AR-3a Possible Tweeter and Hello

[RoyC]

It seems though, that while I may be perceived as placing too much importance on the speaker impedence, the way I'm reading your post seems, to me, as placing too little importance on it.

John

p.s. Here they are with new grills:

Well, you are wrong about that, John, and your conjecture and logic are flawed as it relates to the issue at hand. Do you understand the difference between impedance and dcr? Do you understand the effect the AR level controls have on these parameters? Have you looked at individual driver response graphs and wonder why they are different even though the drivers may have the same rated impedance and/or dcr? How about the capacitor's interaction with these variables? Other parameters such as inductance and (especially) sensitivity are also considerations. With that said, it is very likely the Peerless tweeter would work well enough with the small parallel coil treatment, but the proper value would have to be determined. As it stands you are (subjectively) comparing apples and oranges, and the discussion has become pointless.

There is much better grille cloth than burlap available. Carl has run some tests showing burlap to have a negative effect on high frequencies. Your grilles and badges do appear to be cost effective, however.

Roy

[michiganpat]

take this with a grain of salt, as I'm just a hack, too...

the point I think you're missing John, is that a driver's response is the sum of it's electrical crossover AND it's mechanical response. the AR tweeter has a fairly steep mechanical response slope, so a simple first order electrical crossover (6db/octave) with it's mechanical response sums probably to something similar to a 4th order (24db/octave) at the crossover point. the peerless india and hivi tweeter's frequency response goes much lower in the frequency range....

https://www.madisound.com/store/product_info.php?manufacturers_id=136&products_id=1493

the Q1R is pretty flat from ~2000hz to 20000hz, with about a 6db/octave roll-off at that point....adding a 4uf cap only on the Q1R yields a 6db/octave roll off beginning at ~5000hz, going steeper once it's mechanical roll off makes it more like 12db/octave at around 2000 hz, so you're getting a lot more acoustic information once you sum the mid and tweet with the shallower tweet high pass slope. the paralleled inductor turns it into a 2nd order (12db/octave) roll off, and the .07mh value is quite small, smaller than a typical "textbook" value, and I don't have a good feel on how that impacts the electrical crossover slope, whether that causes a very large round knee in the curve of the electrical crossover beginning pretty high in the frequency response (my guess) or a very sharp knee, where response is flat to very near the crossover point and then takes a header

[JohnHeubel]

Sorry to be so long in getting back to this as it's been a trying week. We lost a friend to cancer after a long struggle, and the next day my wife found out that one of the students she just taught was killed in an auto accident, the visitation/funeral just yesterday. Needless to say, I haven't been on the computer much.

Anywho, after reading and rereading some of the posts I figured out that, while I was using the term impedence, I was thinking more of the dcr values as they are usually listed (if known) right next to the rated impedence, so sorry if I caused even more confusion. My approach was to match as many of the specs as I could (realizing that there would be some tradeoffs), and thought that based on the dome material, size, freq. response, dcr, impedence, etc that it should be a close substitute. And though I have a basic understanding of some of the other items mentioned, that's all it is, a basic understanding. Like I said before, I won't begin to claim to know as much about these topics as most of you here. It would have been easy enough (and undoubtedly truer to the original) for me to just have gone with the HiVi since much research had been done by you all, but as in many other areas of life, I don't typically just say, "this is how it's done...don't ask questions or wonder why." So, if that outlook offended anybody, I'm sorry. I do like the sound of the Peerless, so despite what shortcomings it may have in mimicking the AR tweeter, that wasn't really what I was looking to do afterall.

I also realized that, though this discussion has been mainly centered on the AR-3a, my question about replacement drivers is actually a more general question, that of finding a suitable replacement for "x" driver, whilst being limited in what testing and refining can actually be accomplished. For instance, I come across a pair of speakers that I want to restore, and for the sake of argument let's assume that one tweeter is blown. Let's also say that information is limited as to original specs. Would a reasonable course of action to find a replacement be to measure the dcr of the good one and then try to reverse engineer the crossover to get the crossover frequency and possible slope? Then try to compare drivers and see how many "might" work? Remember, the only test equipment is a multimeter. Obviously, if it's a woofer that's bad, taking cabinet measurements to determine volume would come into play, but what's a reasonable strategy to take when trying to find replacements with limited information or ability to test? PE has their driver replacement guide and wizard, but are there other good strategies to use?

It's fine if this should be moved to another thread, but these are the sorts of things I'm thinking about after the different reading I've done on various forums.

John

[genek]

Offhand I think you have no chance at all of reverse-engineering a crossover using nothing but a multimeter.

[JohnHeubel]

Offhand I think you have no chance at all of reverse-engineering a crossover using nothing but a multimeter.

What else would be needed then? Measuring inductance is pretty much out of the question since I don't have a signal generator.

[genek]

If you talk to our more exert crossover people, I think you'll find that most of them combine measuring frequency response with trial and error component swapping, at least until they become experienced enough to be able to make more educated guesses about what a given component value will do.

[Rlowe]

What else would be needed then? Measuring inductance is pretty much out of the question since I don't have a signal generator.

As has been mentioned, if you like the sound of the tweeters, great, enjoy, listen to the music.

This site has a lot of good information and very helpful people.

I have found the following posts helpful in me in appreciating some of the effort required to "match" woofers to tweeters,

it is not as easy as I first thought.

I hope they assist you, if not, they're still a good read.

David Smith is an experienced speaker designer with lots of background in the industry.

He describes his development experience of "replacing" a tweeters in a pair of $58, AR4x,

obviously not the same as your project, but it covers a lot of the speaker designer's craft.

http://www.classicspeakerpages.net/IP.Board/index.php?showtopic=5864&st=0&p=83421&hl=dave&fromsearch=1entry83421

The detailed posting can be found at;

http://www.classicspeakerpages.net/IP.Board/index.php?showtopic=5815&st=80