Can Thiel CS1.2s be upgraded?

[Boden]

I suppose you are referring to acoustic odd order filters.

I am, however, afraid it is not that simple: first of all drivers are usually non-coincident , furthermore there is cone size, and then there is driver-inherent on and off axis behaviour.

Boxsim is remarkebly optimistic in all simulations. Never have I seen nicer and smoother off axis graphs, the off axis results not based on real measurements but -I assume- based purely on piston/disk modelling.

Kind Regards,

Eelco

Regards,

Eelco

 

[system7]

Well, you politely asked a question about third order filters, and I answered it politely.

I was, of course , talking about acoustic third order filters. The main thing to understand about odd order is it overlaps the drivers more than even orders. Hence it fixes the dip in power response. At the price of a certain amount of vertical lobing.

Now if you have any criticism of Boxsim or myself, you can go and get stuffed really. I don't care. The science will win.

 

[Scottmoose]

You know those old classic speakers with BIG boxes and baffles? What is special about them? Mainly that you can get away with low value bass coils. Which has the knock-on effect that you can use nearer Zobel value shunt circuits.

Wide baffles have their advantages, that being one. Ideally, moving toward 30in to shunt the -3dB step-loss frequency sufficiently low to start blending with ~average room-gain is the way forward, if a trifle impractical for most.

So what I am finding is that to get a 8" bass to perform well, you need a big baffle. Which allows you to use some serious shunt capacitance, which sounds better.

Depends how you define performing well, given that this is a floating / subjective point unless specific criteria are adopted. Me, I tend to like very low XO frequencies, which tend to require fairly 'extreme' filters; swings & roundabouts. That certainly works well if the tweeter can handle it. I like some of the wide baffle types too of course.

And something odd then happens, which is that you then need a bass driver with lower Le inductance to work well.

I like low VC inductance, period.

So in general, I am moving away from narrow baffles.

Fair enough, there's plenty in wider baffles to work with & lots of advantages.

 

[lousymusician]

I think I'll focus instead on stereo bass reinforcement for the Thiels. What would you guys suggest? Perhaps a sealed 10" for each side? Or should they be ported since the Thiels are too?

This is a good read:
https://mehlau.net/audio/multisub_geddes/

Adding subs and relieving the little Thiel woofers of bass duty might be the best improvement you could make in the midrange.

Bill

 

[norman bates]

true.

But it is tough to install something that works as a crossover and doesn't degrade the sound.

Norman

 

[lousymusician]

true.

But it is tough to install something that works as a crossover and doesn't degrade the sound.

Norman

Maybe passive line level to roll off the mains and DSP on the woofers only?

 

[bmwparts]

I contacted Thiel and they sent me the crossover layout!

They even keep track of things like wire length -- awesome.

 

[bmwparts]

More from Thiel:

"The PP stands for polypropylene and all of the caps are 1.0 uf. The “s” stands for polystyrene, and those values are 10 or 15 nF. (nano-Farad)."

 

[norman bates]

Yea, .01uF caps bypass the caps.

Pretty slick, then again, they weren't cheap when they came out many years ago.

 

[Scottmoose]

Ah, cascaded bypass. Interesting.

 

[bmwparts]

Ah, cascaded bypass. Interesting.

They call it cascaded bypass? Cool, i've never heard that term before.

Can it be applied to other things like input and output caps on amps/dacs to good benefit?

 

[Scottmoose]

I don't know if Thiel call it that, but that's what it's often known as, i.e. a main value with multiple descending smaller value caps -in this case two. 'Better', 'worse', 'different', 'no difference'? YMMV.

 

[system7]

Cascaded Bypass (Capacitors) have nothing to do with anything important here! A red herring, for sure. Let's focus on the important stuff.

This is a very old speaker with an early metal dome SEAS tweeter. It also has a Vifa P17WJ polycone with copper damping rings and vented polepiece. Very like the famous old Vifa P13WH in fact.

The bass crossover uses impedance correction (10R + 34uF) on the reflex woofer. The 0.65mH coil and parallel 10R resistor apply about 5dB bafflestep around 700Hz. But the woofer is essentially free-running, and relies on its natural 4kHz rolloff to behave.

The tweeter filter is a bit odd altogether. The Fs of the tweeter is slightly notched by the LCR at 650Hz. It then gets a fairly unsurprising first order filter @ 4uF and 6.5R. The 1uF lifts the overall tweeter response about 2dB. The 20R is very odd, even wrong, and I don't know why it is there at all. It lets a lot of bass into the tweeter.

The crossover can be improved IMO. I'd also look into replacing the SEAS tweeter with a new one if it is a ferrofluid design, which has possibly dried out.

 

[sreten]

Hi,

The 20R in the tweeter circuit is very perplexing, but that
doesn't mean its wrong. I'm guessing its lowest value
they thought they could get away with chasing the
6dB/octave acoustic goal, though TBH I don't like it.

FWIW the whole circuit implies messing about with
the tweeters bass rolloff, so under no circumstances
change the tweeter model, in an area where the
distortion is, to put it mildly, not that great,
and tweeter responses vary a lot.

I don't really agree with the principles used in the 1.2,
but for sure I agree with don't change it. If you don't
like them, sell them as original and move on.

The only caps worth changing are the electrolytics.*

rgds, sreten.

* And be wary, its impossible to distinguish electolytics
selected for value from the numbers printed on the case.



Pers

 

[norman bates]

yup.
So, if the surrounds are good, replace the caps.

If it sounds ok and you don't like it, sell it.
Or you may get more for the working drivers on ebay.

It will be bass light unless you are sitting against a wall and they are against another wall.

I (among others) enjoy what they do correctly (time alignment thing), even with their lack of output, high distortion, dispersion issues, and unobtanium parts !!!!!!!!!!!!

hate to break it to you, but even the 3-way 8" thiels have little bass............................
Well, a little deep bass at higher volumes anyway.................

It's a 6.5" or an 8", and it is doing voice. Now you want to beat it up doing bass ?

 

[system7]

It seems to be an OEM version (H0420 25TAC/GW TH) of the SEAS H0417 25TAC/GW:
PRESTIGE TWEETERS.

cf the H0400, it looks like non-ferrofluid impedance profile to me. So that's good. It can't have dried out...

25mm High Fidelity dome tweeter with glass fibre reinforced plastic
chassis. The unit features high efficiency and smooth, extended response.
Its aluminium diaphragm with critically designed shape and thickness
behaves like a piston throughout the audible frequency range, resulting in
good dispersion above 10 kHz. A fine mesh grid protects the diaphragm,
and supports a phase plate which compensates for a slight axial roll off
towards 20 kHz. The specially designed soft surround allows excellent
mechanical linearity. A damped hole in the pole piece and a tuned chamber
behind the magnet result in a low resonance frequency. The unit is
equipped with flexible lead out wires, allowing relatively low crossover
frequencies to be used.

That 20R is just plain wrong IMO. It aligns phase better at very low bass levels, but that's no excuse for poor design which will cause distortion or risk physical damage. Removing it will not change treble levels significantly.

With a transistor amplifier, the bass impedance correction is an unnecessary waste of power too.

This sort of design is actually well documented, cf version 3:
18W-8434G00

I suppose it's a LR2 at heart. I have some misgivings about tweeter phase in the Thiel design too. I'd think negative polarity.

 

[sreten]

That 20R is just plain wrong IMO. It aligns phase better at very low bass
levels, but that's no excuse for poor design which will cause distortion or
risk physical damage. Removing it will not change treble levels significantly.

Hi,

The 20R is right and wrong. I don't like it, and think its wrong.

Putting my engineers hat on, its plain wrong and totally stupid.

Arguably add a series cap to it (33uF ?) might work, but the
effect might be so measuredly meaningless not to bother.

rgds, sreten.

 

[Dissi]

I suppose it's a LR2 at heart. I have some misgivings about tweeter phase in the Thiel design too. I'd think negative polarity.

I rather would say it's a classic first order design (-3 dB and 90° phase shift at the crossover point), optimized for a perfect step response.

 

[system7]

I rather would say it's a classic first order design (-3 dB and 90° phase shift at the crossover point), optimized for a perfect step response.

Classic "Mickey Mouse" first order design IMO.

Good at step response, lousy at everything else that makes a good speaker.

It distorts. HORRIBLY. It's a LGWAG speaker.

I don't like 6" bass at all, but maybe Third order Butterworth can rescue it: http://www.diyaudio.com/forums/multi-way/92628-q-heybrook-hb2-xo.html

 

[moebuster]

"peeking in" on this thread since i have CS1.2's in need of a woofer or woofers and found it very helpful.