Vintage Hi-Fi speakers with crossovers? (how to test?)

[gooseEL34]

Hey all-

I have a set of UK made KEF C20 hi-fi speakers.   Each speaker contains a tweeter (8 ohm) and a 6" driver (8 ohm).  I put my Fluke leads on the terminals and measure 1 meg plus on one and 2 meg on the other.   I understand that the crossover is a cap and resistor network but I wonder if the differences in the two readings indicate problems in the crossover.  I removed the tweeters and woofers and they all measure correctly as 8 ohm speakers (5.6 Ohm on the Fluke).   

They sound ok., cannot see any gooey caps.  Just wondering if I should recap either of them

Thanks
Tim

[retrolefty]

 When analyzing crossover performance you have to move from the DC world to the AC world. Nominal impedance of a speaker is made up of both DC resistance and AC inductive reactance, the product being it's norminal impedance.

 The crossover is a combination high-pass (for the tweeter) and low-pass filter network. Lots published on the subject if you want to research. Complete testing and evaluation takes an assortment of test equipment.

 That all said your checks seem to me to indicate that you don't have anything to be concerned with absent any symptoms.

[gooseEL34]

Thanks

[DrGeoff]

Since the impedance of the drivers varies with frequency, it is usual for a Zobel compensating network to be attached to the speakers (in a reasonable quality system).
This network provides the crossover with a more stable impedance, allowing the crossover filter to work as intended.

[Richard Crowley]

I put my Fluke leads on the terminals and measure 1 meg plus on one and 2 meg on the other.

Huh?  What does that mean?  Does "one" and "the other" mean the input terminals to the crossover?
What are you measuring these TO?  What is the other end of the meter leads connected to?
Or are you saying this is the DC resistance reading of the woofer driver?  Your verbal description is very confusing.

Very strongly suggest acquiring or creating a schematic diagram of the crossover. It should be simple enough to draw a circuit just by looking at it.

As @retrolefty said, DC measurements of AC circuit components is of very limited value.

[Mark Hennessy]

I agree that the original post is confusing.

As a 30-second search didn't turn up any evidence of these having bi-wire terminals, then I suspect that he measured one box, and then the other.

But how to explain the high DC resistance?

Most likely, there is a capacitor in series with the woofer in the crossover, converting the bass roll-off from 2nd-order to 3rd-order.

From my quick search, I was unable to confirm this, but I know that KEF did release a few 3rd-order designs over the years. It's relatively unusual, but possible.

If so, then what he is seeing is the difference in leakage in the series capacitor that is making the system 3rd-order.

Either way, there is nothing to worry about 

[SoundTech-LG]

If we are to assume you are measuring across the input terminals of the cabinet, those megohm measurements are so ridiculously far off there is certainly a problem somewhere. On an 8 ohm rated cabinet at DC measuring on resistance scale, you will typically measure a DC resistance of UNDER 8 ohms. The woofer consumes gobs of current at low frequency as it approaches the DC level, which is why a faulty amplifier with rails shorted to the output terminals will FRY your woofer Voice Coil. Take a look at this rather complex KEF Xover, and you will see inductors in series with the woofer, and capacitors in parallel to block High Frequency from the woofer. DC will pass through the inductors as a low resistance. You either have open circuit somewhere, or open components (inductors) or open Voice Coil on the Woofer.

By the way Woofers do not get series capacitors unless you are protecting them from say, below 20Hz signals. Not something you typically see in a Full Range speaker system, and certainly not a subwoofer cabinet. IF there are fuses they are generally accessible near the speaker input terminals on the OUTSIDE of the cabinet. Never seen in crossover mounted fuses, and KEF does not do that.

http://usr.audioasylum.com/images/4/40632/KEFc80x-over.gif

[Richard Crowley]

If they "sound OK", then why would you consider screwing around with them?
The kind of capacitors found in speakers like that are not known for much failure.

Never attempt to repair or replace equipment which is working to the customer's satisfaction.
-- IBM service manual

If it ain't "broke" don't attempt to "fix" it.
-- Common Sense

[SoundTech-LG]

Well it certainly would be strange if they work, and give that kind of measurement. The output level would be so weak no one could hear it because it would be quieter than a pin drop... Better check your measurement technique, or your meter.

[Mark Hennessy]

By the way Woofers do not get series capacitors unless you are protecting them from say, below 20Hz signals. Not something you typically see in a Full Range speaker system, and certainly not a subwoofer cabinet.

KEF made several smaller models with series capacitors in the woofer circuit. Like I explained earlier, it converts a sealed 2nd-order system into a 3rd-order system (or ported 4th-order system into 5th-order). It's a recognised approach, but not often done because of the cost of the large capacitors.

Here's an example: http://www.hifiloudspeakers.info/Anatomy/SpeakerSystems/CarltonII/CarltonIICrssvrSchematic.html

[CatalinaWOW]

What are your pas fail criteria for these speakers?  I have never seen a crossover failure in old speakers.  Occasionally have found connectivity issues around the connection terminals.  Particularly the screw lug type as apparently some of the owners felt that they needed the same torque as the lug nuts on their wheels.  Have rarely run into fatigue failure in the leads to the voice coils.

So overall, rarely have I found electrical problems.  But most old speakers I have encountered are junk for one of a number of mechanical reasons. 

1.  The spiders and other surrounds have failed.  This failure is extremely common.  It is also usually repairable, and not too difficult.  Vendors for replacement parts are easy to find. 

2.  The cones have been punctured by any of number of things.  Not sure how this happens given the speaker cloth over most speakers, but it is common.  Re-coning a speaker is possible in principal, but always difficult and it is virtually impossible to duplicate the quality of the original cone in a high end speaker. 

3.  Finally, water damage to the cones.  Either direct or just the result of condensation from humidity and temperature changes.  Same issues as repairing punctures.

Is this set of speakers worth worrying about any potential electrical problems?

[Richard Crowley]

@gooseEL34 said "they sound OK".  So there is nothing apparently wrong with them.
He asked "should recap either of them"?  To which the answer seems quite resoundingly NO.
And his measurement methodology is suspect, or at least confusingly defined.

It would appear that @gooseEL34 is simply suffering from free-floating anxiety.
And continued discussion of what might be wrong with them is just fueling the anxiety.

[SoundTech-LG]

By the way Woofers do not get series capacitors unless you are protecting them from say, below 20Hz signals. Not something you typically see in a Full Range speaker system, and certainly not a subwoofer cabinet.

KEF made several smaller models with series capacitors in the woofer circuit. Like I explained earlier, it converts a sealed 2nd-order system into a 3rd-order system (or ported 4th-order system into 5th-order). It's a recognised approach, but not often done because of the cost of the large capacitors.

Here's an example: http://www.hifiloudspeakers.info/Anatomy/SpeakerSystems/CarltonII/CarltonIICrssvrSchematic.html

Like I said. Very unusual and expensive to be putting 600uF worth of capacitors in series with each woofer. Leave it to the superbly designed, and built KEFs. 9 times out of 10, NO SERIES capacitors there in your garden variety cabinets. That is not going to be considered a full range system, probably meant for biamp, or triamping, although the actual LF cutoff is going to be very low. At least it's much harder to fry the VC, and maybe that was the point. Ever price out a KEF woofer???

During  DC resistance measurement at the cabinet terminals, you would not see megohm readings for a very long while, during the charging of those capacitors.

"The low pass section consists of a RLC resonance that provides a peak near ~2 kHz and also determines the Q of the system. The 600uf capacitor not only isolates the speaker from the danger of DC from the older transistor amplifier, it provides a peak at 100 Hz to give a deeper and quality bass. "
http://www.diyaudio.com/forums/multi-way/192077-kef-carlton-iii-resurrection-seas-22t-af.html

Flat to 47Hz. Definitely recommend a sub with this cabinet. Take the load off it, and find some bass...

http://www.kef.com/uploads/files/en/museum_pdf/80s/Carlton_II.pdf







"

[TMM]

Looks like a single 600uF electro, I doubt they spent more than a dollar on it given the size:
http://www.ebay.co.uk/itm/Pair-Vintage-KEF-2-Way-Speaker-Crossover-Networks-SP-2053A-C20-/121370250288?nma=true&si=7Mw3IpnyqBB0xyH9DTe2hTenpF8%253D&orig_cvip=true&rt=nc&_trksid=p2047675.l2557

I guess if the speakers are 1980s vintage it would not be too uncommon for the caps to have started to go. Even if they look ok they may have dried up and increased in ESR which will change the tone of the sound. Only way to know is to crack it open and test with an ESR/LCR meter.

[Mark Hennessy]

By the way Woofers do not get series capacitors unless you are protecting them from say, below 20Hz signals. Not something you typically see in a Full Range speaker system, and certainly not a subwoofer cabinet.

KEF made several smaller models with series capacitors in the woofer circuit. Like I explained earlier, it converts a sealed 2nd-order system into a 3rd-order system (or ported 4th-order system into 5th-order). It's a recognised approach, but not often done because of the cost of the large capacitors.

Here's an example: http://www.hifiloudspeakers.info/Anatomy/SpeakerSystems/CarltonII/CarltonIICrssvrSchematic.html

Like I said. Very unusual and expensive to be putting 600uF worth of capacitors in series with each woofer.

So what? In my original post, before you said it, I also said it was unusual:

but I know that KEF did release a few 3rd-order designs over the years. It's relatively unusual, but possible.

And non-polarised electrolytic capacitors built for the loudspeaker industry are not expensive. Probably similar or cheaper than a port. It's a valid design choice.

During  DC resistance measurement at the cabinet terminals, you would not see megohm readings for a very long while, during the charging of those capacitors.

Unless they are leaking very slightly, which is entirely possible. Like I also said in my original post. You did actually read that? It's all there, but it seems that you'd rather be arguing about it than learning from it. Not a problem - your choice

All the best,

Mark

[SoundTech-LG]

Close to 3rd order per KEF support.

No argument on this end. If the caps were leaking, and lost enough capacitance, doubtful they would still sound as good as mentioned, and the low end would be (going, or gone). Also the OP is talking about model C20, not the Carlton II's, but the C20's do use the series caps.

Specification: C20
 System Type: Two-way, bookshelf/stand-mount
 Enclosure type: Closed box
 Size: 340 x 247 x 212mm (13.4 x 9.7 x 8.4 inches)
 Weight: 6.5kg (12 lb)
 Nominal Impedance: 8ohms
 Rated maximum power: 100W programme
 Frequency response: 72Hz to 400Hz +/-3.0dB rising 3dB to 800Hz-20kHz +/-3dB (-6dB at 58Hz)
 Sensitivity: 90dB for a pink noise input of 2.83V (under wall-mounted conditions)
 Maximum output: 110dB on programme peaks under typical listening conditions
 Finish: Simulated Walnut, Rosewood, Black Ash
 System: SP3053
 Drive units: B200 bass unit (SP1200), T33 tweeter (SP1211)
 Crossover: SP2053

http://www.microsofttranslator.com/BV.aspx?ref=IE8Activity&a=http%3A%2F%2Fwww.ebay.it%2Fitm%2FKEF-C20-2-Wege-Lautsprecher-made-in-England-%2F281771109147%3Fhash%3Ditem419addeb1b


Like you said...

 
PostPosted: Fri Sep 19, 2008 9:36 am    Post subject: C20 to C30 modification Reply with quote 
 

C20 used bass driver SP1200 with xover SP2053
 C30 used bass driver SP1206 with xover SP2058

 Although the HF units and xovers were the same the LF xover sections were different, the C20 having a 'Series-C' configuration with a 600uF capacitor at the input of the xover to help shape the LF roll-off to close to 3rd order. This wil be somewhat critical to the natural response of the LF driver in the existing box volume.

 Putting the C20 parts into a larger box will therefore be a bit of an unknown and might need a bit of fiddling with the series C value to get a clear improvement.

 Regards,

 KEF_Support
 
http://hifiloudspeakers.info/speakertalk/viewtopic.php?t=840&sid=6b39c7ca8f6ce2d0301d04f46cb87aae

[retrolefty]

Rated maximum power: 100W programme

 A pet rant to follow, so ignore if wish.   

 What the hell does "100W programme" audio output power mean? Is that unit of measurement like any of the other misleading and/or fraudulent ways of rating audio output power?

 The stereo industry in the US sometime in the 60-70s had to stomp down on misleading consumer on actual audio power output capacity when trying to compare different manufactures models. There were misleading units of 'peak power', 'music power', etc. The FTC then came up with some consumer advertising laws covering the consumer stereo market. Only true RMS units, sine-wave driven and other conditions stated for frequency bandwidth, both channels driven, preheat requirement, etc.

 Now I don't know if the FTC regulations just expired or if the law as written doesn't apply to the 'home theater' industry but when 5-channel stuff started coming out we once again started seeing bogus wattage units used once again. Most 5-channel amps when reading their rear nameplate data state maximum AC power input that can't come close to matching their 'rated' audio output power units if all 5 channels could output such at the same time, so they must have discovered 'over unity' power for audio amp output ratings. 

 Again just a personal rant, but there was a time in the 70s when you could compare output ratings between Marantz, Sansui, Pioneer, Kenwood, etc and be reasonable secure knowing that one was comparing apples to apples rather then apple watts to helium filled balloon watts.   

 

[SoundTech-LG]

Yep.

[retrolefty]

Yep.

Or Yep  ?
Is my rant a righteous one or not?   

[Mark Hennessy]

I did say that my example schematic was just that - an example. But well done for finding that forum post - I only spent 30 seconds searching, so didn't come across it.

And yes, it's very possible for electrolytic caps to leak slightly while still having good capacitance and ESR. See it all the time. In truth, "1M plus" (or 2M for the other one) isn't really much leakage at all.

The OP really has no need to worry

[SoundTech-LG]

Not much leakage but the measurement will approach that final resistance value extremely slowly as 600uF is charging up, which was not mentioned by OP. How could that be missed, and just how long do you get to wait before the meter reads the megohm values??? Calculations...

[SoundTech-LG]

Yep.

Or Yep  ?
Is my rant a righteous one or not?   

Righteous dude...

[SoundTech-LG]

As a quick experiment I put my Fluke 189 (to read DC resistance) across a 1000 uF cap. It charged up much quicker than I thought.. a few seconds until it reached 6 Megohms. So unless you were using an analog meter  , or paying very close attention, you WOULD probably miss the charging of the cap. I still gotta say those KEFs are a pretty unique Xover, and you don't typically see that series cap in the woofer ckt, much at all. 

[SoundTech-LG]

If you are manufacturing 100,000 speaker cabinets and the capacitors run you $5 each ...   well it does add up quickly.

KEF uses the cost is no object approach. Not saying that is bad, or good. But certainly a bit out of step with the norm.

http://www.parts-express.com/500uf-100v-electrolytic-non-polarized-crossover-capacitor--027-378


Spent a few minutes trying to find at least ONE other Xover Ckt. that uses SERIES capacitors. Nope.

See if you can find something besides KEF...
https://www.google.com/search?q=3+way+crossover+design&biw=1280&bih=842&tbm=isch&tbo=u&source=univ&sa=X&ved=0CB0QsARqFQoTCOiwr5aIuMcCFU02iAodPMIF2g#imgrc=_         

[Mark Hennessy]

Obviously there is a massive difference in price between retail and business-to-business.

The 3rd-order KEF designs that I know about are 2-way designs, so no surprise that a Google image search for "3 way crossover design" failed to turn up a result - no idea what you're hoping for with that. "3-way" is clearly not the same as "3rd order"

It's a difficult thing to google for, not least because the term "3rd-order" is more typically applied to the slope of the crossover rather than the bass tuning. Here's one reference - http://www.audiokarma.org/forums/showthread.php?t=186000 - where apparently Infiniti used the same trick. Post #5 is interesting:

There was an article in Speaker Builder (by Tom Nousaine ... IIRC) about this many years ago. The capacitor can (if the woofer/box is designed correctly) resonate with the input impedance and actually provide a bit of boost below Fc, and obviously cut at lower frequencies. The cap and speaker is a 3rd order high pass system. This concept has been covered in the JAES.

Another one: http://www.diyaudio.com/forums/subwoofers/187634-capacitor-coupled-bass-can-sound-tighter-go-lower-frequency.html - have a look at post #8

And here: http://www.audiobanter.com/showthread.php?t=51463

So, not an especially widespread technique - and I never claimed it was - but it's certainly a known practice among loudspeaker designers. The main thing that counts against it is the audiophile's perception of electrolytic capacitors in the signal path, bizarrely. Certainly not cost.