• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

walsh tweeter design/rebuild

Well, I realized I still have the old poly caps and midrange resistor networks I took out of the 2000AXT's, so I didn't have to wait to test. I hooked up both the one I repaired and the "good" one that came with the WTLC's through a 2.7 ohm resistor and 3.5uF cap to my BA-F1, which has fast power meters (@ 8 ohms) and individual channel selectors and attenuators. Both sound the same and with my head between them (they are only about 20" apart) the voices/cymbals image dead center with the attenuators set the same.

I was surprised to hear Buddy Guy and BB King's voices as well as the cymbals and the upper guitar notes, but I guess with a 1st order HPF they actually "play" far below the 10khz xover point, just not as loud. I think I'll let them play like this for a few hours to see if either is going to fail right away.
 
Well, now I not only have a full gold coned set for the WTLC's, I have a spare pair of silver cones for the 2000AXT's. I decided to open the bad one for a look and discovered it was also broken at the glue dot where the positive wire exits the coil. Also like the other, the glue was still holding it in place until I poked at it. This time the wire didn't go under the negative wire so it was quick and easy to unwind one coil, route it to the terminal and solder. On both that I've repaired I wrapped the excess wire around the terminal to hold it in place for soldering. While working on this one I noticed the glue holding the cone on was partially separated so I put some super glue down inside the joint between the original donor tweeter dome and the Walsh cone.

After that went so well I decided to try reattaching the cone that was knocked off in shipping to the good voice coil. Took some time but I got the cone straightened out and shaped at the bottom to mate with the donor tweeter dome. After I got it stuck in place initially I used a sharpened pencil (can't use anything magnetic that close to the magnet) to work the end of the cone and the remains of the dome to fit tightly, then put more glue down in the seam. A flowable epoxy would have been better but the super glue was all I had that doesn't dry semi-flexible, plus it has a long, thin tip on the bottle to reach down inside the cone to the dome/cone joint.

I tested both with my makeshift HPF. I'm getting pretty good at this, even with poor eyesight and shaky hands. :)
 
When you decide to make this a cottage industry let me know. I've got a number of open VC but otherwise fine Walsh's that I wouldn't mind getting renewed.
 
From the sample of two open voice coils I've done, plus the various threads that inspired me to try the "unwind one coil" repair, I suspect the culprit for this failure is the glue dot. Either the glue creates a hot spot or more likely, a single point of flexure where the wire exits the glue that eventually fatigues the wire. Don't know if it's just coincidence that both mine failed at the positive wire (which joins the coil closest to the dome/terminal). Being on the "top" (with the cone down) of the form the negative wire exits the glue more in the direction of VC movement, whereas the positive lead wire exits the glue at 90* to VC movement.

I saw no evidence of excess heat anywhere. Someone had mentioned being afraid of "vaporizing" the wire while soldering to the tab. I had to maintain 800* solder heat for quite a while (and several heating cycles) to get the varnish to burn off enough to effect a good connection. I was afraid of nicking or thinning the wire by trying to scrape or sand off the varnish. No apparent ill effect on the wire from the heat. Heating copper will anneal it to dead soft condition, which makes it less likely to crack. I'm not sure 800* is enough though. I've heated used copper crush washers to glowing with a torch so they can be re-used. I'm not a hoarder, I just hate throwing out anything that might one day have a use.

Coastsider, the cold dark months are fast approaching here so maybe I'll be contacting you. Working on other people's stuff is a fairly guilt-free way to procrastinate on my own projects.
 
FrankPA said:
Coastsider, the cold dark months are fast approaching here so maybe I'll be contacting you. Working on other people's stuff is a fairly guilt-free way to procrastinate on my own projects.
Click to expand...

I'll await your PM. I still have a couple of special "Walsh tweeter ready" shipping boxes and I'd love to be your first customer. I really do admire the way you've risen to this challenge, especially after my botched attempts.
 
superb work indeed, frankpa. not only proved it can be done once, but multiple times. and a successful cone transplant as well. i had faith these could be repaired. there is just not that much to them. sure would like to see some pics of the process.


meanwhile, i've been having some fruitful dialog on the sidelines re: the tweeter. will post more in a bit, when i've got my head around it.


.
 
I noticed someone a few posts back mentioned replacing the center support "rod".
Do not use a ferrous or magnetic material. We originally used a hardwood dowel secured with 5 minute epoxy, and it looks like Infinity was using aluminum tubing with a relatively thick wall. It also looks like they may have used something like "liquid aluminum" as the adhesive.
A mass of magnetic material near the magnetic gap will distort the flux of the magnet and do nasty things to the response.
In a phocon with Bill a couple of weeks ago, he strongly advocated using ferro-fluid in the gap. The last two tweeters built incorporated ferro-fluid and were placed on a two-way "Walsh" system prototype that never reached the market.
The investor ran a big cabinet shop in Vegas, and was only interested in selling his cabinets- he refused to allocate funds for further development or refinements, so the project was scrapped. The difference between High Fidelity and "steereo" was not appreciated.
The prototypes ended up in my living room for a few years until I discovered ESL's and gave them to a friend.
 
Last edited:
Funny you mention the ferrofluid. I saw ferrofluid packets for sale the other day (PartExpress catalog?) and wondered if it would be a good idea to add to these. The center support rod is clearly aluminum tubing. There is some blue writing on both of mine that looks like it's from the tubing manufacturer. The rods on my WTLC pair aren't even cut to the same length. One is 4", the other is .140" shorter.
 
loquatious asked whether I used glue on the foam caps. I left the remnants of the glue after shaving off the old foam with an exacto knife. Initially it was a friction fit on the new foam. I pushed and pulled on the foam until it was evenly flush all the way around. After a couple days I noticed some spots on the first ones I did that aren't "perfectly" even with the top of the cone. The old glue now seems to have bonded to the new foam.
 
Thanks. When you say "to try", do you mean try in my 2000AXT or try to repair? My 2000AXT's now have a full complement, plus a full set of spares. :)

If you want to drop off your dead one on your next visit to PA, I might just eventually, one day, before I die (maybe) get around to it. Or we can pop it open and have a go while you watch/assist. Two extra hands would be nice. Having some idea what I'm doing seemed to reduce the repair time quite a bit on the last two.

That auction is a relisting, 1st time around it was $250/$350. Since the seller at least seems to understand how to test AND ship, I was watching them before I managed to fix my 4 dead ones.
 
i may attempt to do it myself, i wish you had taken pics, but believe me i understand why you didnt, i forget too when im bawls deep in a speaker or piece of electronic equipment, and pics are usually the last thing on my mind.
 
The main reason I didn't take pics is because mine would have shown the same as Coastsider's pics (only more dirt). His are pretty high res. My camera just doesn't do high-mag macro well enough to show any details of the wiring details I mentioned. I couldn't see the details with my face right there, unless I used the 30-power loupe. Of course my work on them tended to be "stop, squint and poke as I walk by", so of course I wasn't thinking about pics. I'm not a smartphone guy, so taking pics means pulling the camera case out of the closet... Hmm, I just re-upped on BT, maybe I should list my 30-yr old SLR film camera setup.

I've been thinking about buying a USB microscope/probe. If I do, I think I'll have a chance to see how it does taking pics of Walsh wiring.
 
I got a chuckle out of your comment about your 30-yr old SLR film camera setup. I have two 30+ year old Canon F1-N camera bodies and a variety of fixed focal length lenses and accessories stored in the same closet as all of my audio gear spares. In fact I actually made my living selling my photographs for 15 years before having to go back to work in IT after the economy took a dump in the early 90's. I don't do smart phones either - the Walsh photos I posted were taken with a very nice and easy to use but not nearly state-of-the-art 12 megapixel Canon SX130IS.
 
Mine's an 8 MP Canon A590IS, so there's another reason I wouldn't have bettered your pics. I envy you having been able to spend 15 years doing what you love for a living. As it turned out, photography just wasn't ever going to be that "thing" for me. I just checked, I still have 2 unused rolls of Fujicolor 400 film in the meat keeper drawer of my fridge. I put them there sometime before the turn of the century. :)
Now there's a phrase I never thought I'd use!
 
I noticed in some photos a few posts back that the "cores" of the tweeter were Al dome tweeters that had the dome cut out very sloppily. I suggest this is very bad, as the ragged edges remaining will cause distortion due to unwanted resonances, maybe even buzzing.
If you are attempting a repair, I would advise that you attempt to remove as much of the ragged aluminum remaining as you can, very carefully. Perhaps a drum sanding attachment on a Dremel tool could be employed.
The difficulty in getting a clean removal of the dome was the main reason why we opted for mylar domes. You could remove the mylar dome fairly easily with an Exacto knife
 
All six of the ones I refoamed had the same jagged cut domes. Removing all the dome would leave little to glue to. The domes are such thin aluminum they'd be hard to cut any better. I suspect a Dremel would just mangle the remaining foil.
The fiberfill should suppress any buzzing or resonance of the remaining dome. Plus it's inside the cone, beneath 4" of fiberfill and foam. I certainly didn't hear anything untoward while listening to them "alone", connected to the amp through a 3.3uF cap and 2.7 ohm resistor. I wouldn't care to hear them without fiberfill or foam though.
 
in my quest to educate myself about the design and function of these tweeters, i've been trying to seek out people in the know. i always prefer facts to speculation, and first hand info to hearsay. you all know i solicited jay's contributing to this thread. since jay is focusing on other things, i contacted john strohbeen at ohm acoustics, but he knew nothing of the tweeters as they were from before his involvement with ohm. i learned marty gersten (who worked with lincoln walsh, and founded ohm acoustics after leaving rectilinear) was involved with developing the production tweeters, but he is dead, as is bruce zayed (ohm's chief engineer at the time). lincoln walsh did experimental work on tweeters, but he has also long since passed. no doubt arnie nudell could shed considerable light on the subject, but i haven't gotten past his firewall yet in terms of a dialog.

i have been having some email conversations with dale harder from HHR exotic, who is the most knowledgeable person i've found on the subject. dale comes from a background in lasers but, (like arnie nudell, whos' background was also in lasers) his passion is speakers. as some may know, dale's speaker company makes its own advanced design walsh speakers (really nice btw, i recommend you check them out) and he's done his homework on all things walsh. he was a good friend of marty gersten, and was also able to interview bruce zayed, and he has gleaned much info about the infinity walsh tweeters. he has been very generous with his time and knowledge, and very patient with my questioning. i'll save you all the back and forth of our emails and just present info specifically relating to the infinity walsh tweeters. these are dale's responses edited and condensed from the quite drawn out question and answer sessions, which frequently veered off on tangents related to walsh drivers (but were not always specific to the infinity tweeters). these comments are specific to the infinity tweeters. the bracketed bits i added to maintain context.


----------
dale's words:



Repairing these drivers [infinity walsh tweeters] is a very simple matter…, really. The beauty of these drivers is that they worked without a great deal of Voodoo Magic. It was simply an understanding of how and why they performed. For what it is worth we [HHR] can re-manufacture them. However, since [your thread] is a DIY thread and forum, most probably want to try it for themselves. I realize there are many out there that have some skill.

Marty Gersten and Ohm Acoustics actually developed the [production] tweeters, but licensed them to Infinity. Most of the prior work had been done by Lincoln Walsh himself. In fact, some of his units reportedly went as high as 40KHz. They are Transmission Line Drivers and not Dynamic Drivers. They do not move air like a dynamic speaker. The are more akin to the Bending Wave driver. The key is to have a very thin cone material with low mass, but extremely high rigidity. Low mass for ease of acceleration. Rigidity to hold its form, and high density to allow for very fast propagation through the cone material. In the case of the Infinity tweeter, propagation speed is 3 to 4 times the speed of sound in air (1,128 ft/sec).

The cone angle is a bit of math, but suffice it to say the angle is chosen to produce a cylindrical wave front that is perfectly vertical in nature emanating from the cone. it also, assures time and phase coherence. I know that the original tweeters used aluminum foil off the shelf. Some was diamond foil and some was scribed. The purpose of the diamond or the scribing was to stiffen the foil, since most Aluminum foil off the shelf has been annealed and is very soft. this does not bode well for the bending wave driver and can cause the cone to fold up or crush under high loads. The production drivers were very simple consisting of one layer of foil with about .0015" to .002" thickness. X-max [linear cone excursion] was on the order of + / – .004" without damage.

Remember, back in the 70's aluminum was readily available off the shelf and was relatively low priced. All the wonderful tempered and alloyed aluminum and foils we have today were not readily available,...and since this was the beginning of this type of driver, even the designers did not know what to use. It was mostly trial and error. Lincoln Walsh himself, investigated, papers, felted paper, aluminum, plastic, and various other materials. Also, looking at the tweeter, even though it was relatively soft material, since it could [economically] be stiffened by mechanical means and it did not have to support anything, it worked reasonably well.

Many [dampening] "goo's" consisting of silicone grease, oil, rubber and felt were tried including a double layer cone with damping between the layers. There never was any "goo" used in the production drivers.

The center posts inside the tweeter were for strength to prevent the cone from folding over or damage if touched, since that had very little actual movement capability. They were non-magnetic and lightweight. [The aluminum tube's] resonance was also very high. I am not convinced that they had thought this out that far at the time. Necessity dictated simply grabbing what was available and low cost at the time. [The supports were] hard epoxied to the VC and the spider was simply an embossed foil to provide some movement.

The open cell foam acted to damp the cone from ringing, provide structural support to the center tube and to help with the air/metal interface where the sound hopefully was absorbed at the annulus of the cone. If sound is not absorbed here, when it reaches the end of the cone and perceives the air/metal interface, physics demands that the propagating wave be reflected back down the cone 180 degrees out of phase. The 3/4" foam plug at the top was glued to the cone at the terminus and it was to provide the damping and absorption of the wave at the air/metal interface and prevent reflection of the wave back into the cone. It was also a strengthening agent to bond the cone to the center support.

The density of the foam does play a role and this is why in our work with all Walsh speakers we tell people, you can't just grab any old thing and throw it in. The foam that was used in the tweeters was an open cell foam, charcoal grey in color and the density of the foam was uniform. Off the shelf foams do not work as well. Some off the shelf foam has a very randomized cell structure with lots of air voids and this foam does not work as well. This leads to break-up and wave cancellation/reinforcement giving the speaker a very muffled or hollow sound. Even muddy, or veiled.


------------



there you go. a few more bits of info into the mix. conversations continue. i'll post relevant additions as they present themselves.







.
 
so i am also listening to a cull compliment of walshes, i took mine apart and it was not too much hassle to get it fixed, it had multiple contact points in the wire which i removed 8 wraps of, and got it back to 7 ohms, then i coated the wire with nail polish and rewound it resoldered it and retested it, it tested and has been testing at about 6.9 ohms, i put the magnet back on and it sounds excellent, i am listening to a Vivaldi chrome four seasons tape on my axt's and it sounds phenominal, i have to say to be honest this was an easy repair compared to some things i have had to do recently like CT-F1000 belts. anyway, i need new foam in the axt walshes, what should i use if they are supposedly so picky about foam? the only pic i happened to get was this pic of the voice coil removed from the magnet.

8C5D60A8-C2ED-49AC-B5D3-3586B70EEE97-2048-000002AC14DDF3E0.jpg
 
You must log in or register to reply here.
Back
Top Bottom