AR 302 Measurements

[administrator]

AR 302 measurements taken by David Rich of The Sensible Sound are now in the AR 303 section of the Library. I will let him comment further on his testing.

Thanks David!

Mark

[Guest richd]

Recently I came into the acquaintance of Tom Tyson through David Moran who is contributing editor of The Sensible Sound magazine. In November 2000 I wrote an article for The Sensible Sound (Issue 97) on the AR 302, AR 338 and the AR-5. The AR 302 and 338 where available at 75% discounts at the time from an on line store at the time and I felt the price to performance ratio was so compelling that a review was appropriate. Since it is 6 years after this review it should be understood the stock of the AR 303 series is now long depleted from this on line store. The Sensible Sound review should be posted soon in the AR 303 section of the library

As part of my discussions with Tom I re-measured the AR 5 and 302 (I no longer have the AR 338). The frequency response runs used in the original article (they were not shown in the original article) were not done to the same quality as the ones now posted in the AR 303 library. I know how to better control early reflections better than I did 6 years ago so the curves more closely represent the anechoic response.

The new curve set, now posted in the library, is measured using the same procedure I use to produce curve sets published as part of my current Sensible Sound speaker reviews. For example Issue 106 (Dec/Jan) contains reviews of the Infinity Beta 40 and Phase Technology 3.1 II and 6.1 II. Issue 107 contains a review of the DCM TFE 200.

Note the plots of the measurements are truncated at 10kHz. This is because the software I use (ETF 5) places significant limits on how a graphs horizontal axis can be presented. If I tell the software to give me a graph from 200Hz to 20KHz it produces one in which the major and minor divisions are placed at points that are not where they would normally be placed. Note that room effects come into play below 300Hz and the graphs are not representative of the speaker’s performance below that frequency.

The ETF software I use is very inexpensive yet it performs as well as much more expensive systems that I have tested (www.etfacoustic.com). It can be used with a Radio Shack level meter (calibration curves are on the sight) but a calibrated microphone specifically designed for speaker measurements gives much better results.

I have not supplied similar curves for a restored AR-5 that I own. The results were very different than the AR 302 and I could find no place to put the microphone to get anything approaching a flat response. As Tom Tyson pointed out the AR-5 is not measurable using quasi-anechoic measurement techniques. Tom sights a number of reasons for this and pointed me to an AES journal paper. The paper is Allison and Berkovitz (“The Sound Field in Home Listening Rooms” JAES Volume 20 Number 6 pp. 459-469; July/August 1972). Figure 9 (anechoic measurement) of the paper looks very similar to the curves I made of the AR 5. The Allison and Berkovitz paper presents flat measurements with the microphone well out in the room. An average of 16 rooms is presented and the curve looks much better. I thus do not think it advisable to supply you with my measurements of the AR-5

The AR 302 (as are all speakers past the AR 9) is designed to provide good axial response. The classic series is designed to provide good power response. As Ken Kantor published recently in the Classic Speaker Pages it is impossible to have both without a complex design such as the Magic speaker. Of interest is the fact that the AR 302 drivers and crossovers are very similar to each other (near field measurements of both speaker will appear in a separate post). It is the driver deployment on the front baffle that appears to make all the difference.

The subjective section of the Sensible Sound review needs a little clarification. The AR 5 was subjectively evaluated with the speaker placed as I placed the AR 302 (and other floor standing speakers I test) a couple of feet from the back wall and at least 2 feet from the side wall. The speakers were level matched with pink noise. I would have been seated away from the rear wall 8 feet back from the speaker. The AR 5s and 302s were put on 18 inch stands. An 18 inch stands puts my ear height at midrange height of the AR 302 (the optimal spot for flat response). I then did the same thing with the AR 5. In this placement the midrange and tweeter level controls had to backed off considerably. Also note that I know now that one of the AR-5s tweeters probably has a 4 ohm tweeter. I did not know such an issue existed at the time I purchased the speakers.

Tom pointed out I had the placement of the AR 5 all wrong. The AR-5 should have been horizontal on a book shelf against the wall with the speaker surrounded by books. I had know way of knowing that in 2000 (no Classic Speaker Pages or Tom Tyson to guide me in this back then). The AR literature from the 1960s is not at all clear on the placement restrictions of the classic AR speakers.

I should also clarify the subjective review of the AR 302 and AR 338 in my article for the Sensible Sound. The speaker is sensitive to the room it is placed in owing to its complex radiation pattern in the vertical axis and the wide dispersion in the horizontal axis. The room I tested them in at the time of the review appears to have been less than ideal for the AR 302 and 338 In the rooms I now use the AR 302 is a little reserved when it is listened to at its optimal axis (refer to the vertical radiation curves in the library to see the limited angles that the speaker produces its best response characteristics).

The vertical average response curve I have provided correlates with this subjective observation. The vertical radiation pattern of the AR 302 is not as tight as the better modern speaker in its price range. The result is a limited range over which the speaker can be listened to in the vertical plane. Since it is not floor mounted the limited range can be compensated for by adjusting the stand height to insure optimal operation for a given seated ear height. The loss of upper midrange energy in the vertical average response of the speaker is a consequence of the loss of energy seen in the individual vertical response curves that are off the optimal axis. It appear it is this loss of upper midrange energy that gives the AR 302 its characteristic AR sound.

Of significant interest is that the vertical radiation characteristics actually improve dramatically when the speaker is listened to well above the tweeter axis. I do not know of any other speaker that does this. This characteristic is important for the case where the speaker was placed on the floor. It also has relevance to the first reflection to the ceiling.

The horizontal curve family and the curve average are very flat and completive to any modern speaker I have measured to date. The exposed ¾ inch exposed dome clearly has good off axis response with very little change out to 10kHz.

David Rich