Old Radio Shack SPL cal

[Testpattern]

I have an older (first gen?) Radio Shack SPL meter which has, and is, serving me well. I'm about to embark on setting up my sub via my new BFD 1124P/REW and have become acutely aware of the need for a calibration curve for the RS meter. Problem is I have not been able to find any info on my model (42-3019).

Does anyone have a curve for this beast?

Thanks,

[Sonnie]

Yeah... that is the ancient model there.

It may be that the old correction values will work with it... it seems that this was the model they were originally designed for.



It's still a **** shoot though, and if you want to be certain you are getting true measurements I would go ahead and invest in a Galaxy CM-140 for about 100 buck.

[Testpattern]

Thanks, I give these a try..... If the results don't fair well I'll bust out for a decent mic as you suggest.

Al...

[Testpattern]

Just thought I'd kick out some feedback on the old RS meter cal.

I have a friend who has a 33-2050 (latest analog I believe). I ran and plotted a set of curves manually using a low freq test disc with 1/6 octave steps. I then ran my old RS meter (42-3019) and the numbers were generally the same. A few steps did show some differences but it was always less than 1 dba. It would appear by this analysis that the curve for the 33-2050 will work just fine for the older 42-3019 meter.

[brucek]

Can you tell us how low you measured? Did you test down to 10Hz?

brucek

[Testpattern]

Sure...

I utilized a Rives Audio CD-2 test disc which covers the audio spectrum in "1/3" octave steps (not 1/6 as previously reported).

I performed this comparison over a range of 20-400 hz. Meters were measured in my listening room mounted in a fixed location (sweet spot). Two sets of curves were collected manually for each meter by noting the meter deflection at each test tone. Readings were virtually identical between the two readings for each RS meter. When the two passes for the 42-3019 were completed, it was removed from the fixture and replaced with the 33-2050 and the same sweeps repeated twice more. The average for the two 42-3019 sweeps were compared to the average for the 33-2050 sweeps. The differences were minimal with any observed varances less than 1 dba.

Although the older meter seems to reflect the same values as the newer model, it isn't as useful overall. The 42-3019 minimum scale is 70 dba where as the new 33-2050 includes a 60 dba selection in addition.

I hope this is useful.

[brucek]

Yeah, thanks. I don't think we've ever had anyone measure the old one against the newer one, but once in a while someone asks the same question as you and we just answer that we don't know. Thanks for doing that test...

brucek

[Chrisbee]

Tp

Measuring test tones needs the patience of a saint!

Is it impossible for you to run REW?

REW could have graphed the two meter's response traces in just a few seconds and then you could have saved the graphs to post here.

[Testpattern]

I'll be glad to but it'll be a bit. We've had a family emergency which will delay my acoustic endeavors for a while. You're correct... "patience of a saint" but in a quiet environment the results are very repeatable.

[Testpattern]

As promised, here's a direct comparison of the 33-2050 and 42-3019 Radio Shack meters. A few words of how things were done may be in order.

I utilized a stand alone Boston Acoustics SW-10 sub standing next to my computer in my office. I had set this up to provide convenience for playing around with REW to attempt to absorb the learning curve before taking on the BFD 1124P. Anyway, the meters were set on a tripod at 1 meter near field. The setup position was carefully maintained between meter exchanges.

After loading and performing a sound card calibration on my new SB Live 24-bit External, I loaded in the mic calibration for the RS 33-2050 as obtained from the forum download page.

Levels were adjusted and a set of three sweeps were averaged for each meter. These two averaged trace sets were displayed together and one was adjusted vertically to negate a small volume setting difference.

The results confirm my previous manual measurements in that these two meters are basically identical in their performance. Further comparison of the respective schematics provided with each device supports this because all components are virtually identical in value. The two schematics are identical except for the added range step on the 33-2050 which provides a 60 db selection the 42-3019 does not have. Some of the scaling values were changed in the respective voltage divider to accommodate this added range.

radio shack meter comparison.jpg

The green trace is the older 42-3019 meter. The red trace is the 33-2050 meter. The observed differences around 16 -17 hz and 21 - 23 hz are a bit misleading. This is an area of the curve were repeatability varied between each run. Had I ran a large number of measurements to be averaged, this observed difference would virtually disappear.

The SW-10 isn't that impressive just setting there but it accommodated this task just fine.

Hopefully this info will be of some assistance to someone. At least now, I'm comfortable using the cal data for the 33-2050 to calibrate my older 42-3019 RS meter.

Cheers.....

[Chrisbee]

Hi Tp

JohnM recently advised me to use longer REW sweeps to get better repeatedly at very low frequencies.

When you click on the "Measure" box the sweep details in the drop down box can be altered. You can change the sweep length to 256, 512 and M.

There are also options to change end frequency, level and number of sweeps if you click on the chevrons. (the sergeant's stripes)

M is the longest sweep and great fun as it climbs very slowly in frequency from the inaudible 2Hz starting point into full audibility and beyond.

In my recent "REW Repeatability" thread I showed how the multiple traces closed together right down to just a few Hz using the longer sweeps. The effect was quite dramatic in comparison with the default sweep length.

[Testpattern]

Good to know Chrisbee. I'll check it out next time I'm running a curve.

Thanks....

[brucek]

Hopefully this info will be of some assistance to someone

Thanks for doing that test. I think we'll add the 42-3019 as a meter that can reasonably use the 33-2050 calibration file on the download page.....

brucek

[Sonnie]

I also thank you for completing the comparisons.

I have notated these findings on the Downloads page and referenced this thread.