines. More often one would see a pair of 2.2kohm, if not even higher values. But if those values are quite so low, what sort of voltage is this whole thing running on?
[Reassembly and connection...]
Five volts...
A piddly, pathetic five volts - a tad under that, even! The voltage on the XLR2/3 lines gets dragged down all the way to 7.6V or so, indicating that this whole thing is sucking almost TWELVE milliamps (of the ~14mA that would be available into a dead-short). But wait a minute, never was the 5532 intended or spec'd to be a rail-to-rail opamp.
[Browsing through datasheet...]
Uuuuuuhum... When powered from 30V (+/-15V), the 5532 "typically" clips about 2V away from its power rails, ie. 26Vpp. Subtract 25V from both sides of that equation, and for a 5V-ish supply, we're left with a near-insignificant 1Vpp of signal, before the opamp clips. Greeeeeeeeat...
Who approved this design, i wonder? Even a lame old LM358 can swing 3.5Vpp before clipping, when powered from 5V. Nevermind that the output stage is asymmetrical and biased very "cold", and has loads of crossover distortion... And a legit 5532 isn't even cheap; i'm surprised they didn't use at least some Nisshinbo (JRC, NJM) part.
Yeah, this will need (and will be getting) a Schoeps-type circuit - just as well that i have some really tiny boards on the way, for exactly such purposes (internal-JFET electret capsules and a board small enough to fit inside an XLR connector shell). I can do without the switchable notch, it's not that big a deal...
No comments:
Post a Comment