RS DIGITAL SPL METER (33-2055) MODIFICATIONSThanksNoah Katz was kind (and trusting) enough to offer up his digital meter up for possible sacrifice. It came in the mail yesterday so opened it up and promptly began exploratory surgery (don't worry Noah, I used anesthetic!).
Looking Under the HoodClick here for the schematic. I traced each component and drew up a a schematic. Several pain-filled hours later, wherein much cursing of RS management could be overheard emenating from my work area, the bulk of the signal path was traced out. A pole frequency analysis determined which capacitors would need to be changed in order to move all poles significantly outside of the passband. I picked 1Hz and 100kHz as targets for these calculations. Before I changed anything, I measured the frequency response of the electronics of the stock meter by removing the microphone element and driving the pads where it was connected with a function generator, coupled through a 10k resistor and a 100uF capacitor. The electronics of the stock digital RS SPL meter rolls off the bass (many poles) somewhere around 28Hz (-3dB), and rolls off the treble (two poles) at 12kHz (-3dB). The "A" weighting setting introducesa pole at 500Hz when engaged, which rolls off the bass beginning at this higher frequency. The frequency response mod as described below improves the frequency response of the electronics for the "C" setting of the A/C switch. After perfoming the mod, direct meter readings should be accurate for bass SPL measurements down to 10Hz or so using the internal mic without using any correction factors. The treble response of the stock meter is another matter. For some reason, the designer of this meter decided to use the same cruddy mic element as used in the analog RS SPL meter, and this is responsible for a major treble roll off somewhere between 10 and 15 kHz. While the designer dispensed with the skanky stone-age BA 301 op-amp found in the analog meter, (s)he inexplicably settled on using the (in some ways even skankier, particularly for this application) LM324 quad op-amp for most of the electronics, and this is the main cause for the high frequency response of the electronics to be down 3 dB @ 17 kHz even after the mod is performed. Equally inexplicably, the single transistor input stage is almost exactly the same as the analog meter. Jeez, I could have done a better design with one brain lobe tied behind my back. You had a whole quad there buddy! Anyway, for these reasons I am disinclined to recommend installing a jack for an external mic mod, and would suggest you purchase the analog meter if you are interested in using the meter itself as a mic preamp. If you have already purchased the digital meter and desire a better high end response than this mod will produce, you could try replacing the op amp with a higher quality quad unit with the same pinout. The specs to look for would benear rail-to-rail input and output, along with reduced supply voltage operation and high slew rate. I tried replacing the LM324 with a TL074, but the rectifier formed by two of the op-amp sections stopped working, so I put the original op-amp back (I installed a 14 pin socket, and so was able to check more than one TL074, but none worked in-circuit). The design biases the rectification section asymetrically and rather low (+1.5V from the '-' terminal) and the TL074 probably just won't work this close to the supply rail. Anyway, the moral of this story is to use a socket so you can easily put the old chip back if the new one doesn't work (and of course that a TL074 won't work without possible further modification to the circuit). Regardless of the high end though, this mod will at least make bass measurements easier (since no correction factors are necessary between 10Hz and 10kHz or so) and should extend any high frequency readings somewhat also. Note also this mod should not affect the calibration of the meter, since only the poles in the circuits have changed, not the ac gain in the passband.
The ModFull speed ahead into the mod! But first you have to get the thing open to work on it:Case Disassembly:
Now that the case is open, look as the component side of the board for the printed component numbers listed below. These designations are printed immedately next to the component they refer to.
Here is a picture of the inside of the meter with the back off. Note the brown disk, which is the wafer switch.
To extend the LF response:SOLDER IN PARALLEL WITH THE OLD VALUE, on the back (non-component side of the pc board). Use electrolytic caps (except where noted), observing polarity:
To extend the HF response:CHANGE (replace, use ceramic or NPO cap):
Here is a picture of the meter after the mod. Note the electrolytics soldered onto the back of the PC board near the microphone silver holder. The SPL meter electronics (including the digital meter itself and the signal at the output jack, but not including the built-in mic capsule) now respond in the following way (regardless of the range switch):
"MAX" response switch position, "C" weighting:
Parting ThoughtsThe digial version of the RS meter has one thing over the analog meter: it reads +/- 10 dB within a single range setting. If it dispensed with the range selector switch and read fractional dBs, I would consider purchasing one, but with the selector switch, I consider the analog meter more useful and the better value. The analog meter also goes higher in terms of frequency response before rolling off, and so makes a better preamp for external microphone capsules.
FAQ SECTIONThis section reserved for future questions regarding this mod... |
