Thiel CS3.6 speakers vs. High Power Amplifiers

[wrf 4,980]

Both Papajoe (actually his friend) and I have experienced driver failures in Thiel 3.6s recently.  I thought I would capture the details and a pretty good diagnosis of the problems in a new thread.

 

SPOILER -- It's NOT Carver amps causing the problem.  Its NOT the age of the amplifier.  It's the crossover.

 

Papjoe's failures occurred with a TFM-42 amp (TFM-45 with no meters).  Mine occurred after a succession of amps, but the M-1.0t mk II opt 2 was the one holding the bag when the failures occurred.  Bear in mind that the Thiels are very inefficient speakers.  They are rated at 86db.  REALLY inefficient in the grand scheme of things.  That means that at even moderate volume levels, you are pushing some serious power into them  Crank them up high, and it can get crazy.

 

All that power goes somewhere.  Some into the drivers, some into the crossovers.   The Thiels have three completely independent crossovers mounted on two boards.  All three are loaded with capacitors, inductors and resistors.  Some of the resistors will get HOT under heavy load.  Unfortunately, the early Thiel 3.6s mounted capacitors right next to and on top of the potentially HOT resistors.  The resistors are actually touching the caps.  That means that when the resistors get hot, the capacitors will melt.  When they melt, they short - perhaps intermittently.  When they short, the drivers are exposed directly to the amp.  Way more low frequency energy than they were designed to adsorb. This results in physical damage as well as heat damage to the drivers.  This is especially the case with the mids and tweeters in the Thiels.  See photo below for my collection of melted caps.

 

Since the Thiels are so inefficient, you tend to run them at fairly high power levels.  That means that even moderately high listening levels will generate lots of heat in the crossovers.  Lots of heat means melted caps which means failed drivers.  This doesn't happen over night.  It's cumulative.  Over the space of 25 years it adds up.  Pretty soon a cap shorts and a driver fails.  It doesn't take much to kill a driver after a cap shorts.  

 

After the failure, somebody gets fired up and fixes/replaces the driver without checking the crossovers. BANG.  It happens again.  Everybody points fingers at everybody else.  It's the amp!  It's Bob Carvers fault!  The amp needs a recap! It's clipping too much! The dealer lied! Thiel did a crappy rebuild job!  NOT!!!!

 

How can you tell if your crossovers suffer from this problem?  Visual inspection is good, but you may not want to tear your expensive speakers apart to reveal the crossover boards.  An easy check is to disconnect the amp from the speakers, and remove the driver.  Remove as in completely disconnect the wires.  Measure the resistance between the positive connection of the speaker and the positive connection to the amplifier.  The tweeter should be about 20 ohms.  Significantly lower or higher means trouble in the crossovers.  The midrange should be about 17.5 ohms.  Woofers should be pretty low at 1.5 ohms or less.

 

Thiel eventually figured this out and started building their crossovers with some space between the heat prone resistors and the capacitors.  But there are sill a lot of 3.6s out there with the problem.

 

 

 

Here is a Before photo of the tweeter crossover in one of my Thiels.

 

 

 

 

 

 

 

Here is one after the rebuild.  Note the spacing.

 

 

 

 

 

Here is my collection of fried capacitors (and resistors) from all six crossovers.

 

 

[aslan7 954]

Wayne

That was a fantastic post that helped even the electrically challenged like me understand the issue. Those pics are also worth a 1000 words.

 

Thank you for sharing your wisdom and knowledge my friend. I love learning from the "greats".

 

[RodH 4,866]

Maybe physically larger, higher wattage resistors would help spread out the heat load as well.

[Papajoe 221]

Wayne,

 Thanks for taking the time to explain the technical end of my problem. The photos were the icing on the cake. We will be removing the drivers and crossovers and sending them to Thiel service to be evaluated. Thiel service had rebuilt those same drivers. If he finds that a failing crossover, which should been serviced last time, was at fault, he will rebuild them again for free. He will also rebuild the crossovers as Wayne did.  More about this after the removal.

Papajoe

• 

[wrf 4,980]

Thought I would add that after rebuilding the speakers, I've had no trouble.  They have been powered by TFM-45's, M-500 mk II, M-1.0t Mk II opt 2, TFM-75 and a couple of Mitsubishi amps.  All good.

[fill35U 1,847]

FWIW, during the Stereophile review of the Thiel CS3.6 back in '93 ( HERE ), they managed to blow apart a midrange driver. Even more interesting is that Thiel had already provided a spare midrange with the speakers...

 

Thiel certainly did innovative stuff! I can't seem to find much info on that midrange driver. Did it have a special name? I think the hollow cone design is really cool- any cross-sectional views? Why was it abandoned and never copied?

[4krow 6,633]

   I have looked at this post numerous times and the same question keeps coming back. Why would anyone ever pack a 10 watt resistor into a stack of other components and not think that there would be serious trouble down the line? It comes to my mind to at least get these firecrackers off of the board by at least 1/2" and let them breath! It even occurs to me that I might have mounted them on the opposite side of the board to shield most of the heat from the unintended victims in the first place. Not to mention some kind of heat shield between the resistor and the Masonite board (flammable). I don't think that I am out of line for pointing this out. I should also say that Thiel speakers are among the best that I have ever heard, and the build quality was excellent. This came as a bit of a surprise to me.

[loner_t 2,568]

 I don't think that I am out of line for pointing this out.

 

This one seemed to do very well in terms of a layout, modularity and component replacement. 

 

 

[4krow 6,633]

LT,

 

 That is absolutely a stellar example for the kind of layout that would be practical. Thanks for sharing. 

[PDR 1,194]

The Leonidas adjustable crossover.....pictured above....is what I wanted to use in my Trios.

I thought the  $650.00 US DOLLAR   price tag was just to.....pricey

 

Why I'm going the DSP route. 

[loner_t 2,568]

 

 

The Leonidas adjustable crossover.....pictured above....is what I wanted to use in my Trios.

I thought the  $650.00 US DOLLAR   price tag was just to.....pricey

 

Why I'm going the DSP route. 

 

Absolutely. It is bliddy expensive.  

[fill35U 1,847]

 

 

This one seemed to do very well in terms of a layout, modularity and component replacement. 

 

 

 

Looks to me like the two inductors are mounted way too close together.

[Papajoe 221]

WRF- thanks for the info that everything is going OK after the crossover and driver rebuild. We have been trying to figure out what to do with all the power and the Thiels? I thought of using the 'NASA' TFM-42, which has Carver installed volume pots, to keep the power under control. We will see.

We will be having the crossovers rebuilt by Thiel and I will attach a photo of what they have done, when they get back. As many of you have noted, that crossover board is just too crowded. It is interesting to me that the woofer and tweeter share the same board. Easy route to disaster? After removing the boards from the cabinet, that was the first thing that struck me, why so packed together? There is lots of room in that cabinet to spread the components out or add additional boards. Even my wife, who has seen only a few crossovers, said "Didn't they realized that putting them so close together was going to cause a problem?"

Papajoe

[xavionics 2,094]

Even my wife, who has seen only a few crossovers, said "Didn't they realized that putting them so close together was going to cause a problem?"

Papajoe

 

Sounds like she's a keeper

 

For comparison, I drove a pair of Klipsch KG-4's (rated for 100 watts each @ 6 ohms) with a m-1.0t (rated 200 watts/ch @ 8 ohms) for more than 2 decades. Big amps don't kill speakers...