Useful Modifications to the RadioShack Analog SPL Meter
|Useful Modifications to the RadioShack Analog SPL Meter|
29 March 2002
So you think a review of a DIY project involving about $12 worth of parts for a RadioShack SPL Meter listing for just $40 is not esoteric enough for StereoTimes?
It's easy for audiophiles to be fooled by inaccurate bass response. Some may classify a given speaker's bass response as "tight as a drum," when, in fact, measurements reveal a significant roll off beginning at 80 Hz or so. Other speakers may be praised for their subjectively rich and powerful deep bass response, but the meter shows a peak of 5 to 10 dB between 80 and 120 Hz, with a slow roll off below 60 Hz or so.
For some listeners, an overall downward slope of more than 10 dB in frequency response from the low bass to the high treble sounds authoritative, accurate, and pleasing, particularly on large-scale classical works, and especially when combined with a midrange suck-out, which adds spaciousness. Admittedly it takes extraordinary air-moving ability in the bass to be able to both produce this kind of bass and play at high volumes, usually meaning a large enclosure or bass towers. Anyone remember Robert Harley's praise for the Genesis II.5 and wonder how the measured frequency response (see Stereophile, Volume 18, No. 1, January 1995, page 88) of such a highly rated speaker could slope down 15 dB from 20 Hz to 10 kHz? Harley justified his admiration for this type of balance by saying: "The response is smooth and flat, but with an overall tendency toward an uptilted bass and a downtilted treble [plus a measured suck-out along this slope of 5 dB between 400 Hz and 2.5 kHz]. Loudspeakers that measure flat tend to be too bright, in my experience. The curve could be summed up as 'flat with lots of bass,' which corresponds to my overall impression of the II.5."
Audiophiles who really want to get the best bass (and other frequency) response from their systems and who want to train their ears to recognize good versus anomalous frequency response need a fairly accurate and convenient way to measure the frequency response of their audio systems in their listening room from their listening position.
It's even more important to me since my Legacy Audio Whispers are designed to need electronic low-frequency equalization for flat bass response. One convenient measuring device that some audiophiles, including me, have been using for years is the venerable RadioShack Analog-Display Sound-Level Meter, Catalog Number 33-2050.
But is the RS meter accurate enough to use for this purpose? Here there is great debate documented on the Internet--just use "Radio Shack SPL meter" as an exact phrase search term in Google'sAdvanced Search mode to get a feel for the controversy. Few seem to dispute the basic accuracy of the meter from mid-bass to mid-treble, and some sources claim even better performance. For example, one test compared the RadioShack to two B&K (the calibration microphone folks) reference instruments over the 125 Hz to 20 kHz range and the RadioShack held its own quite nicely over this range. At least one source, however, claims the meter is wildly inaccurate above 1 kHz (scroll down to near the end). Others mention "correction factors" which should be applied, but there is some disagreement about the size of such correction factors, as well. Compare the values in the above link with the ones mentioned in the Web instructions for the modification under review.
The audiophile literature I respect, such as Robert E. Greene's comments in The Absolute Sound, suggests that while it takes considerable expertise to correlate subjective listening tests with measured results in the midrange and treble, in the bass, what you measure with a good meter is basically what you hear. Thus, using a good meter is a much quicker and more accurate way to determine the flatness of bass response than listening to a wide variety of recorded performances, especially since the relative weight and extension of bass response varies considerably even among otherwise fine recordings.
My primary focus in using the RadioShack meter has usually been to measure the bass response of the system, given particular loudspeaker placement and listener position. Unfortunately, this is the area where there is most disagreement about the accuracy of the inexpensive RadioShack meter. While I have always thought I heard good correspondence between the RadioShack meter readings and my subjective impression of bass weight from a system compared to my aural memory of live unamplified bass in orchestral concert venues, others obviously disagree.
Thus I read with great interest Eric Wallin's Internet discussion ofmodifications he developed for the RadioShack meter which he claims make the bass response ruler flat down to a very few Hz -- certainly low enough for complete confidence in main loudspeaker and subwoofer setup. He also describes modifications to make the meter response very flat out to the limits of hearing and beyond.
This is the only DIY project I have seen documented on the Web for correcting the meter's supposed high and low frequency deficiencies. The low-frequency modifications seemed simple enough to perform, even for a "half-thumbs" (as opposed to "all thumbs") solder-slinger like me.
I decided not to attempt most of the high-frequency modifications (I did replace capacitor C12). They are a bit more complex, involving replacing the microphone element in the meter, and I have never used the meter to adjust high frequency response at the listening position anyway.
The low-frequency modification only involved replacing eight capacitors with new ones of different values. And, if I didn't like the results, what would I lose? Only an hour of two of my time and $52 or so for the parts and a replacement meter. I decided to proceed.
Cutting to the chase, the low-frequency mod is, in my opinion, quite worthwhile. Quite.
I won't belabor the modification details. They are available at the referenced link and are very detailed and accurate, as such things go. So is the description of the more skittish operation of the meter following the mod. The only modification details I would add are:
Unless your spatial relations are a lot better than mine, I strongly recommend removing the stock capacitors and replacing them with the new ones on the top side of the circuit board. This may take longer, but it is easier to see what you are doing this way and easier to be sure you are connecting the new capacitors into the proper holes in the circuit board. The top of the board is nicely labeled with the capacitor numbers referred to in the instructions and plus and minus polarity markings which match the plus and minus markings on the caps.
It was difficult for me to determine which board connections to desolder when looking at the board from the bottom -- there are a lot of connections spaced closely together and nothing is labeled. Unfortunately, all soldering and desoldering requires working on the bottom of the board. Once I determined which connection probably needed to be desoldered next by repeatedly comparing the top and bottom of the circuit board, I used a 30-watt soldering iron together with desoldering tape to remove the solder from that connection on the bottom of the board. Once that hole was clean of solder, the wire lead of the stock capacitor could be pulled out of the hole from the top of the board. Once both leads of that capacitor were free, that cap could be removed and replaced with the new value.
When inserting the wire leads of the new capacitors into the circuit board holes, be careful to match the labeled plus and minus polarity on the cap with the plus and minus labels on the board holes for that cap.
Soldering the new caps into place was easier than desoldering the old ones since I could see the new capacitor's lead wires poking through the board holes when looking at the bottom of the board and thus knew at a glance exactly where to apply heat and solder.
Working carefully, I took almost two hours to complete the project. But then, as I said, I am somewhat ham-handed at this sort of thing, and that time included correcting one cold-solder joint I located after the meter didn't work at all following my initial reassembly.
The new caps are somewhat larger than the stock ones and I had a bit of trouble getting the case back together because of lack of clearance of the new parts on the top side of the circuit board. I corrected this by carefully bending the wire leads on a couple of the new caps so the caps leaned over a bit allowing the case to again fit around the now-more-stuffed board.
Meter Set Up
I have used the RadioShack meter the same way for years. The bottom of the meter is threaded for a standard camera adaptor screw used on camera tripods. I attach the meter to the camera head of a Vivitar camera tripod so that the meter is centered on and covers the camera head and is positioned as far forward on the camera head as possible so as to place the microphone in a maximally free-air position.
I position the meter in the listening chair by removing the cushions, placing an MDF shelf from an old Target equipment rack on the seat of the chair, setting the tripod on that shelf, and adjusting the height of the tripod and spread of the tripod legs so the front of the meter's microphone is positioned in 3-D space at the "listening position," which I define as the center of my head at the height of my ear canals when seated in my listening chair which is centered between the sidewalls, 1/3 of the length of the room from the wall behind the listening position, 37 ½ " from the floor. I then place the cushions back in the chair seat in front of the tripod to mimic as closely as possible the regular furnishings and presence of a body at the listening position. The meter's microphone points straight ahead, parallel to the floor, to the center point between the two speakers.
I usually operate both channels simultaneously when making measurements, not attempting to separately measure sound from each channel. This works well since I have always arranged my listening rooms for maximum left/right symmetry, believing this is necessary for best perception of all things related to stereo imaging and staging.
The exception is for setting the levels of stereo subwoofers. Since the level settings of these are not ganged but are individually controlled by volume controls on each sub, each must be individually adjusted for correct loudness at some frequency or frequencies within their operating range. To do this, I turn my preamp's balance control all the way left or right and establish reference midrange levels and relative bass levels that way.
I establish a reference level of 80 dB at 1 kHz using the warble tone at track 17, index 7 of Stereophile's Test CD 2. I adjust system volume so that the meter reads O dB when set to the 80 dB range. All measurements are made with the meter's switches set for C weighting, and Slow. When reading the meter, I stand away from and to the side of the meter as recommended in the meter's instructions. The tripod mounting yields much more consistent measurements than any attempt at handholding it can, especially now that the meter's bass response is extended to near DC.
The Measured Measuring Results
The correction factors published in Wallin's article seem to be in the ballpark. If anything, the meter readings after the modifications varied by an even greater amount at the 20 and 25 Hz frequencies. The modified meter, as the article states, seems to have significant response right on down to DC. Waving your hand once in front of the meter so as to produce a single air movement produces a significant reading in the 80 dB range of the meter.
As an example, where, before the modification, the meter indicated that using only my Z-Systems rdp-1 digital EQ/preamp, I needed to apply about 12 dB of boost to my Legacy Audio Whispers to achieve flat response at 25 Hz, after the modification, that figure dropped to 7 dB, indicating that the meter was now 5 dB more sensitive at 25 Hz after the modification. Or, applying the low frequency EQ strictly with the Legacy Audio Steradian (now known as the Step One), this difference corresponded to the difference between setting the control at the 3:30 position prior to the meter mod, versus 2:00 after the mod.
These days, I no longer use the Z-Systems rdp-1, having found that a combination of a Bryston BP-25 preamp and the Legacy Audio Steradian provides even better results in my system. And since I now also have added a pair of ACI Titan IILE subs to the system, I no longer feel the need to equalize the Whispers for flat bass down to 25. I now prefer to EQ the Whispers so they are flat down to 50 Hz. With the modified meter, this corresponds to 11:30 on the Steradian dial, versus 1:00 prior to the modification.
The Listening Results
Given the up-to-several-decibel magnitude of the difference in bass output I was targeting after the meter mod, I was surprised that my system's bass really didn't sound any leaner, less full, or powerful. The primary effect seemed to be an "uncovering" of all bass tones. While the 80 Hz region was reduced in output by only a decibel or so, this made a significant positive difference in the realism of male voices, both recorded and on live studio and location events heard on classical station WFMT. Plucked string bass lines became even more distinct, and organ pedal tones came through seemingly as powerfully as before, and with an even better sense of pitch.
Part of what may be at work here is that the Whispers 15" woofers and the driving Bryston 7B amps are not being stressed as much in the bass range as before, with a corresponding reduction in distortion, especially at very high levels. But the added clarity I note is present at all listening levels, so I believe the frequency response modifications are primarily responsible.
As a specific example, on track 11, "Postmodern Blues," of Patricia Barber's Modern Cool [Premonition PREM-741-2], beginning about 3'00" into the cut, there is an acoustic bass solo with some extremely powerful (and I believe electronically synthesized) sub-harmonics. The entire album has unusually high levels of both low string bass and kick drum, and both are quite percussive throughout with tremendous crest factors. But this particular cut is extremely challenging for any speaker with full-range pretensions to reproduce at high volume. Mini-monitors don't bother trying to capture the low notes and are not as bothered by it; "full-range" woofers with inadequate air-moving capacity or uneven bass response tend to bottom out at even moderate levels, with ugly distortion and/or scary snapping sounds as the bass drivers reach their mechanical excursion limits. Of course, my Whispers have no such problems (said he, immodestly, giggling just a little) and even before the re-EQing were able to play this passage at 100 dB+ levels with powerful deep bass and very little audible distress. But the sound got cleaner yet and the melody of this solo and the fingering and dynamic shifts of each note became better defined after the bass EQ was readjusted for flat response with the modified meter, especially after 4'00" when the notes get louder and more powerful, and the subharmonics are turned up.
This modification gets a hearty "thumbs up" from me. Do it and use the RadioShack analog SPL meter the way I've described and you can trust it up to 8 kHz or so and right down into the sub-basement. Your ears will be better trained to recognize flat, extended bass response in a loudspeaker. Never again will you be fooled long enough by a craftily tailored bass response to make an expensive purchasing mistake.
Don't forget to bookmark us! (CTRL-SHFT-D)