"Straying from Original"

[frankmarsi]

10-18-06

Ladies, gentlemen and poor-slob AR addicted associate friends of mine.

Lend me your ears and indulge me if you will.

After I installed a couple of the ‘official-AR’ replacement tweeters from that place in MA. - I don’t have to state their name. True AR cognoscenti are aware of who I refer to, so no offense is intended.

This is hard to say; after I spent so much money on 18 of them- I almost feel these drivers not only offer what seems like an acceptable high end, but they may project a bit too much mid-range also. But the most painful realization is that the few original 1972 tweeters I still have operating- definitely have a crisper and believe it or not- more extended top end, albeit at lower db’s. This opinion is based on ‘close-ear’ observation only, but still it’s undeniable noticeable. And although their actual volume is approximately 2-3db or so louder than the originals, it’s not much of a consolation. Although they sure handle more watts which is what I require.

I was hoping if there might be some helpful soul out there that can recommend or advise on some sort of miniature pot that I may mount either with adhesive, screw or how-ever to the wood front baffle and actually adjust the tweeter’s response while live listening with varied program material? I’m not certain where to start in terms of actual electrical values and specs.

Yes of course I have tone controls on my pre-amp, but I want to do it directly on each ‘driver’ as this is ‘their’ problem I’m trying to correct directly. When I think about how much I actually spent on all of these tweeters I’m almost moved to installing my low-end equalizer back into service. I pulled it out about a year ago and have been happy with out it thus far, but now reconsidering its use. I will in any event install it once I get my cassette and reel-reel set-up ready to use. Yes I still use cassette and ‘Revox’ 10 inch when I past the time making extended ‘all day long’ play-back tapes during the winter months. My main sources are CD and vinyl when I’m going for ‘quality’ listening, although that old A77 type4 does put out fine quality of its own, almost un-discernibly so when recording from CD to tape.

So getting back to the tweeters, sorry, you see I avoid the pain of thinking that I have 18 of them little bastards. There must be a tiny set-up with a variable resistor pot or capacitive controller to exaggerate the tweeter’s limits. I’ll spay it flat black so it won’t stand out next to the tweeter, but it must be those real small components with tiny shafts and bodies. Maybe a little slot ‘trim-pot’ or something giving me an ability to adjust things. I use AR-LST’s and AR-3a’s. On the LST, I rather keep the ‘autotransformer’ in the flat #2 position in order to keep the bass flat. Even though I have spare woofers I really try to be careful using what I have as I usually listen at high volumes. While I’m here after being on this site for a short while now- I will ‘tell-all’ and reveal that my LST’s were originally installed in a New York City Broadway theatre and decommissioned after fifteen years of service when I became their extra proud new owner. Actually there were 5 along with about ten or more Ar-6’s or AR-7’s, they were taken by someone else, I got the 5- LST’s from which I made four operational. The fifth cabinet had a blown woofer and high-ends and was very ‘smashed-up’ in front and everywhere else, badly so. All surrounds were long gone and in 1990 when I finally ‘re-commissioned’ them there was only one guy who offered surrounds from Flat Rock, N.C., called “Ken’s Hi-Fi Clinic” to my knowledge, now there’s several million sellers. I stupidly in 1987 threw it out as I had no real room to store it and the poor thing was sad to look at, so I put it down like a good ole horse, it hurt me as well.

So there it is, my ears are open to all and any opinions.

Your friend and mine, Frank Marsi

[dynaco_dan]

Hi Frank;

I am not 100% certain, but, I believe that the new tweeters are rated at 6 ohms impedance, not 4 ohms.

If this is true, then they will be about 1 or so db down in level from the original tweeters sound level.

You must use the same impedance tweeters in all enclosures or there will be a very slight imbalance.

Setting you LST's at 1 should brighten up everything a very, very slight amount.

Good luck, Frank.

[RoyC]

Frank,

Your ears aren't deceiving you, and you are not alone. The "Tonegen" version replacement tweeter you purchased does have an extended frequency response downward into the midrange compared to the original. It is similar to the AR-11 era tweeter, and was used in the AR-3a Limited (in Asia) in the early 90's with a different crossover.

Other than modifying the crossover, it may be time to dust off the equalizer. In the AR-3a that tweeter requires a smaller value cap (4uf) and a parallel coil (.16mh) to sound more like the original.

Roy

[frankmarsi]

Thanks gentlemen.

I should've also stated that my goal here is to 'bring-up' the high-end and 'take-down' the extra mid-energy these newer tweeters have. Am I asking for an impossiblity and will an equalizer adjust anything above 16K. mine doesn't. I haven't ever seen a consumer one that does except for really expensive pro-studio stuff. Those older original tweeters do perform above 15-16k or more even though there's the obvious 'fall-off'. My little gentle porcelin like and danty ears believe they can at times go way up there, forget my age, I 'will' my ears and have all these years.I've been training these 'Golden-Ears' long before it became cool to claim it publically by anyone else. Or at least while I still can and delude myself that I can. If you fellas want to just, well even if you gotta fool me into thinking there's a way I can accomplish my goal here I'll be happy with that. Hold-on! STOP the Music, STOP the Music!! Wow I must be 'tripping' here thinking about cleaning all of my woofer frames with acetone as was suggested in another post here in '05. I've got 8 need-in fix-in right now!

Frank Marsi

[onplane]

>Thanks gentlemen.

>I should've also stated that my goal here is to 'bring-up' the

>high-end and 'take-down' the extra mid-energy these newer

>tweeters have.

>Frank Marsi

Frank, what you seek might NOT be too hard to achieve on the AR-3a's. If the tweeters are producing too much in the mid-frequency area 3,000 to 5,000 Hz, Roy's suggestion of reducing the cap and adding a coil will improve the 5,000 Hz "cut-off".

Basically what Roy's suggestion does is starve the tweeter for current in the frequency range that you don't want the tweeter to produce sound. As usual, Roy's idea is a good one!

Now, to compensate for the tweeter being too sensitive, can't you use the pot to reduce the tweeter output?

Finally, you want more "high-end". That is, the tweeters are LESS sensitive in the upper frequencies. The solution, as you suspect Frank, is to send more voltage to the speakers in the "high-end". The simplest solution is the treble tone control on a good amp.

I have several amps where the treble tone control operates in the 15K to 20K range with a boost of 10+ db. With 10+ db you really ought to be able to achieve more than sufficient boost, Frank.

Strange as it seems, in this case the old fashioned tone controls may work better than an EQ!

Hope this helps ...

Jerry

[RoyC]

Frank, there may not be an easy solution to the problem short of digging into the fiberglass. Taming the midrange may enhance the higher frequencies. It can't hurt to try the EQ if you have one handy. It should give you a general idea of the offending or missing frequencies.

Vern, the dcr of the replacements measures 4 ohms, but considering the original tweeters were around 2.5 ohms or less, I agree that the impedance difference is a factor.

By today's standards the old tweeter seems to have acted more like a "supertweeter" providing substantial output and reinforcement at the highest frequencies, but rolling off quickly at lower frequencies, while the midrange driver also had much high frequency duty. The new tweeter simply overlaps the midrange more, yet has no more (maybe less?) output at the higher frequencies.

Roy

[onplane]

>Vern, the dcr of the replacements measures 4 ohms, but

>considering the original tweeters were around 2.5 ohms or

>less, I agree that the impedance difference is a factor.

>

>Roy

>

Roy, if the new tweeter has a higher impedance, that will tend to lower the xover frequency. Your suggestion to reduce the cap to 4uf, will certainly help.

I'm wondering if testing a 3uf cap might would be worth a try?

Frank could do this by putting the two 6uf's in series. Now, this would just be for a test and he could see without having to purchase caps, if it achieves his objective.

What do you think, Roy?

Regards,

Jerry

[RoyC]

>I'm wondering if testing a 3uf cap might would be worth a try?

I agree:-) that it is worth a try. 3uf would be in the ballpark if no parallel coil is used....only one way to find out.

Roy

[Carlspeak]

>Frank,

>

>Your ears aren't deceiving you, and you are not alone. The

>"Tonegen" version replacement tweeter you purchased

>does have an extended frequency response downward into the

>midrange compared to the original. It is similar to the AR-11

>era tweeter, and was used in the AR-3a Limited (in Asia) in

>the early 90's with a different crossover.

>

>Other than modifying the crossover, it may be time to dust off

>the equalizer. In the AR-3a that tweeter requires a smaller

>value cap (4uf) and a parallel coil (.16mh) to sound more like

>the original.

>

>Roy

Frank/Roy. You know, it's a shame we all have to find out these little 'gory' details about drivers sold by that MA outfit from this and other chat rooms. I deal with a number of raw driver suppliers nationwide and 'that MA outfit' is the ONLY one I know of who doesn't provide T/S parameters of the speakers they sell on their website. I wrote about this weeks ago in this same chat room and now here is another example that drove me to get on my soapbox again. Most or all of their drivers are made in some Asian country and you can bet the manufacturer has their specifications which could be shared with us just like the rest of the driver sellers do.

I think most of the posters here are familiar with T/S parameters and would find that information useful when making a purchase from 'that MA outfit'. Yes, their modern drivers may not be exactly the same as original AR drivers. I think that's okay. We simply need to know what those differences are. Even some basic information like SPL, Fs, VAS and Rdc would help. They should publish a table covering all the drivers. It should include T/S parameters on both their new stuff and original AR specs.

Granted, there were a number of changes to original speakers in the early years, but the archives should be there for each model change. It seems a bit cruel to sell someone a $90 woofer based only on their recommendation that it is a replacement for the original in some particular model speaker. Physically, it may fit fine. Acoustically, it may need some crossover work to make it sound more like the original. For example, Frank's 18 tweeters could have arrived with a recommendation for a simple attenuation circuit that would address the increased efficiency of the new tweeter.

I just finished repairing a pair of AR4ax's. One of the 3/4 inch tweeters was dead. I ordered a new one from 'that MA. outfit'. Upon receipt I tested it and found it was +7 dB more efficient than the working old one in the other speaker. I called the customer and told him about it and asked if he wanted me to get another new one so they both would be a better match. He declined. Fortunately, I was able back off just enough on the pot feeding the new tweeter to tame it down to match the old one.

It's understandable there were no T/S parameters in the early days of AR. I believe the famous Theil and Small papers weren't published until the early 70's. But, there must be some fine working original speakers available that could be tested to generate a database and derive useful average guidelines of performance.

Remember, it's all about the music

Carl

Carl's Custom Loudspeakers

[onplane]

>I just finished repairing a pair of AR4ax's. One of the 3/4

>inch tweeters was dead. I ordered a new one from 'that MA.

>outfit'. Upon receipt I tested it and found it was +7 dB more

>efficient than the working old one in the other speaker. I

>called the customer and told him about it and asked if he

>wanted me to get another new one so they both would be a

>better match. He declined. Fortunately, I was able back off

>just enough on the pot feeding the new tweeter to tame it down

>to match the old one.

>

>Carl

>Carl's Custom Loudspeakers

Carl, now that is one legitimate use for those pots. That is matching drivers of different efficiencies!

Your point on speaker parameters is a good one. Efficiency is just one measure as Frank has learned. We need to be concern with frequency coverage as well. Indeed, frequency coverage may require xover network changes to bring things back in balance.

Unfortunately, it's very hard to do this when we are "flying blind" with no specifications or parameters to make those calculations.

Regards,

Jerry

[frankmarsi]

10-19-06

If I may say so; a healthy out-pouring of advice and opinions, I am grateful to all.

When I first received these tweeters I decided to measure their impedance

just to look for uniformity, etc. They averaged 3.6 to 4.10 ohms through-out the lot.

As I mentioned earlier these guys seem to have more mid-freq. than I need or like and seemingly less sweet and delicate sounding top-end as their vintage counter parts.

Be that as it may, my main concern is not to change the ‘voicing’- of the 3a’s and LST’s -a new-ish term to me, voicing, must be some dumb yuppie word?

Anyway my situation is as I’ve said and I feel perhaps I would try what has been suggested to me in this post, which is as I was hoping for at the on-set. A ‘cap’ or two in line or something similar and just as easy would suit me fine. Question is my main concern is that sweet extended top-end, which is what I want to maintain. Is it possible that these replacement tweeters don’t have the high –end to give, anyone doing a more qualified test? Don’t want to burden anyone, just looking for more suggestions.

Thanks, FM

[onplane]

>10-19-06

>

>Question is my main concern is that sweet extended top-end,

>which is what I want to maintain. Is it possible that these

>replacement tweeters don’t have the high –end to give, anyone

>doing a more qualified test? Don’t want to burden anyone, just

>looking for more suggestions.

>

>Thanks, FM

Frank, my sense is you can test it yourself. Turn your treble controls full up. With 10db and possibly 15db additional energy you'll be able to hear if the tweeters can produce that top end.

Most speakers don't have a sharp frequency "cut off". It's more like a "roll off". So by boosting treble (especially a treble control designed in the 15K to 20K range), you might be able to offset the roll off.

On the LST's, this is going to be a more difficult "fix". Frank, have you ever considered bi-amping? Unfortunately to bi-amp the LST's you need two good amps. When I bi-amp my AR-3a's, I'm able to use a low power, low cost, "chip" amp for the mid & tweeter. It has more than enough power to drive the mid/tweeter.

You aren't so lucky on the LST, Frank, as you'd need a decent 2nd amp driving all those mids and tweeters.

Regards,

Jerry

[Carlspeak]

....."Be that as it may, my main concern is not to change the ‘voicing’- of the 3a’s and LST’s "......

If you succeed in matching the 'voicing', it will be pure luck. In other posts I've postulated that the physical pakeup of a driver is the most important determinant of a speaker's 'voice' or sound to the listener. Obviously, the original tweeters and new ones are quite different in makup and thus may never sound the same. I just hope you find a way to be satisfied with what you have to work with.

Remember, it's all about the music

Carl

Carl's Custom Loudspeakers

[dynaco_dan]

Hi Frank;

Quite a variance in DCR for the 18 tweeters.

I would expect to hear a difference, if the highest and lowest impedance tweeters was used in a pair of AR-3A's for example, rather than less so in a group of 4 per LST.

The original AR 3/4" 4 ohm tweeters DCR was approximately 2.75 ohms.

Why is everyone avoiding the suppliers name, there is nothing to be ashamed of being written here, just the truthful facts.

[Pete B]

>10-19-06

>

>If I may say so; a healthy out-pouring of advice and opinions,

>I am grateful to all.

>When I first received these tweeters I decided to measure

>their impedance

>just to look for uniformity, etc. They averaged 3.6 to 4.10

>ohms through-out the lot.

>As I mentioned earlier these guys seem to have more mid-freq.

>than I need or like and seemingly less sweet and delicate

>sounding top-end as their vintage counter parts.

>Be that as it may, my main concern is not to change the

>‘voicing’- of the 3a’s and LST’s -a new-ish term to me,

>voicing, must be some dumb yuppie word?

>Anyway my situation is as I’ve said and I feel perhaps I would

>try what has been suggested to me in this post, which is as I

>was hoping for at the on-set. A ‘cap’ or two in line or

>something similar and just as easy would suit me fine.

>Question is my main concern is that sweet extended top-end,

>which is what I want to maintain. Is it possible that these

>replacement tweeters don’t have the high –end to give, anyone

>doing a more qualified test? Don’t want to burden anyone, just

>looking for more suggestions.

>

>Thanks, FM

Hi Frank,

I've cloned several speakers over the years, in some cases replacing a fabric dome type with an aluminum dome for reasons that would take too long to get into right now. It can be done and it's not that hard when selecting from the many drivers available on the market.

What matters is what we call the cascade, or complex multiplication, of the crossover electrical response, as loaded by the driver, times the electro-acoustical response of the tweeter. Make the total responses match and it is very difficult if not impossible to hear a difference. Generally, assuming that we have "good" drivers with smooth passband response, I find that matching level or amplitude, and tilt across the passband is a good start. Also very important is to match the transition region, magnitude and phase of course, at the crossover point. I use a circuit simulator, and sometimes an automatic optimizer to adjust circuit values and obtain the desired response.

No AR-3a or LST has come my way, yet, so I've not measured the tweeter. I've also not measured or heard the "replacement" tweeter, and so I don't have the required data to investigate what's required in the way of crossover mods.

[RoyC]

>As I mentioned earlier these guys seem to have more mid-freq.

>than I need or like and seemingly less sweet and delicate

>sounding top-end as their vintage counter parts.

My experience as well, Frank.

>Be that as it may, my main concern is not to change the

>‘voicing’- of the 3a’s and LST’s -a new-ish term to me,

>voicing, must be some dumb yuppie word?

:-) No need to worry about that. The term has been around a long time...("character", "personality", whatever)...

It is typically used in conjuction with the design goals of a speaker builder.

>Is it possible that these

>replacement tweeters don’t have the high –end to give, anyone

>doing a more qualified test? Don’t want to burden anyone, just

>looking for more suggestions.

The response curve is definitely different. The highest end is no better and may not even be as good as the original. That is why simple attenuation won't work...been there.

Try disconnecting the 3a woofer jumpers and turn the midranges all the way down, so you can hear just the tweeters. You can compare the old with the new in this way, and give you an idea of what you are dealing with. It is amazing how little sound seems to be coming out of the old tweeter...but that is part of its charm, and attributable to its unique construction. It really is restricted to the highest frequencies.

As discussed above, the lower cap value is the easiest way to start out, and that will be an improvement. You will still have more extension into the midrange, but it will be less bothersome. Some folks prefer it over the original.

My personal experience is that in order to get close to the sound of the original tweeter, an inductor coil in parallel with the tweeter is necessary. A .10mh inductor coil was used in the AR-11 and AR-3a Improved, and a .16mh coil was used in the 3a Limited. The associated series caps were 10uf, 6uf anf 4uf respectively. A parallel coil would roll the new tweeter off more rapidly (like the old one did naturally), and offer more protection from lower frequencies as well.

I have been comparing notes with other forum members offline for some time on this issue. The consensus, so far, seems to be conversion to the tweeter circuit of the AR-3a Limited crossover. As Carl pointed out, we have little to work with.

Roy

[frankmarsi]

Very kind of everyone in trying to help me here.

Now can anyone offer a simply drawing of the circuit.

Is mounting a simple 1uf or greater 'cap' in line with one for each of the tweeters a possibility?

Thanks, Frank Marsi

[Carlspeak]

>Very kind of everyone in trying to help me here.

>Now can anyone offer a simply drawing of the circuit.

Sorry Frank. I don't have one to offer. However, you need to find out what value you currently have for the tweeter cap. Simply follow both wires from the tweeter back thru the circuit. You will encounter a cap or a resistor first. If it's a resistor keep going back along the wire and you will eventually locate the cap.

>Is mounting a simple 1uf or greater 'cap' in line with one for

>each of the tweeters a possibility?

No. but to successfuly tame that excessive midrange in the new tweeter, you need to lower the cap value which, in turn, will raise the tweeter's crossover value. I suggest you start by lowering the cap value about 20%. For example, if your tweeter cap is 6uF, then try an inexpensive 4.7 uF cap. They are very common and inexpensive. Metallized PP caps are not terribly expensive. When you buy the 4.7 cap, get a few others in an intermetiate range betwen 4.7 and 6. Say a 5.0, 5.3 and 5.6. Try each one separately IN PLACE of the original. You should find one that works for you.

With regard to the sweet high end you seek from the new tweeter, I think that will come with break in of the new tweeter and cap. It may take a few month's of playing music thru them, but it should happen.

Remember, it's all about the music

Carl

Carl's Custom Loudspeakers

[onplane]

Carl's approach is good. I'm a little concerned, however, that any cap without a coil will NOT give you the sharp cutoff to really "kill off" that nusiance mid-range.

Frank, why don't you try a simple experiment. You have 4 LST's. Open two and remove the 6 uf cap from one and put it in series with the 6 uf cap in the other LST. This will give you the equivalent of a 3 uf cap in the "modified" LST.

Now, you have a "modified" LST in one channel and an unmodified in the other. Turn on your amp and select mono so the same signal is going to each LST.

YOu should hear a significant difference in tweeters. The modified LST tweeters should have far less mid-range.

If you can hear no difference then we know you'll have to get a coil as well.

Hope this helps ...

Regards,

Jerry

[Carlspeak]

>Carl's approach is good. I'm a little concerned, however,

>that any cap without a coil will NOT give you the sharp cutoff

>to really "kill off" that nusiance mid-range.

>

>Frank, why don't you try a simple experiment. You have 4

>LST's. Open two and remove the 6 uf cap from one and put it

>in series with the 6 uf cap in the other LST. This will give

>you the equivalent of a 3 uf cap in the "modified"

>LST.

>

>Now, you have a "modified" LST in one channel and an

>unmodified in the other. Turn on your amp and select mono so

>the same signal is going to each LST.

>

>YOu should hear a significant difference in tweeters. The

>modified LST tweeters should have far less mid-range.

>

>If you can hear no difference then we know you'll have to get

>a coil as well.

>

>Hope this helps ...

>

>Regards,

>Jerry

>

>

>

Adding a .1 mH air core inductor (as Roy suggested earlier) in parallel to the neg. side of the tweeter will turn the single cap idea I proposed earlier from a 6dB roll off to a 12 dB.

This will accomplish two things. It will pad the midrange frequencies better and, also serve to protect to tweeter better from low frequency information that could cause it to fail.

Remember, it's all about the music

Carl

Carl's Custom Loudspeakers

[johnieo]

>Very kind of everyone in trying to help me here.

>Now can anyone offer a simply drawing of the circuit.

Frank:

There are at least three AR speakers (maybe more?) that used the -12 dB/oct rolloff (an L-C crossover instead of an R-C crossover circuit). The European "AR-3a Improved" (second version) and the AR-11 both used an L-C circuit. The "Improved" used 6 uF and 0.105 mH(AR #14 coil), whereas the AR-11 used 10 uF and 0.105 mH. Both of these speakers used fixed resistors instead of potentiometers to achieve two (or three) levels of attenuation.

The L-C circuit that may be the most like an AR-3a or LST might be the tweeter section of the crossover used in the Asian "AR-3a Limited," -- a 1991 design. It also used an L-Pad; an easy replacement for the AR-3a pot. (See attached x-o circuit.) This circuit used a 4-uF capacitor and 0.160-mH coil. It both increases the crossover frequency slightly and causes its low frequency output to fall off more sharply as the frequency decreases toward the midrange.

Of course, any change in the tweeter crossover affects only the sound from the tweeter and not that coming from the midrange driver; however, it will affect the speaker's overall upper mid-range. One thing that is nice about this crossover is its ability to vary the output smoothly from max to min, rather than in 2 or 3 fixed steps such as in the AR-11 or "AR-3a Improved."

[RoyC]

Attached is an LST schematic.

The new tweeter issue is probably going to be more pronounced with the multiple tweeter configuration of the LST. It appears that the LST had the same 6uf cap in the tweeter circuit as the 3a.

My experience using that tweeter in the 3a is that a cap of no higher value than 2uf to 3uf is somewhat suitable, if not using a parallel coil. In the LST it may be closer to 2uf.

Roy

[frankmarsi]

>Attached is an LST schematic.

>

>The new tweeter issue is probably going to be more pronounced

>with the multiple tweeter configuration of the LST. It appears

>that the LST had the same 6uf cap in the tweeter circuit as

>the 3a.

>

>My experience using that tweeter in the 3a is that a cap of no

>higher value than 2uf to 3uf is somewhat suitable, if not

>using a parallel coil. In the LST it may be closer to 2uf.

>

>Roy

>

>

>

Hi Roy, by adding a 1uf to 2uf in line with the positive terminal to each tweeter might do it I'm presuming here, is that correct? I can go to RS and buy a couple and try it out, I will post my results as soon as I can. I was hoping there'd be some sort of variable control I could use, to tune it to my tastes?

Thanks Roy and everyone.

Frank Marsi

[RoyC]

Hi Frank,

You really need to disconnect the single 6uf cap from the tweeter circuit and replace it with a new one of the desired value. It means removing the woofer to get to the crossover. I'm not sure what you would end up with by connecting separate caps to each tweeter.

Roy

[onplane]

Frank:

You could try something like this:

http://www.partsexpress.com/pe/showdetl.cf...tnumber=269-950

It's cheap enough and gives you the 2uf cap plus a coil that's pretty close to put across the tweeters.

This ought to really cut off the mid-range from those tweeters. You'll have to re-wire this thing a little, but not much. Plus everything is attached to a board.

Regards,

Jerry