KEF 105.2 Crossover Info?

[bob p]

I'm looking for info on crossovers for the Kef 105.2. A friend of mine has a pair, and one crossover seems to be bad -- it intermittently cuts out. I tried contacting KEF and the USA distributor to obtain a schem and/or service data sheet, but its been weeks without a reply.

Has anybody got crossover specs for the KEF 105.2?

I know at least one person here has a set of KEFs from that era:

Slightly off topic but...

I loaned my daughter my hi-fi speakers (Kef 104 II's at 200W R.M.S.) and she managed to blow the tweeters with a 30W R.M.S. amp - that square wave thing.

She also banged them with the vacuum cleaner, leaked water from her plants on the tops and trained her cat to claw the fronts but it's probably my fault.

TIA.

[Enzo]

Forgive an old man a weak moment. When I saw this thread I thought, hmmm FM radio station programming note?

Radio KEF, 105.2 on your FM dial, now playing your favorite crossover hits: Faith Hill, Garth Brooks, etc.

I usually trace out Xovers by hand, they are not all that complex, and they all tend to be very similar circuits.

Does the whole thing cut out or just one driver?

Xovers are exposed to all the vibration, so certain things tend to fail. In no order:

Solder on everything.

COmponent leads break, and since parts are often glued down, sometimes the lead breaks in the pc board hole and isn't apparent and they work when the lead touches.

Pull on all the leads.

Twist all the caps in case the lead has come free of the foils.

Make sure power resistors have not cracked.

ANy electromechanical stuff like breakers and switches are immediately suspect.

[Sock Puppet]

Great advice from Enzo again there.

I dug out the manual for what it's worth - it shows in much detail a bloke wrestling the speaker out of the packaging and fitting the feet.

No schematic though.

S.

[bob p]

thanks for the help, guys. its too bad that crossover schems are so hard to find. i guess i got lucky when i found those infinity service documents a couple of months ago.

Enzo, you should really do a standup routine. I'm still grinning from the time that you said, "Thanks, I'll be here all week. Try the veal."

AFAIK the intermittent cutout is just limited to one driver, but I'm not sure on that. I've never actually seen the speakers. a friend has them in Royal Oak Michigan, and they'll be moving to Clawson. the fellow who owns them probably isn't as good at tracing the circuit as us Ampagers might be. I might end up trying to help him fix them via remote, kinda like Enzo has done with my Carver amp. like you said, Enzo -- its a lot easier when you can actually look at the broken item, instead of telling the other guy to bring it closer to the phone...

if all else fails, fixing them might require a roadtrip with some test gear -- and it will probably cost more to travel than the speakers are worth. :-/

[Enzo]

The key to any effort is to define the problem, then isolate it. One driver intermittant is a different thing from all drivers. And if the driver is intermittant, we can work on the xover all we like and it won't help.

Since there are things like inductors in shunt across some points - which of course have about zero DC resistance - chasing a problem down cold can be a problem. My usual method with xovers is to apply a signal from my amp to the input, then I have leads from my bench speaker patch panel that have clips. I can clip one to the xover common, and follow the signal through the xover with the other clip. Looking for present on one side of a part and not the other. The tweeter signal will come out my 212 bench cab more than well enough for tracing a loss of signal.


And if nothing else, your guy is about an hour from me. I would be wiling to try and help him while he goes somewhere for dinner, save two trips. Of course then yuo would no longer have an excuse to visit... Detroit... Make him visit you, CHicago is a much cooler deal.

[bob p]

Enzo, are you closer to Royal Oak or to Carson City? I seem to remember you being in Lansing or Grand Rapids. Am I right on that? If that's the case, I'd guess that you might not be too far from my friend after all. He was planning on moving the speakers from his brother's place in Royal Oak to the family farmhouse in Carson City (I think) to make the trip shorter for me. You might actually be on the way from one place to the other. Its good to know that if all else fails, that he can find a good shop locally. I didn't think that it would be worth trucking them from Detroit to Chicago.

[bob p]

on the subject of troubleshooting, when I repaired the 5-way crossover on my Infinity speakers this summer, i ran into a bad 2 uF cap in the tweeter crossover circuit. i was lucky enough to find it by signal tracing. i tried hooking up the output of my HP signal generator and an amp to the speaker binding posts, and tracing points on the crossover board with my oscilloscope. i also looked for signal on the speaker terminals. doing it that way i was able to find the segment of the board that wasn't passing signal. then i tested individual components and found the bad cap.

as it turned out, infinity (the dumbasses) had soldered the cap to the board, and then mounted a big hollow core inductor on the other side of the board. to hold the inductor in place, they put a cable tie around the inductor, through the board, and around the cap on the other side. then they covered everything with hotmelt glue. it was only a matter of time until the weight of the big heavy inductor used the skinny cable tie to destroy the cap. when i repaired the unit, i used a cable tie and glue to hold the inductor to the board, but i didn't go around the new cap. i mounted the new cap separately. its funny -- Arnie Nudell is famous for being a great speaker designer, but somebody in his manufacturing plant used a really kludge method for assembling the crossover boards.

[bob p]

My usual method with xovers is to apply a signal from my amp to the input, then I have leads from my bench speaker patch panel that have clips. I can clip one to the xover common, and follow the signal through the xover with the other clip. Looking for present on one side of a part and not the other. The tweeter signal will come out my 212 bench cab more than well enough for tracing a loss of signal.

Enzo, i just thought i'd ask -- it sounds like your normal approach is to perform audible signal tracing with a speaker, right? If that's the case, are you using a full range multi-driver speaker on your bench? or just a guitar cabinet? IME it can be really hard to hear some signals out of a guitar cab. The xover that I was working on crossed at 10K, and i really needed a scope for that. but even with a scope it can be confusing as the rolloff slopes didn't seem to be as steep as i thought they would be when i viewed them on a scope.

the other thing i wanted to ask about was your "bench speaker patch panel." i'm just wondering what kind of gear setup you're using, just in case i can steal a good idea or two from you.

[Enzo]

Yes, feed an amp to the xover and use a plain old speaker and clip wires for audible trace. My bench speaks are a couple old 2x12 closed back guitar cabs I bought from SOvtek when they first came upon the scene. New Sensor was blowing out the cabs so I got like 10 of them. Very sturdy inch thick birch ply. A couple of them came with Russian crap speakers. 12" guitar speaks with whizzer cones of all things. Half of the cabs I bought empty. I loaded them with pairs of 12" speakers that Seymour DUncan used in the convertibles - nice sturdy EMinence.

DO they do 10k? Not well, but it is a go/nogo test, not listening for fidelity. Not many over 10k xovers in PA speakers. If I hear sizzle, it is likely working, check with a scope then to be sure. My method is for sorting out xovers with a missing output, not for checking slope or freq - on the other hand, those aspects are rarely wrong.


My bench patch is simple. By the way, absolutely nothing wrong with plugging a speaker cord into the bench speaker and hanging the free end on a hook nearby. I have two cabs under the bench, two 250W loads under the bench, and a small stereo receiver tuned to a music station. Each speaker and load connects to a 1/4" jack and a pair of banana binding posts on the panel. The posts and jack are in parallel. Two more sets of jacks/posts are connected to left and right speaker outs from the little receiver.

Those jacks are mounted on a THICK 3x12" slab of plexi, and the plexi panel is itself mounted on the front of my bench riser below my scope and to the side to clear the scope cables. The wires feed out the back of it out of the way, and are routed under the bench to the stuff. The speakers and loads are thus accessible through the panel, and the stereo receiver provides a reasonable low power signal to drive speakers - or xovers. I also have some line level signal trailing out of the record outs of the stereo to use as general test signals, but they are not mounted in this panel.

I made test cables. A number of foot long banana to banana patch jumpers so I can parallel speakers or loads, or even connect the receiver out to a speaker. I can also use them to series loads or speakers. The main cables are maybe 5 feet long made of like 16ga wire. All have dual banana on one end. A couple have 1/4" Male on the other end, a couple have clips, a couple are left stripped bare. Plug the banana end into the panel, then the other end fits most things. I do have others like one with RCA male and some Molex things, but they are special purpose and I rarely get them out. Even have an XLR4F one for old Ampegs.

Of course nearby are also a couple plain speaker cables, and a couple 1/4 female to clips short cables(Slap that on the end of a regular speaker cord and it turns into a long test cord). I don't keep SPeakon cables, I have a couple 1/4" female to speakon adaptors. MAybe eventually...

I like the banana and 1/4' combination cables. For guitar amps, I plug the banana end into the panel, and the 1/4M into the amp output. For PA power amps, I plug the 1/4M end into the panel and then the banana male end is free for the binding posts on the rear of the amp. Same cable gets dual use.

All the cable hang next to me.

One channel of my scope lives with a dual banana on the end of its input cable, and it hangs by the panel - It can then plug into the banana posts of any speaker or load to monitor it. This saves me having to connect a scope probe to the amp under test direct unless I want to - one less wire trailing across the bench. I usually plug it into the receiver out as a resting spot and let the screen show the music going on when it is not busy - customers find it technical looking. And if the banana posts are already occupied, that's OK, banana dual males will stack nicely.


ANd Carson City? Maybe 40 miles NW of me, ROyal OAK not quite twice that to my east. I'd go right through Lansing making that trip, unless they wanted to go through FLint. A lot more interstate my way though, FLint would be a long drive on surface roads. Yes, I am in LAnsing... like it says just left of here under my name on each post box.

[bob p]

thanks for the tips, Enzo. i don't have a permanent setup, as i have so many projects and so little space that i can't leave things set up for an extended period of time -- if i do that one project ends up blocking the other! besides, it always seems that i'm working on troubleshooting something on-site (or on a kitchen table) rather than on a bench. even so, its always helpful to know how a pro does it to make things simpler and more efficient.

i think that the first thing that i'm going to do in following your lead is to build some more adapter cables. one that i have that you didn't mention is the dual banana to male RCA, or the BNC male to male RCA. i seem to use an RCA adapter more often than anything else because lately i've been working on consumer stereo stuff, and they all have RCA phono type inputs. i could see where dual banana to 1/4" would be really handy for working on MI gear.

Lansing, eh? why wasn't i smart enough to look under your name???