Early Capacitor Manufacturing History

[johnieo]

AR chose capacitors based on what was available. I just returned from a technical meeting in Dallas and was fortunate in locating two 'well-aged' engineers who worked in capacitor manufacturing during the 1950s and 1960s. Here is a brief summary of how early caps were made, and why, like the dinosaur, some disappeared. This give me some idea of how to select replacements for pre-nonpolar anodized capacitors.

Oil-filled capacitors:

=============

In the period ~1950 to ~1964, stable capacitors were manufactured by first rolling layers of aluminum foil / paper / aluminum foil / spacer. Next the assembly was placed in a steel can, which was heated in a vacuum chamber so that residual moisture could be baked out and pump away. (Moisture increases ESR and voltage recovery.) While hot, the capacitor was filled with oil through the two terminal fittings. The process was completed by sealing the terminals with hot solder, cooling and venting. The oil most used was Arclor 1254 (pentachlorobiphenyl). It is one of a class of chemicals known as PCBs. PCB was the favorite oil for filling capacitors and large transformers because its equivalent series resistance or dissipation factor was *extremely* low; this meant low heating during service. A typical value found in my dusty 1956 handbook gives ESR = 0.00035 for the audio frequency spectrum. Great dielectric--except for one major flaw. By 1964 it was identified as a cause of cancer in employees working in capacitor and transformer manufacturing plants. Manufacture of PCB-filled capacitors ceased almost immediately.

AR used these capacitors in its loudspeakers up to that date; that included early serial-numbered AR-3 speakers made through the mid-1960s.

Note that these capacitors, if destined for audio use, could be manufactured as "non-inductive" by simply winding the layers clockwise for half and counterclocwise for the remaining turns.

Wax-paper capacitors:

===============

With the demise of PCBs, capacitor manufacturers turned to a different dielectric--paraffin wax, composed of mostly aliphatic saturated hydrocarbons. These capacitors could employ the same winding techniques, including non-inductive for audio use. The wound package and the paper box in which it was to be housed, were placed in a vacuum system, exhausted and baked to remove all traces of water. (This is very important.) Next hot wax was admitted to the chamber filling the package. Next, the chamber was vented to force the hot, liquid wax into all the intersticies and then it was cooled. A steam pocket in the middle meant a bad capacitor. Parafin wax was chosen, because it too had an extremely low equivalent series resistance--typically ESR = 0.0002, a value somewhat lower than PCB. However, these capacitors were big and heavy--the dual 150/50 uF unit I removed weighted 5.3 pounds! The different weights printed by FedEx on my two speaker boxes is literally reason why I "opened this can of worms." Like PCB-filled units, they were expensive to manufacture. Unlike liquid-filled units, paraffin cooled to a solid and could be contained in a cardboard box. But the days of this dinosaur were to be numbered.

In the early 1970s new, light, very small, low-cost electrolytic capacitors (anodized aluminum) were developed and manufactured. Non-polar electrolytics used for ac are just two polarized units back-to-back within the same aluminum can. When charge flows in one direction, one side is a low resistance and the other a capacitor and vice-versa. Manufacturing cost, size, etc., resulted in the replacement of the wax-paper-impregnated capacitor with the NPE. Their low cost was offset by their very high dissipation factors. For small caps, up to a few uF, ESR ~ 0.02-to-0.05. For hundreds of uF ESR ~ 0.15-to-0.22. This is due to the physical size of the anodized surface, and the fact that large-sized caps are typically made with a lower voltage breakdown resulting in a thinner oxide with higher leakage. These ESR values are so high that it is obvious to me the speaker designer of the 1970s had to include that resistance in the calculation of the crossover frequency, and thus the legitimate concern when they are replaced with low ESR polypros.

AR's switch to waxpaper capacitors seemed to occur during the time when it was manufacturing its late-model AR-3 and early-model AR-3a speakers.

--AR-3: we observe that the extremely low ESR, typical of both PCB and paraffin, meant that there would be NO change in the crossover properties between early- and late-model AR-3.

AR switched from waxpaper to NPE some time in the 1970s.

--AR-3a: we observe that the ESR of paraffin caps was much lower than for NPEs and that the crossover frequency WAS affected--it increased from 525 Hz to 575 Hz from early to late models.

Replacement of old caps:

=================

Based on what I measured and what my elderly friends have taught me, I conclude the following:

Oil-filled: ESRs were originally 0.00035; in 2004 I measured ESR = 0.033 and 0.1, respectively on two units. Not bad for forty years; the increase is small because water cannot permeate steel, rather only leak through the terminal insulation joint. But, now that I know its original value, I see that the change is great enough to affect the crossover---Recommendation: replace!

Wax-paper paraffin: ESRs were originally 0.0002; in 2004 I measure ESR = 0.044, 0.3, 0.8, respectively on three units. Terrible! This is due to the huge long-term water vapor permeation through wax. Impregnated paper caps were designed with a life of 1 year continuous duty at a service temperature of 40 degrees C! Recommendation: replace!

My simple-minded conclusion is that by replacing the NPE caps installed in my AR-3 when it was converted to an AR-3a in 1976, I am most likely returning the crossover frequency to its original 525 Hz. Everything else, inductors, resistor, etc., appears to be the same?

Replacement of NPE capacitors:

=====================

NPE devices will age in their own way. One well-known phenomenon is the decrease in anodized oxide thickness when a high voltage is not applied (soft music or long-term storage). In this case C will increase, and the crossover will most definitely be affected. Elecrolyte could also leak and cause C to decrease. So here one has two choices:

---Replace the NPEs used in the crossover after, say 5 or more years and continue to do so every five or more years.

---OR (here I'm going to get flack I may well deserve) replace with polypropylene caps and a small resistance equal to the ESR.

One advantage of the polypro units is the excellent polymer encapsulation that will most likely result in decades of service with no change in capacitance or ESR. Also, polypropylene is extremely non-polar and therefore has an extremely low voltage recovery (dielectric absorption)--it's about 1 part in ten thousand--so absolutely no volume compression would ever be possible.

My conclusion is that AR, like any original equipment manufacturer, purchased what was being manufactured and this is the cause of our concerns with early AR speakers. I am quite happy with the two 3a speakers now using Solen caps; they produce beautiful sound.

John O'Hanlon, who often wonders why he ever noticed the weights of his two parcels.

[Guest Bret]

>John O'Hanlon, who often wonders why he ever noticed the weights of his two parcels. <

John, I for one, am extremely grateful that you did notice the difference, that you did consider what you were doing, that you did all this background research that I couldn't have, and that you cared enough for your fellow man to report your findings.

I don't know that you'll catch flack for your suggestion of adding resistance to the crossover in-line with a poly cap. I really hope that the only couple of people I would expect to have a knowledgable opinion about this will chime-in; even if it is to give you flack. ;-)

So you think an ESR of .15 would have been taken into account in the crossover's design. Very in-ter-esting (as Arty Johnson used to say).

Okay, so. . . since the ESR of a capacitor changes with. . . is it heat or voltage?. . . now how to I find a variable resistor!?

I'll bet your 3a's do sound great with the Solens. I'm thoroughly convinced that just putting "C" back in spec makes old things new again.

Do you suppose that it is possible to put-together a list of "suggested replacements" for the caps of the 70s with resistor chasers?

Thank you, again. Your post was extremely informative and greatly appreciated.

Bret

[tysontom]

John,

That’s a very authoritative and informative “dissertation” of sorts on the state-of-the-crossover capacitors. It is a scientific explanation, and that is vitally important. As to its recommended implementation… that will have to be the decision of the end-user. Generally speaking, unless one has an obvious audible or measurable problem with their speaker, changing-out capacitors probably doesn’t make a lot of sense. This goes on the assumption that the components are reasonably close to their original specs. Moreover, there are many variables, such as the changes that occur to the adhesives that bond drivers together, the compliance changes of the foam materials used for suspension pieces and a myriad other things in these 40+ year-old loudspeakers that keep them from sounding exactly as original. It is probably a shame that few, if any, of these original or restored AR speakers would now meet AR’s original specs when the speaker was manufactured, but that’s reality and it doesn’t keep us from enjoying the sound of these speakers.

I do have some additional thoughts on your message:

1. The oil-filled capacitors used in the early AR speakers were originally war-surplus items, readily available. The good thing is that they would have been mil-spec, which tends to enhance quality control on such items.

2. AR discontinued the oil-filled capacitors (as you indicated, the oil was a PCB-carcinogenic oil that made the headlines years ago) around 1962 or 1963 and began to use the paper capacitors from ICI in Chicago. This actually happened four or five years before the AR-3a was introduced.

3. The AR-3a crossover changes, as reported by SteveF in an earlier post, note that the changes went down from the original 575 Hz to 525 Hz, with a similar change to the AR-5 crossover. So the changed frequency went down, not up.

4. I still can’t understand how one of your AR-3a’s actually weighed nearly ten pounds less than the other, despite the difference in the 150/50 capacitors. That is a puzzle.

--Tom Tyson

[soundminded]

While we can't necessarily restore every parameter of these loudspeakers to the original, it seems to me that it is a good idea to try to restore as many of them as possible especially where there is a practical way to do it. Generally, multiple changes don't tend to compensate one another but if anything, each tends to make the others worse. I am therefore a little surprised that you would not endorse johneio's suggestion. He has demonstrated both the need to look at these parts for significant changes and has given us a practical method to fix the problem.

The real risk of PCB is if it becomes involved in a fire. The products of combustion include dioxin and difuran, both highly carcinogenic. At least 3 buildings in the US were sealed up after fires where PBC transformers were involved bacuse the toxic fumes got into the masonary and the buildings were therefore considered too dangerous for human habitation. I don't know if they've been cleaned up. It is technically illegal to just throw these capacitors in the trash. I was involved in disposal of 16 large PCB filled transformers about 18 years ago. I am trying to think of a way to get rid of them safely and legally. I'd suggest contacting your local recycling authorities or your power company. That would be a good place to start.

[tysontom]

>While we can't necessarily restore every parameter of these

>loudspeakers to the original, it seems to me that it is a good

>idea to try to restore as many of them as possible especially

>where there is a practical way to do it. Generally, multiple

>changes don't tend to compensate one another but if anything,

>each tends to make the others worse. I am therefore a little

>surprised that you would not endorse johneio's suggestion. He

>has demonstrated both the need to look at these parts for

>significant changes and has given us a practical method to fix

>the problem.

>

Well, it's not that I do *not* endorse John's suggestion. I am sure that there is validity to his comments and those from other fellow-scientists with whom he has had contact, and I believe his comments. However, in the absence of finding exact-type capacitor replacements -- and that is a real issue -- I suppose I am too conservative to go in and do the change on this basis alone. And quite frankly, I have really not seen many AR crossovers (particularly the earlier ICI types) that were notably faulty, at least in my experience.

>The real risk of PCB is if it becomes involved in a fire. The

>products of combustion include dioxin and difuran, both highly

>carcinogenic. At least 3 buildings in the US were sealed up

>after fires where PBC transformers were involved bacuse the

>toxic fumes got into the masonary and the buildings were

>therefore considered too dangerous for human habitation. I

>don't know if they've been cleaned up. It is technically

>illegal to just throw these capacitors in the trash. I was

>involved in disposal of 16 large PCB filled transformers about

>18 years ago. I am trying to think of a way to get rid of

>them safely and legally. I'd suggest contacting your local

>recycling authorities or your power company. That would be a

>good place to start.

I remember an instance back the late 1970s in which an industrial oil-recycling company was dumping -- shall we say "spraying" -- PCB oil along the roadside. For a fee they removed PCB oil from various transformer companies, and then they had to dispose of it. They were using spray nozzles to evenly spray and dissipate the PCB on the shoulders of the road, thinking the oil would be absorbed in the dirt and would disappear. They were doing it at night, of course. State and local authorities quickly determined that something was amiss due to the stripes of dead grass along the shoulder, and apparently followed the trail to a tanker truck that had become stuck in the quagmire of its own oil/sand mixture along a roadside.

--Tom Tyson

[Guest Bret]

>I have really not seen many AR crossovers (particularly the earlier ICI types) that were notably faulty, at least in my experience.<

Have you checked or are you surmising that's the case because they don't sound too bad? If it hadn't been for my reading so many posts here, I would have come to a different conclusion about these 90s I "fixed."

I would have concluded that AR-90s just don't sound very good by today's standards.

Tom, as much as I don't want to have to spend the time, money, or energy to change capacitors, I hate the idea of removing original components even more. (you remember how picky I am about even a tweeter's paint job)

So we've got a bad situation. I lean toward making them "as right as I can" which, if these 90s are any indication, means the capacitance numbers can't double here and there and be ignored (which I realize isn't what you advocate).

Where'd Ken say the AR anechoic chamber is now? Do you think Audiovox has a lab? ;-)

Bret

[soundminded]

At some time before I was involved in the removal of 16 PCB network transformers from a main communications hub and software development center for the Bell System in central New Jersey, I was advised that managers at Todd Ship Yards in Los Angeles had contracted someone to dispose of PCB for them and the contractor brought it to a brass foundry to be burned. Those involved were convicted and got a collective 212 man years of prison time. The criteria for reclassifying PCB transformers to PCB free was to get below 50 parts per million and keep it there for six months. The deadline was October 1, 1988. 50 to 500 ppm was considered PCB contaminated. The units we had had a concintration of 700,000 to 1,000,000 parts per million. In the end, the removal and replacement of the transformers was contracted to GE who had the PCB burned in one of only two high temperature incinerators approved for this purpose in the US. They took title to the empty transformers and maintain liability for them in perpetuity. It's too bad. There is no other known liquid with electrical properties suitable for substitution that can even begin to compare with PCB. (Likewise asbestos but that's another story.)

[tysontom]

>

>Have you checked or are you surmising that's the case because

>they don't sound too bad? If it hadn't been for my reading so

>many posts here, I would have come to a different conclusion

>about these 90s I "fixed."

Bret,

It probably is “surmising” on my part, not because “they don’t sound too bad,” as you say, but because I have not detected audible frequency-response changes or distortions in any of the AR speakers to which I regularly listen. It is entirely possible that something subtle is happening to these speakers, and perhaps I need to check into this more thoroughly. I will try to be less conservative and more open-minded going forward.

Consider the following:

In 1994 after the AR 40th Birthday Party in NYC, I was contacted by AR to send a pair of original, mint-condition, 1967 AR-3as to Julian Hirsch of Hirsch-Houck Labs in order that he could compare them to the then-new AR-303 speaker designed by Ken Kantor (and Bill Bush) of AR/NHT. It had been suggested that this comparison would be appropriate to confirm expected improvements in materials and technology in the intervening years. It was important that the AR-3as perform properly and as close as possible to original. Therefore, before sending my 3as to the late Mr. Hirsch, AR had me ship them to their plant in Benicia, California in order to be carefully tested by engineer Bill Bush. The last thing that AR wanted was a malfunctioning or misbehaving pair of AR-3as for that test. Bush performed a series of measurement tests on both the AR-3a and the new AR-303. Predictably, the results of those tests confirmed the superiority of the AR-303 in terms of overall flatness, especially the 1- and 2-meter, swept-noise test. But it was also clear that *both* speakers were very flat and uniform across the frequency band. As expected, the AR-3a suffered from cabinet-edge molding diffraction effects on the first-arrival, impulse testing, and some anticipated interference effects, but there were no problems noted with overall performance. Once Hirsch received and tested this pair of AR-3as (*Stereo Review,* June 1995), he also noted the close correlation in the sound of both speakers: “…In fact, their composite response curves (a close-miked bass measurement spliced to a room-response measurement made with a sweeping warble tone averaged from the two speakers at a single microphone position) were amazingly alike. The AR-303 had significantly flatter response at the high frequencies, however, about 3 to 5 dB stronger than the AR-3a above 6 kHz….” The AR-303 was flatter out to 20 kHz, and essentially this new design corrected the slightly downward-sloping, high-frequency response of the AR-3a.

I made no changes in the AR-3as sent to Hirsch other than to clean the level controls, something I felt necessary to do before sending them to be tested. Granted, this pair was (and still is) in like-new condition, with little use, and had actually been packed away for four or five years before this project arose. The crossover components -- Industrial Condenser Inc. capacitors -- were all-original, 21-year-old devices. Perhaps subtle changes had occurred in my pair of AR-3as, but if so it did not seem to affect the measured-frequency response or comments on sound quality.

So the question has to be: do I want to go in and change the crossover capacitors based on what we have learned to this point, *if* there are no detected problems? Perhaps I should, but I think we need to be able to quantify the results.

>I would have concluded that AR-90s just don't sound very good

>by today's standards.

>

>Tom, as much as I don't want to have to spend the time, money,

>or energy to change capacitors, I hate the idea of removing

>original components even more. (you remember how picky I am

>about even a tweeter's paint job)

>

>So we've got a bad situation. I lean toward making them "as

>right as I can" which, if these 90s are any indication, means

>the capacitance numbers can't double here and there and be

>ignored (which I realize isn't what you advocate).

>

>Where'd Ken say the AR anechoic chamber is now? Do you think

>Audiovox has a lab? ;-)

It seems that many people mention that they are hearing problems in their AR-90s and AR-9s, so perhaps in this case there are serious crossover problems here. Also note that these speakers are usually played at significantly higher power levels than most of the bookshelf speakers, and this may also be affecting the crossover capacitors.

The anechoic chamber was moved from Acoustic Research in Massachusetts to the AR/NHT facility in Benicia, California. It was installed there. Unfortunately, the reverberant-test chamber was lost or hoodwinked somewhere along the line and no one seems to know of its existence today.

--Tom Tyson

[Guest Bret]

>I will try to be less conservative and more open-minded going forward.<

What's that old saying? Something like "If you get too open-minded you'll soon be empty- headed?" Please don't abandon your conservatism on this subject due to me. In fact, if I thought I caused you to move toward free-wheeling I'd have to either jump off a cliff or go to Confession looking for the most severe penance the Priest could dish-out (and I'm not even Catholic). [Leave it alone, Ken. Gimmes are beneath you.]

Your 3a's tested well, which would seem to indicate that things hadn't changed much internally. I also noticed that Joe has said he's never run across a bad capacitor in his speakers. I've also noticed that those speakers were pre-Teledyne, pre-LPJ "I will not seek. . .", and made with good-ole non-filter carcinogenic vintage components. Shoot, a Plymouth Barracuda still had that strange rear glass and it was possible to buy a real station wagon with fake wood on the sides. Americans hadn't even invented Japanese cars or Jimmy Carter yet! (Or did the Russians invent those first?) I was washing wounds with pHisoHex. [Leave it alone, Ken.] Grandma gave me good-tasting Fresca complete with cyclamates because it was healthier than all that sugar! [Leave it alone, Ken.]

There was a lot of bad, good-old-day stuff that really was pretty good stuff.

I think there are circumstances and instances where doing the best you can really is the best thing you can do. Judging from these AR-90 crossovers, sometimes the best you can do is to replace things.

I'm still not settled on just what the best replacements are. Have you ever tried to find either product data sheets on these older caps or the technical data for the Truth In Listening crossovers? No sense in beating my head against the wall if you've already developed a flat forehead.

I was just kidding about the anechoic chamber, but it really makes me sad to hear that the reverberant test chamber is gone. That's like hearing that we can't build and test a Saturn V rocket engine anymore. No wonder we can't build a pyramid like the ancient Egyptians. Someone threw away the instructions just because the papyrus was getting yellow.

Bret

[johnieo]

Thank you all for your comments including corrections like getting 525 and 575 Hz reversed. Yes, there are many other things that can change in addition to capacitor values!

Getting data is the easy part of a project. Deciding what to do with it is the hard part! We all filter information through our own experience and act accordingly. That's why we come to different conclusions and that's fine. If Tom's speakers worked well when tested in 94, then no further action was needed. My issue was different; my 3a speakers were not the same and I wanted a matched pair so I installed polypropylene caps. They now appear to be an fine replacement for the wax filled caps designed into the 3a, as their electrical properties are much the same.

I'm kinda glad I found the old PCB caps glued to the wall of the 3, which were not removed during conversion. I'm happy to know they are out of the house; will take them to hazmat recycle and not to the landfill.

War surplus PCB-filled capacitors would now be ~60 years old. One of these capacitors (the 'Pyramid' 4 uF) showed a C decrease to 3.7 uF and an ESR increase from 0.00035 to 0.003 in that time. Wow, that's *impressive* stability; too bad PCB was a carcinogen. However, I don't think wax-filled caps will be that stable after 40 years. What I don't know is how large a change in ESR or C is necessary to see a discernable change in performance of a crossover of the type used in the 3a. Someone with modeling software could perhaps demonstrate that? If these filters are like other circuits I have studied, then extent to which a component's drift in value affects overall circuit performance would depend on the individual circuit design--some circuits are designed with just that issue in mind.

Cheers and enjoy listening -- what ever music you like!

[Guest lansolut]

If anyone is still paying attention to this thread in 2007, I would love to find out what value resistor to install in series with the Poly replacement for the Wax caps in the AR-3a.

Did that question ever get answered?

[Carlspeak]

>If anyone is still paying attention to this thread in 2007, I

>would love to find out what value resistor to install in

>series with the Poly replacement for the Wax caps in the

>AR-3a.

>

>Did that question ever get answered?

I'd say from the posts above you're looking at a miniscule DCR - say in the range of .1 to .25 ohms. I you really feel you must get resistors, work out an L-pad arrangement of resistors that will get you that low value. Most cap ESR's change over time so measuring original 3a caps now which may be 30-40 yrs old could give you misleading info. You really need to know the original cap ESR to do the job right.

FYI, below is a link to a web site where a guy has subjectively reviewed the sonics of over 25 different capacitors. No science here, just his personal opinion - but never-the-less intriguing reading.

http://www.humblehomemadehifi.com/Cap.html

It's all about the music

Carl

Carl's Custom Loudspeakers

[dynaco_dan]

Hi Carl;

Thank you very much for the link.

It does make for some interesting reading.

Thank's to Steve, Tom, Bret, etc for keeping this topic active.

[rrcrain]

I've read a few opinions that many, if not all capacitors tend to sound extremely similar after a year or 3 once they've "settled in", have a similar ESR.

Do you know of anyone who's attempted to prove or disprove this opinion?

[Carlspeak]

>I've read a few opinions that many, if not all capacitors

>tend to sound extremely similar after a year or 3 once they've

>"settled in", have a similar ESR.

>

>Do you know of anyone who's attempted to prove or disprove

>this opinion?

I'm not so sure a 'similar ESR' has any significant impact on sound. Certainly a significant ESR shift of an ohm or 2 that can occur with old electrolytics and other types will.

Check out Pete Basil's "Capacitor Myths" post located down the list on the AR discussion home page 1.

He references a study done by a group of knowledgeable DIY's that concluded there were no perceptable differences in capacitors evaluated.

As I've written before: "SONIC BEAUTY IS IN THE EAR OF THE BEHOLDER"

It's all about the music

Carl

Carl's Custom Loudspeakers

[rrcrain]

Thats effectively the same thing I've read in several places as well. I've not experimented with various exotic capacitors and to be honest, don't intend to nor do I have a desire to drop a fortune on so called magical / exotic caps. Granted, they "might" make a difference, but my personal opinion, until proven otherwise, is the exotic caps provide someone a means of having their bank account lightened.

With the home improvement projects we still have to complete, I do not need any assistance lightening my bank accounts.

[RoyC]

Ken Kantor makes some interesting comments about capacitors used in restorations. For those who have a nagging feeling that they have spent too much for little or negative benefit, it is a must read.

http://kkantor.spaces.live.com

Experimentation with small resistance in series with a number of brands of film caps has led me to believe that it has a desirable effect in the restoration of the sound of our old timers. Resistors of those values are inexpensive and easily obtained from places like Parts Express.

Roy

[dynaco_dan]

Hi Roy;

I had the the time to visit that site last night.

Great opportunity to see the business and fun side of, Ken Kantor.

I'll be short with this write-up, heh Ken.

The suggestion of between 0.5 and 5 ohms in the application mentioned is a consideration once the woofer is out and the cap is being replaced.

Basicly a few minutes of time and the fixed close tolerance resistors, not trimmer pots.

Perhaps we can find a similar, listening only test, published with the insertion of ever slightly increasing resistance values and the same caps.

[RoyC]

Hi Vern,

Actually I have been using only .27 and .33 ohm resistors on poly tweeter and mid caps, and .5 ohms on large woofer caps, as I am a bit nervous about over-doing it.

We really don't know the specs of the various types of original caps, and installing a new, reliable cap should be a good thing. Presumably, in the restoration of a "classic speaker", the downside to a new cap would be to affect original crossover points, frequency response and/or phase relationships, resulting in alteration of the speaker's original design or character. This is essentially the same reason that it is not wise to replace original inductors with super-duper low gauge, low resistance types. Fortunately, inductors do not degrade as caps do.

Roy

[dynaco_dan]

Hi Roy;

Thank you for your feedback, Roy.

It may be too much work but have you had a remote switching arrangement, so that you can just switch between different resistances, outside of the box?

[dynaco_dan]

Hi again;

I don't know where to properly post this, so here it is.

It is to add to our pool of information.

In the Dynaco forum, I just added to a Dynaco A-50 topic a few minutes ago.

Dynaco claimed that each step in the 5 position switch, the A-50 was specified, but should also be valid with A-25 and A-35 speakers as well, is about 3 dB+/- @ 10khz.

Dynaco used only 2.2 ohm 3 watt rating sandcast resistors, no tolerance specified but maybe 10 - 20%, for each step.

The tweeter is rated at 8 ohms.

I copied this here to show that a small 2.2 ohm resistor makes a 3+/- dB change in series with this 8 ohm tweeter for reference only.

This adds to the repeatability of settings, which is much more consistant than a pot.

The Dynaco switch, nor the area around it, ever seems to suffer from heat problems compared to the Aetna-Pollock pots, at least in my experience.

The Dynaco switch has exposed contacts as well.

[RoyC]

>It may be too much work but have you had a remote switching

>arrangement, so that you can just switch between different

>resistances, outside of the box?

No, I use all original "reference" speakers (both versions of the AR-3a as well as stock AR-4x, 2ax, and, soon, 5's) with which to compare, as tweeter and mid caps/resistors dangle behind the project cabinets :-(.

I'm not making any definitive claims here, just sharing some thoughts.

Roy