Jump to content

AR91 repair story - new mids and tweeters identified


Carlspeak

Recommended Posts

I have to smile regarding Tom's comments about the placement of ferro fluid on one side or the other of the mid's voice coil. Once the faceplate adornments are removed, there are no practical differences in performance between any of the backwired 4 ohm dome mids. Further, any of them are a much better choice for repair or restoration of all AR models originally equipped with a version of this mid in contrast to the mismatched replacement currently sold by AB Tech, Simply Speakers, etc. It is just a matter of transferring the donut ring or the metal grille from the dead one.

I agree, Robert, this thread probably should be in the mods section.

Roy

Link to comment
Share on other sites

  • Replies 80
  • Created
  • Last Reply

Roy,

on the post 80's woofers with the weak spiders did you recone them to fix the problem?

Hi Harry,

The spider is replaced with one very close to the mid 70's version, and the foam must be replaced again in the process. The voice coil and cone remain.

This solves the spider problem, but the woofer will still not perform exactly the same as earlier versions.

Try the AR-91 woofers in your AR-11's. They may be as good as they can be. Stuffing type and amount, along with Carl's comments regarding the crossover differences are still in the mix.

Roy

Link to comment
Share on other sites

Your observation regarding the low "Q" would suggest that the units were functioning true to design intention, as previously outlined by Tom Tyson:

See also the following comment concerning the AR-58s (which had the same 1.37 mH coil in place):

Given the premise and direction of this topic, it would seem perhaps better placed in the "Mods, Tweaks, and Upgrades to the Classics" section of the forum.

Robert_S

Thaks Robert for sharing. I was hoping TT would chime in on the 91 at some point, but you took care of that nicely. I wondered about the placement of this thread when I created it.

Link to comment
Share on other sites

I would like to thank you for placing it here to start with Carl because I would most likely not have seen it for some time. I don't check the mods section to often.

I think it has been a very informative thread. I know I have learned a few things and my 91's sound better now then they did before.

Link to comment
Share on other sites

Thanks Roy,

I will have to try that. Maybe I'll try changing the stuffing in the 91's to fiberglass one day when I feel like being itchy. LOL

dumb question, but wouldn't replacing the polyfill with fiberglass actually increase damping and decrease system Q, which would thin the lower bass out even more? a better alternative might be removing enclosure volume, maybe temporarily by taping a few smallish sealed shipping boxes or bricks in the cabinet....a system Q of .8 over .56 will "fill in" the bass a little bit in the 50-100hz range at the expense of maybe 1-2hz of -3db....

Link to comment
Share on other sites

dumb question, but wouldn't replacing the polyfill with fiberglass actually increase damping and decrease system Q, which would thin the lower bass out even more? a better alternative might be removing enclosure volume, maybe temporarily by taping a few smallish sealed shipping boxes or bricks in the cabinet....a system Q of .8 over .56 will "fill in" the bass a little bit in the 50-100hz range at the expense of maybe 1-2hz of -3db....

Hi Pat

The difference isn't great A target Q can be achieved with either type of stuffing by simply varying the amount. In my experience, however, Q can measure the same for 2 identical speakers, one filled with poly and the other with fiberglass, and still have different subjective bass characterisitcs. The fiberglass filled cabinet tends to have a fuller upper bass response, and seems more controlled. When we were gathering data for the AR-3a restoration guide, we found that 10oz of poly-fill and 20oz of fiberglass produced a Q of .8 in one of our test cabinets. We also found that adding poly-fill dropped Q at a faster rate than fiberglass. On the other hand, the cabinet had to be emptied or very stuffed to produce significant changes in measured Q.

As Carl mentioned at the beginning of this thread, reducing the amount of poly-fill seemd to help. Harry should try that first before purchasing fiberglass.

You may be onto something regarding cabinet volume. The original AR-12 inch woofer was used in a 1.4 cubic foot cabinet...a bit smaller than the AR-91 cabinet. I've never measured system Q at much less than .7 in any AR-3a size cabinet, regardless of (ceramic magnet) woofer version or stuffing differences. Of course, as Carl mentioned, the woofer inductors in the crossovers were over twice as large as the AR-91 counterpart. The one constant has been that the earlier ceramic magnet woofers always measure higher.

Roy

Link to comment
Share on other sites

Thanks Roy for answering that. This does raise another question for me though.

Would getting the most amount of bass possible by changing the inductors through the balance of the sound off?.

No, Harry...Not unless you are planning on designing a new speaker system. Doing something like that would require changes to other crossover components as well. :unsure: Replacing crossover components or drivers with those not original to the design is a recipe for disaster. I think the earlier AR woofer and some playing around with stuffing is as far as you should go. If it proves unsatisfying, it is time to move on to other speakers.

Roy

Link to comment
Share on other sites

Roy,

That's the point I was trying to make. I think there should be a limit to what mods and tweaks are done to a speaker. The objective should be to get it as close to original as possible.

I don't want to change the sound of my 91's, I like the how they sound. The AR-11 woofers and playing around with the stuffing is all I will do.

Getting them to the best they can sound as they were designed is all I want.

Link to comment
Share on other sites

Roy,

That's the point I was trying to make. I think there should be a limit to what mods and tweaks are done to a speaker. The objective should be to get it as close to original as possible.

I don't want to change the sound of my 91's, I like the how they sound. The AR-11 woofers and playing around with the stuffing is all I will do.

Getting them to the best they can sound as they were designed is all I want.

Sorry, Harry, maybe I misunderestood you...It didn't seem like the point you were trying to make. :) The suggestions made in this thread have been in answer to your questions. Your 91's probably sounded as close to "original" as possible before you swapped woofers. Nobody is saying you *should* make changes.

It has been firmly established above that the AR-91 is a low Q speaker system. They were not designed to have the bass character of the AR-3a or 11. Changing the stuffing and/or the stock inductors certainly won't bring you closer to an "original" AR-91. Using an earlier version of the AR 12 inch woofer was/is a pretty safe thing to do, but doing any of the other things we've discussed would be clear modifications of the original design.

Roy

Link to comment
Share on other sites

About 9 years ago I dropped off AR woofers for Bill LeGall of Millersound to work on and compare. http://www.millersound.net/about.htm

His opinion was reduced cone mass, stiffer spiders, and (possibly) higher magnet strength of the later woofers accounted for all of the differences.

Roy

Hi Roy,

LeGall's comments are in line with what I was saying many years ago on here,

it is interesting how many small changes can add up to a very different sounding woofer.

Do you have a number for the change in cone mass?

Remember I actually put the cone/VC assembly on a scale for an AR11 woofer and

reported the number on here somewhere, I think it was 79g. Do you know the correct

number for the 60s and 70s woofer moving mass? I calculate about 100g but that

involves an educated guess for a few parameters.

Link to comment
Share on other sites

A closed box woofer is a 2nd order high pass filter and the total system Q

or Qtc for the woofer in box is directly tied to the amplitude response by a

simple formula for the relative SPL at the closed box resonance:

SPL = 20 log Qtc

Qtc = .707 (-3dB at Fc) is what is referred to in filter design theory as Maximally flat

it is the highest Q with absolutely no peaking in the response, sometimes referred

to as a monotonic roll off.

Qtc = .5 (-6dB at Fc) is referred to as Critically damped because it is the highest Q

possible with no overshoot in the step response.

Qtc = 1.0 (-0dB at Fc) was liked by Villchur and Kloss and was loosely referred to

as "flat" down to Fc. There was slight peaking above Fc but small, it was flat in the

sense that the amplitude response was down by 0dB at Fc relative to the woofer

passband. Here are the numbers for the attenuation at Fc relative to the woofer passband:

20 log (.5) = -6.02 dB

20 log (.707) = -3.01 dB

20 log (1.0) = -0.0 dB

These are huge differences, consider that a 3dB difference is twice the power and

6 dB is four times the power. So, if you push 100W into a system at Fc with a Qtc

of 1.0, it will take 400W to produce the same SPL with a Qtc of .5. The system is

much less efficient at box resonance with a lower Q - all else being equal.

Many audio people do not use the terminology correctly when they say that a Qtc lower

than .7 is overdamped. Really from a mathematical standpoint it is lower than .5, anything

higher is underdamped. Many audiophiles view critical damping as best, however there

is such a loss in bass output at Fc that this cannot be overlooked in my opinion. I prefer

.7 or .8 in systems like these. Rooms ring far worse that any small difference in transient

response.

Link to comment
Share on other sites

Pete, WT2 tests produce a T/S value, Mms with units of grams for the mechanical mass of the moving diaphram. It may not be identical to the actual gravimetric weight you got using a scale, but it may be useful for comparative purposes.

For example, looking thru some of the WT2 test records for the 11" AR woofer, a vintage AR3a Alnico magnet Mms was 99.8g. An AR3a ceramic magnet woofer Mms was 107.5g and the AR91 I refoamed was 99.7. Not all that different I'd say.

As luck would have it, I weighed a cone, VC, spider & foam surround assembly from and LST woofer and got 82.7g. The spider and rubber surround wasn't all there due to cutting from the frame. I'd guess and say the weight would be close to 90g if there was a full spider and surround.

Link to comment
Share on other sites

Mms is the total "system" moving mass and includes the radiation impedance

mass load which is about 6 grams for that woofer. We measure Mmd on a

scale or the moving mass of the diaphagm which excludes the radiation mass

load.

Our figures are not very far off, I'm not even sure if an LST would have a different

woofer than an 11 - anyone? The mass difference theory could be wrong, not

sure because it is difficult to know based on these measurements of such old

drivers. I did measure with the spider and I also measured the foam and as I

recall it was very light, perhaps 3 grams. We would not want to add the full value

of the surround since only about half of it moves.

It would be good if anyone who removes a cone/VC assembly weighs it and

reports here. I'm happy to take blown drivers from anyone who has them and

try to take some measurements.

By the way, the best way to test a woofer in system is to bring out wires directly

connected to the woofer, measure the DC resistance of the woofer inductor and

add that in series. Then measure with a woofer tester and you will get the actual

system Qtc. The full crossover network introduces some error, Fc will measure lower

than actual by a few hertz due to the added inductance of the XO coil. Not sure

how the cap will alter the measurement but it being designed for the top end of

the response and we are measuring at the bottom end suggests that it will not

completely throw off the measurement.

Link to comment
Share on other sites

Sorry, Harry, maybe I misunderestood you...It didn't seem like the point you were trying to make. :) The suggestions made in this thread have been in answer to your questions. Your 91's probably sounded as close to "original" as possible before you swapped woofers. Nobody is saying you *should* make changes.

It has been firmly established above that the AR-91 is a low Q speaker system. They were not designed to have the bass character of the AR-3a or 11. Changing the stuffing and/or the stock inductors certainly won't bring you closer to an "original" AR-91. Using an earlier version of the AR 12 inch woofer was/is a pretty safe thing to do, but doing any of the other things we've discussed would be clear modifications of the original design.

Roy

Roy,

No need to apologize, I have a habit of not being very clear. Only half of what I'm thinking comes out when I talk and it's worse when I type. LOL

This thread has helped me to understand more about how little changes can effect the sound of a speaker.

For me doing things like this is how I learn. I can read the instructions a hundred times but until I do it, I don't really understand them. I also don't tend to do things I can't easily undo.

I do like the sound of the 91's with the 11 woofers. I don't know if they will stay there,I may put the originals back. I also may change the stuffing in them to see if I can hear a difference. If I like the difference I may leave it that way or maybe I won't but in the end I'll know what the difference is.

You and Carl and many others are very knowledgeable and the more I read the more I learn but until I do some of the things you guys talk about, I don't really know......... and I want to know.

I hope this was clear, it took an hour to type it. LOL

Thanks,Harry.

Link to comment
Share on other sites

Hi Roy,

LeGall's comments are in line with what I was saying many years ago on here,

it is interesting how many small changes can add up to a very different sounding woofer.

Do you have a number for the change in cone mass?

Remember I actually put the cone/VC assembly on a scale for an AR11 woofer and

reported the number on here somewhere, I think it was 79g. Do you know the correct

number for the 60s and 70s woofer moving mass? I calculate about 100g but that

involves an educated guess for a few parameters.

Hey Pete,

Glad you showed up. This stuff is right down your alley! :) I do remember your comments about the AR 12 inch woofer, which have been right on the money.

I don't have any new information on cone mass, but I have been seeing lots of woofers lately, and will attempt to gather more info. I've also been intending to take a closer look at the new AR 12 inch woofer re-cone kit. Of the other basic measurements to date, the biggest differences between earlier and later versions of the ceramic magnet woofer have been for Qms and Zmax, which are both consistently three to five times higher in earlier woofers.

Roy

Link to comment
Share on other sites

Hey Pete,

Glad you showed up. This stuff is right down your alley! :) I do remember your comments about the AR 12 inch woofer, which have been right on the money.

I don't have any new information on cone mass, but I have been seeing lots of woofers lately, and will attempt to gather more info. I've also been intending to take a closer look at the new AR 12 inch woofer re-cone kit. Of the other basic measurements to date, the biggest differences between earlier and later versions of the ceramic magnet woofer have been for Qms and Zmax, which are both consistently three to five times higher in earlier woofers.

Roy

Thanks Roy,

The good thing about the Qms differences is that they don't matter much at all.

The higher Z peak at resonance makes perfect sense since that is how Qms is

measured by the height of the peak. Ideally, we want Qms as high as possible

it is one of two components that contibute to the total woofer system Qts (and in

system Qtc) the other being Qes. Once Qms is significantly larger than Qes, say

by a factor of 5 or 10 then we reach a point of diminishing returns - making it any

larger has virtually no effect on the performance.

I've been under the impression that all the large AR woofers have metalic formers

of some type, Bronze alloy in the very early ones and Aluminum later. Do you know

if there was ever a non-conductive former? The other thing to look for is differences

in venting around the voice coil, is the dust cap porous, do they all have the vent

holes at the top of the former. Qms is a measure of the mechanical resistance but

there is no way to exclude things like air drag, or eddy currents in a conductive

former. Generally, drivers with conductive formers (which are good for cooling) have

lower Qms.

Also, most of the replacement voice coils are wound on high tech non-conductive

materials such as Nomex. Perhaps those woofers had their voice coils replaced. I

would try to avoid non-conductive formers since they will not have the thermal power

handling of the correct conductive type.

How do the very early ALNICO woofers measure as far as Qms goes?

Recone kit? I'll have to take a look for that. I'd be very nervous about using such a

kit unless it has been very carefully tested. It is very difficult to have a 12" cone go up

to 500 - 1K with minimal response deviations so even if it looks the same there is a

very good chance it will have response issues unless it was made in exactly the same

way. Good replacement voice coils, spiders, and a jig to assemble it all onto an

original cone with proper alignment is the best way to go.

Link to comment
Share on other sites

Gents, I have some conflicting test data. It's from the same three tests I referred to in my last post on the Mms cone ass'y mass. The two latter tests - ceramic magnet woofer and the AR 91 woofer had Z max.s around 28 ohms and QMs's around 2.5. Whereas, the older, cast frame, Alnico magnet 3a woofer had a Zmax of 51 ohms and Qms of 5.2. BTW, on the older woofer, Fs was a very low 15.4 hz. I suspect a very tired spider was the main contributor.

The LST cone ass'y I weighed yesterday has a hard pape or nomexr former. I've only seen the metal formers on the oldest of 3a woofers. I have 3 spare 3a recone kits in my inventory and weighed one of them (i.e. cone w/surround, VC and spider) and got 52 g. The cone is definitely lighter and thinner than original.

With regard to non-metalic formers being a concern about cooling, AR took steps to address this via perfing the periphery of the VC with holes just above the windings and below the spider. The combination of the holes and very porous spider served to vent the heat. Not the absolute best, but, I suspect, effective.

Link to comment
Share on other sites

This brings back some memories. I think TT said that the later woofers used a Nomex

non-conductive former yet, the main woofer that I was looking at from an AR11 had

a black anodized aluminum former. This woofer was an AR factory replacement from

way back and perhaps they provided a more robust driver to people who where having

failures. I think another member here who had refoamed quite a few ceramic magnet

woofers from AR11s and 9s also said that everyone he saw had an aluminum former.

So? What's going on here?

I believe that the industry has gone to non-conductive formers because they handle

abuse better and so there are fewer returns and failures. However, this is because as

they are driven hard the VC heats up and draws less power, a self limiting condition.

The VC wire and Nomex former use modern materials that can handle very high temps

and the non-conductive former does not wick the heat into the glue joints. However,

such VCs simply do not have the heat capacity of the conductive types and even at

half the power below abuse levels will heat up more and compress. If you know when

to back off on level or listen at moderately loud levels the conductive type are better.

Carl: I would not use those recone kits it will no longer be an AR woofer, even if you

add mass to the cone, it will not break up the same way as the original.

Have you worked with many AR11 or AR9 woofers to see if they went back to AL

formers?

Link to comment
Share on other sites

I have a couple of woofers here for repair....one alnico and one VERY early ceramic magnet version (notice the cone damping ring, which was carried over for a short while). See photos...The alnico magnet woofer had a metal former, and the ceramic version had a paper voice coil former until much later. I don't know when it was changed back to metal. I don't beleive I have seen any 1970's ceramic magnet woofers with a metal voice coil former. The attached photos show a typical failure of the paper former....tearing apart at the top.

I just dug out more data...It appears the alnico magnet woofer Qms and Zmax fall right between the early and later ceramic magnet woofers, and are consistent with Carl's numbers above. They also had very low Fs. I recall Tom T. saying the early alnico magnet woofers were designed that way. (Later ceramic magnet woofers specs called for Fs to range from 18 to 21hz.)

Qts of the alnico woofer is consistently closer to.2, and the ceramic magnet woofer closer to .3. Carl, your Qms and Zmax meausurememts above are also consistent with mine for most versions of the ceramic magnet woofers....but not the earliest.

I knew that replacement cone had to be quite a bit lighter. I've used it on 2 woofers with destroyed original cones.. Subjectively speaking, it has leaner bass response and excessive midrange response. It is best to salvage the old cone whenever possible.

Roy

post-101150-0-58781900-1350669575_thumb.

post-101150-0-81962300-1350669587_thumb.

post-101150-0-35479800-1350669598_thumb.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.


×
×
  • Create New...