Refurbed KLH Model 30 speakers sound "harsh"

[tgmartin]

I ended up with a pair of KLH Model 30s with no tweets. I acquired a couple of compatible Model 20 tweeters off Ebay and pulled the Model 30's screw-in woofers to gain access to the crossover. The front panel on these is made of fairly coarse particle board and appears to not like having the screws removed and replaced. It flakes off very easily and the screw holes can strip out. I reinforced the screw holes by squirting a whole tube of Super Glue in each cabinet's screw holes while the woofers were out leaving them to cure overnight.

The rear aluminum panels with RCA phono connectors were pulled, larger penetrations were routed out and new panels were made from 3/8" MDF/Masonite. 5-way banana terminals were installed on them and they were mounted with double-stick foam tape and screws on the inside of the cabs to keep the terminals from sticking out of the back. The foam gaskets for the woofers seem to have wandered when installed at the factory so I used a little rubber cement on the front panel between the screw holes to hold the gasket in place while the woofer was being screwed in. Also, another piece of self adhesive foam weatherstrip was applied to the inside "step" routing of the front panel to further seal things. The crossover caps were replaced with Dayton 5% film units, original fiberglass stuffing and protective cheese cloth reinstalled, RoyC's magic surround dope applied, Velcro attached to the grills and they were buttoned up.

These 30s sound kind of "brash" or "harsh" to me compared to the Model 6s that I have. The 30s do seem to have the same tweeters and some similar crossover parts to the Model 6s, minus the low-pass inductor. The Model 30 crossover looks to have been designed to act like the "NORMAL" setting on the Model 6.

Although the Web has many mentions suggesting that the Model 30s supposedly share the same woofers with Model 6, this appears to NOT be the case. The metal frame's outer flange is different and the 30 has the extra "hot cross bun" pattern of glue or something on the front of the cone. I'm not sure about the magnet size or voice coil length but I suppose the Model 30 woofers may not be as efficient as the Model 6 and the Model 30 cabinet has 1/3 less internal volume. If they ARE similar in efficiency I wonder if the woofer is contributing to the perceived harshness by being run "wide-band" ... I can't find anyone remarking about this "harshness" in the Model 17 which also has similar components, smaller internal volume and runs the woofer without the inductor ... any thoughts?

Since I just got them put back together I'm not all that hot to pull them apart again and add the inductors but I'll probably get to it eventually. I'll try to make it switchable and maybe investigate padding the tweet down a little too.

Thanks for reading and I hope there's something in this tome that you can use!

-Tim

 

[Nat]

I just recapped two pairs of 6s, and replacing the electrolytics with film caps definitely makes them brighter. I would guess the smaller volume of the 30s would limit the low end, which also would make them sound less balanced.
The difference between bright and harsh is subjective, and probably influenced by the rest of the system. What electronics and source are you using?

 

[tgmartin]

The Model 6s that I referred to are ones that I recently recapped as well so I'm somewhat familiar with the change to expect. Sure, without any real, measured numbers, it's all subjective. My impression is that the area between approximately 1KHz to 6KHz seems to be "too loud" in comparison to everything else on the Model 30s. I'm listening through a Carver PM600 power amp, Spectro Acoustics 217 preamp, JVC XL-V211 CD player and Technics SL-Q2 turntable w/ Stanton 681EEE cartridge. Nothing exotic and all fairly middle range, by now vintage, type stuff in good shape, for sure. I mostly listen to Dynaco A25s but I do have a number of other speaks. Lots of different program material ... Pop, Jazz, Country, Classical ... I haven't been able to find much objective info like Frequency Response/Impedance/Directionality graphs for these, or most of the KLH stuff. It would be illuminating to see numbers on the individual drivers as I don't have the instrumentation required.

 

[loudnoises]

Did you reseal the woofers? These tweeters aren't harsh, I'm interpreting it as being too loud vs the woofer.

Caps generally drift down, so before your recap the tweeter may have been getting lower frequencies than intended. You've now removed them and it's able to focus on the higher frequency. If the woofer isn't sealed. It's not working as efficiently as it should. So in the end you've lost (really moved around) a bit of low end now, making the highs seem too much. But I think it's probably the woofer not keeping up. Recapping will generally move the crossover point up to is original spot. More pronounced highs isn't unexpected. Before you do anything more (beyond woofer sealing) you need to give your ears some time to adjust, yours are still attached to the old sound.

This is one of the reasons recapping has become such a muddled thing. It's a repair, but so many look at it like an upgrade. It will not improve your speakers, it will (ideally) return them to a more original condition.

 

[Randy_P]

I ended up with a pair of KLH Model 30s with no tweets. I acquired a couple of compatible Model 20 tweeters off Ebay and pulled the Model 30's screw-in woofers to gain access to the crossover. The front panel on these is made of fairly coarse particle board and appears to not like having the screws removed and replaced. It flakes off very easily and the screw holes can strip out. I reinforced the screw holes by squirting a whole tube of Super Glue in each cabinet's screw holes while the woofers were out leaving them to cure overnight.

The rear aluminum panels with RCA phono connectors were pulled, larger penetrations were routed out and new panels were made from 3/8" MDF/Masonite. 5-way banana terminals were installed on them and they were mounted with double-stick foam tape and screws on the inside of the cabs to keep the terminals from sticking out of the back. The foam gaskets for the woofers seem to have wandered when installed at the factory so I used a little rubber cement on the front panel between the screw holes to hold the gasket in place while the woofer was being screwed in. Also, another piece of self adhesive foam weatherstrip was applied to the inside "step" routing of the front panel to further seal things. The crossover caps were replaced with Dayton 5% film units, original fiberglass stuffing and protective cheese cloth reinstalled, RoyC's magic surround dope applied, Velcro attached to the grills and they were buttoned up.

These 30s sound kind of "brash" or "harsh" to me compared to the Model 6s that I have. The 30s do seem to have the same tweeters and some similar crossover parts to the Model 6s, minus the low-pass inductor. The Model 30 crossover looks to have been designed to act like the "NORMAL" setting on the Model 6.

Although the Web has many mentions suggesting that the Model 30s supposedly share the same woofers with Model 6, this appears to NOT be the case. The metal frame's outer flange is different and the 30 has the extra "hot cross bun" pattern of glue or something on the front of the cone. I'm not sure about the magnet size or voice coil length but I suppose the Model 30 woofers may not be as efficient as the Model 6 and the Model 30 cabinet has 1/3 less internal volume. If they ARE similar in efficiency I wonder if the woofer is contributing to the perceived harshness by being run "wide-band" ... I can't find anyone remarking about this "harshness" in the Model 17 which also has similar components, smaller internal volume and runs the woofer without the inductor ... any thoughts?

Since I just got them put back together I'm not all that hot to pull them apart again and add the inductors but I'll probably get to it eventually. I'll try to make it switchable and maybe investigate padding the tweet down a little too.

Thanks for reading and I hope there's something in this tome that you can use!

-Tim

By the way, I also recommend checking out this website, where I found many useful suggestions and interesting entertainment options. I hope you will also find something for yourself and share your impressions.

Keep in mind that differences in cone and magnet design will definitely affect the response and how "brash" it feels. If you’re adding an inductor, it might be worth making it switchable so you can do a real-time A/B test. That way, you can dial in the perfect balance between the lows and highs without ending up with too much harshness. Also, playing around with some light damping inside the cabinet could help smooth out those peaks and make the overall sound feel more natural.

 

[spark1]

In my experience, the KLH 24, 30 and 32 all sound pleasant enough at lower to moderate volumes, but get "shouty" as volume increases.

 

[tgmartin]

The woofers HAVE been resealed using RoyC's Butyl compound. I hadn't had the opportunity to listen to these speakers previous to putting in the replacement tweeters and new caps so I have no baseline for comparison other than a pair of Model 6s that share the same model tweeter, same crossover components as the "NORMAL" position and same size (although not model) woofer. I see no mention of what I perceive as these Model 30's "spectral imbalance" or "shouty-ness" in any other postings about Model 30 sound. Seems they should sound more similar to Model 6 than they do. That's why I'm puzzled. I'm going to have to pull them apart again and investigate further. Oh Joy! Thanks for the observations. Still looking for efficiency figures on the various 10" KLH woofers as it appears there are a few different models. I don't imagine that they would vary THAT much ... but, who really knows?

 

[spark1]

I didn't check other commentary about their sound. Just my own opinion after having owned and restored all of them (and just about all of the other "classic" KLH models).

Good luck with yours.

 

[SmileyFaceEQ]

I find this era of speakers had that style of sound overall in my experience. This includes dynaco, AR, KLH, and Leak. That is quite a generalization. I realize that.

After refoaming the woofers I usually do the caps / test / add stuffing to get more bass and mimic a larger speaker. I tend to add a lot of stuffing - more than one should and incrementally remove it to achieve the desired result.

Sometimes I add some bracing and reseal all the joints.

IN the case of the leaks 600s I upgraded the tweeters to modern soft domes. Much better to my ears in my space.

Running the speakers for an extended period over a few days will help get them reconditioned after the redope or new surrounds.

+1 on padding down the tweeter. It wold be nice to add it via a switch.

You can also experiment with felt around the tweeter on the front baffle.

Playing with speakers is a lot of fun.

 

[tgmartin]

I got around to pulling one of the Model 30s apart again and modified the crossover, adding a .4mH inductor to the woofer and 7.5 ohm resistor to the tweeter, like the Model 6. I put in a couple of SPDT mini-switches allowing these items to be shorted, effectively removing them from the circuit and returning things to a stock condition. I was able to cram everything on the original crossover board but had to make a new connector panel out of 1/8" double-tempered masonite (an old clipboard) to accomodate the switches.

Adding the inductor was inconsequential to my ears. The difference between having it in or out of the circuit was barely discernable with white noise when listening to just the woofer and when the tweeter was added in I couldn't perceive ANY difference. I've read elsewhere that this model woofer supposedly "drops like a stone above 1800Hz". Just running tones and listening leads me to believe that's probably true but I haven't seen any test data.

I found the tweeter pad helped with the quality I called "harshness" but it would be nice if the area above 6 or 7KHz wasn't attenuated. I find that true of the "DECREASE" setting on my Model 6 pair as well. I'm left with the impression that these are actually just kind of "harsh" sounding speakers. It may be that some more tailoring of the tweeter circuit might be in order like possibly adding a bypass cap to the 7.5 ohm resistor allowing some more extreme top end through when padded down.

Any guidance?

 

[SmileyFaceEQ]

Unless you plan on putting them in a museum find modern soft tweeter replacement options.

 

[tgmartin]

Thanks for the suggestion. They ARE in a museum, it's my house!

 

[odcics2]

There are guys on the Polk site that say to avoid Dayton caps because they can be harsh.
Did you listen to them before the cap change?
Personally, I bought a pair of Ohm L speakers that had Dayton caps installed. I didn’t like them, but the guy that bought them LOVED them!
YMMV

 

[ARC Tech-109]

Only thing I can see with the switch across the low pass inductor is more high frequency energy on the voice coil and nowhere for it to go. I have a pair of Model-20's and they do get shouty when pushed but nothing unpleasant until I start coming to their power limits driving them with a QSC 1400.

 

[tgmartin]

odcics2: I did NOT listen to the speakers before the cap change because they had no tweeters when I got them. I checked the original caps and they had drifted enough to where I thought it prudent to change them out while adding the stock replacement tweeters and setting up the new back panel connectors. Guess the Daytons are one more variable to check ...

ARC Tech-109: The Model 30s are like Model 20s and 17s in their STOCK configuration in that they have NO INDUCTOR for the woofer, unlike Model 6s. I was adding one (with a defeat switch) as a test to see what effect it would have on the upper end of the woofer in the crossover zone and I could detect very little impact on that aspect when switching between having the inductor in circuit and having it shorted. NOT NONE, but very little. Seems the woofer on this model naturally rolls off in the same area, at a similar rate as the low pass inductor causes. Leaving out the inductor seems like a good, economical design choice. Again, this is all subjective as I don't have any sophisticated test gear. Just ears and an audio generator. I'm really not driving these things very hard ... "shouty" seems like pretty good way to characterize the spectral imbalance that I sense.

Thanks for the responses!

 

[SmileyFaceEQ]

I just put dayton caps in a set I refurbished.
They sound fine. no issues.
I would love to see a double blind test on capacitors performed almost as much as cables.

Do you have another tweeter you could test in the cabs?

 

[tgmartin]

Tried another naive/bonehead crossover mod on the KLH Model 30 and, at first blush, it seems to provide what I was looking for which was a little more "extreme" top end when the tweeter is attenuated. Adding the 7.5 Ohm resistor and switch to the Model 30's tweeter circuit helped tame the "shouty-ness" or "harsh" quality but seemed to nearly eliminate any "air" or "clarity" left in the extreme top end.

In reading various explanations of how the stock circuit was supposedly operating, the term "bypass" caps came up describing the use of the two 2uF caps in parallel with the 3 Ohm resistors. It may be that this is an incorrect term to use in this case, I don't know. What I read was that while the resistors were attenuating the entire signal, the cap(s) were acting as high-pass filters allowing the only the top end through which is then added back.

Since I had the parts on hand I just thought I'd try what you see in the schematic.



Ignorance Is Bliss! The end result seemed to be what I was shooting for but I'm curious what any Crossover Gurus out there may have to say about this. Again, I have no analysis equipment so I'm just using my ancient ears and it sounds pretty good to me. I have no idea about how these changes may affect the impedance of the system, phase shift through the capacitors potentially causing distortion, etc.

Any takers?

SmileyFaceEQ: I do have some other tweeters around here but I was kind of focusing on trying to work with the stock drivers. I wasn't anticipating any quantum leaps, just hopefully some minor improvement. I was never a KLH fan when I was younger and these were current product (I'm 70) but I'm starting to appreciate the sound of the woofs these days ...

 

[tgmartin]

Got to thinking that the Model 30 crossover mod I did above could be simplified. The previous mod provided the audible effect that I was looking for but I'm unsure of resultant changes in impedance/phase/etc.


The next naive/bonehead idea is that the second 2uf cap's place in the circuit would be switchable so that it could be applied AFTER the 7.5 Ohm "DECREASE" resistor is engaged. If "DECREASE" is NOT engaged the "Lift" switch would do nothing. This adds no extra components so I thought it may have less effect on the electronic specs while still providing the desired increased level at the extreme high end. Of course I'll have to take the damn things apart again ... The woofer screws are gonna need some help!

I haven't done this yet and was wondering if anyone might want to predict the outcome. An added bonus would be data from simulations of the mods!

Well? Again, any takers?

 

[tgmartin]

I guess not. Oh, well. I'll pass this last bit on anyway - As I've been obsessing over this Model 30 project and pokeing around I've noticed a few things that I found illuminating.

It's fairly well known that although KLH advertised that some of their speakers used 12" woofers they were really always closer to 10" in diameter. After abandoning the unit construction, epoxied-in-place approach it appears (from pics on the web) that they made use of just two different types of steel baskets for their 10" woofers. It looks like there was a 10 3/4" diameter basket and a 10 1/4" basket.

The larger 10 3/4" one was used for Models 5, 6, 12 and 23 (maybe late 7s?) with different magnets, cones and maybe voice coils. This basket has a simple flat mounting flange around the outside edge. All these woofs have driven cone diameters of 8". That's without taking into account the "skiver" or surround. Even the older unit construction Model 6 version had only 8" of driven cone surface area. That's about 50 square inches.

The smaller 10 1/4" basket appears to have been used for everything else that had a nominal 10" woofer like Models 17, 20, 22, 30 etc. This basket has a more complicated mounting flange with a "step" or "roll" and supports a driven cone diameter that is only 6 3/4"! That's only about 35 square inches.

There is a lot of misinformation floating around that could lead one to believe that some of these smaller woofer equipped speakers share the same LF drivers as the Model 6. It's true that Models 17, 20, 22 and 30 share the same tweeters as the Model 6 and very similar crossover design and components but they DON'T share close to equivalent woofers AT ALL.

I've been puzzled by what was perceived as a "harsh" or "shouty" quality to the basic sound of the Model 30s compared to the Model 6s. Now, I suppose a major contributor to this is the fact that the Model 30 woofer has much less driven surface area ALONG WITH being in a smaller box AND having a smaller magnet assembly while using the same tweeter and crossover as the 6. I was thinking that the Model 30 and it's cousins 17 and 20 had close to equivalent "motor skills" as the Model 6 on the bottom end but simply in a smaller box. NOT TRUE!

Maybe this is all old news to you, but if not, there ya go ...

 

[tgmartin]

Here it is a couple of weeks later after tumbling to the fact that I needed to get at least a little bit of a grasp on SOME of this crossover stuff since nobody was volunteering , so ...

I took some time to get LTSpice working and stumbled through enough to set up a few ultra simple models of the high-pass side of the Model 30 crossover that I've outlined above. I realize that my simplified models are too crude to reveal the "reality" of the electrical system and certainly not the resultant acoustic output of the whole electromechanical system but ... maybe it would provide some meaningful insight?

Anyway, what I found was that a simulation of the "Model 30 Mod" schematic from above shows the added 7.5 Ohm resistor would attenuate the overall tweeter level by about -3dB. This was expected. First the "Stock" version:




Then, the added attenuation:




The "Mod V2" schematic simulation with the added 2uF cap causes an increase of about +4dB at 15kHz, starting at about 6kHz. This is pretty much what I was initially shooting for and kind of what it sounded like to me. The V2 Mod:





Further, it seems that the "Mod V3" schematic simulation provides essentially the same high frequency "Lift" as V2 without any added components. I have yet to pull one of the speakers apart again to try this out as it would appear to be essentially the same as the V2 Mod currently in there. The V3 Mod:





The components used in LTSpice were just "ideal" caps and purely resistive loads.

I was hoping to make these speakers just a little more to my liking without destroying their historical sound. After messing with this stuff I realized that I probably should have started by simply replacing the caps with non-polarized electrolytics like were originally installed.

If anyone has any pithy comments or observations, SHOOT! Thanks ... there is so much to learn ... every day.