High-end Peerless 2Way HDS, Exclusive

[Rounder]

Hello,

I've been planning a high-end peerless two way for quite some time and am getting awfully close to purchasing time. I have settled 100% on the peerless HDS tweeter, the 810921. To match, for a long time i had been considering the 6.5 HDS Nomex cone for bass in approximately 10 litres give or take, vented of course. But as no one has measured distortion and t/s specs available online for them i have recently reconsidered for the better motor design of the HDS exclusive, 830883 to be a better match. Based on the same box size and 1st order electrical crossovers at 2khz i have two main questions.

What tuning would provide a solid bass kick with extension in the 50-55hz range give or take up to two litres in box size. For reference i very much like the impact of Thiel speakers in the bass department. The box will have a medium fill of damping material and a strong box.

What crossover filters would u deem necessary to fix the response of the woofer and tweeter (woofer mainly) to sound good with first order at 2k. (notch filters? zobel? contours?)


P.S. How would u expect this system to sound when compared to a vifa xt25 and usher 8945A combination in the same size box. Is there a better overall woofer for the same price as the HDS exclusive?

I am open to any other comments or suggestions that are relevant.

PLEASE when basing opinions use only measured data from accurate sources... such as Zaph's site.. who has all these drivers.

Thanks in advance

[5th element]

Mark K has a lot of drivers measured on his website too, that you may wish to look at.

http://www.markk.claub.net/index.html

I wouldn't recommend 2khz on any 'normal' tweeter, first order.

[Rounder]

Thanks for the reply,

yes mark k does have some excellent tests and i visit often. as for the 2khz x-over, i hardly feel that the hds is a normal tweeter based on how it measures and i feel confident that the tweeter will not be damaged as i use a first order crossover at 2100hz on a 5 dollar dayton neo without any problems until extremely loud volumes are reached... someone once did a calculation of power handling that stated that the xmax on the tweeter would only be reached at 106db at 1200hz and 110db at 1567hz.

So ill take my chances until it sounds poor.

[5th element]

It has nothing to do with the driver operating within its Xmax.

In all the tests on Zaph and Mark K's websites, the drivers will be operating within Xmax, yet the distortion rises as frequency decreases. If you look at the majority of tweeters, even the most expensive and most capable, you will see that their distortion starts to rise as frequency decreases.

This is what you are wanting to avoid, not running into Xmax issues.

Your little Neo may not hit Xmax, but its distortion is probably high. This is why Zaph consistently says this tweeter will work well above 3.5khz to the lil neo's. The tweeters frequency response keeps useful up till about 1.5khz, but no one would ever use them here, due to the distortion.

According to the excursion spread sheet from Linkwitz, the HDS will manage 89dB @500hz before you reach Xmax.

If your intention is 1st order slopes @500hz your crossover is going to be giving you 12dB attenuation at 500hz. Add this onto the 89dB figure and you get a max of 101dB.

If you were using a 2nd order crossover here then you'd get twice the attenuation, giving an SPL max of 113dB before the tweeter runs into problems.

At least if I've interpreted the spread sheet correctly.

Edit - Sounding bad. It may not sound *bad* as per say, but you may notice the difference if you alter the xover higher or increase its order.

[Rounder]

excellent point, i hadnt taken the calculation that far as to consider what the db max would be a different frequencies. But there is one point where i am confused. Because i am referring to first order electrical slopes, this rolloff combined with natural rolloff of the driver wouldnt the spl be even lower than 89 db at 500hz? or am i missing something. Im not saying ur wrong, in fact i like ur thinking, i would just like clarification for my own understanding.

P.S. The movie rocks.

Oh and as for the distortion, you are correct especially about the my little neo's the distortion is high, but only annoyingly so at loud volumes. But the benefits to this are a really low and simple x-over point that results in uncanny off-axis and imaging, with great vertical dispersion considering its an mtm. Worth the tradeoff by my ears.
Anyway i re-checked mark k's and ZAph's graph and both seem to reaffirm that with this tweeter the distortion remains consistently low at all frequencies. Im sure the tests were also done at around 90db so im not overly worried about the rising distortion with increased spl.

Right?


Any input on the woofer's crossover tho?

[5th element]

Thats what I factored into the equation by adding in the 12dB of the first order electrical filter. The tweeter would be working 12dB less at 500hz, if 2khz was the target. Thus letting the tweeter play 12dB louder then it normally would.

Providing the tweeters frequency response is flat down to 500hz, it can produce 89dB before exceeding xmax, and hence its also producing 89dB at 5000hz too - with a broadband signal.

Now if you put the filter on, you've chopped the response at 500hz down by 12dB, meaning you need to put in 12dB more power at 500hz to reach its limits again. With the filter now on, the broadband signal is playing at 101dB at 5000hz before the tweeter reaches Xmax at 500hz.

This tweeter however isnt flat down to 500hz, not that it matters, but its excursion is still going to increase as frequency decreases. So even if the combined driver and filter response arrives at a 2nd order roll off, the tweeter is still only 'protected' by a first order network.

Remember that music is full of transients. You may be listening at an overall level of 95dB, but if a transient comes along that wants to play at 105dB and at 500hz, your tweeter is out of luck because its going to exceed its xmax.

Yes the results of Zaph and Mark both indicate the tweeter keeps its distortion down very well. So if the device is playing within Xmax a 1st order eletrical at 2000hz shouldn't sound 'bad' due to increasing distortion as frequency does the opposite.

The woofer.

I cant say anything about crossovers 'really' unless I've got measured data in my hands. See post -

http://www.diyaudio.com/forums/showt...29#post1172329

as to why.

BUT giving it an educated guess, a simple coil on the woofer wont do, as its not going to be anywhere near capable of sorting out that rising response. A zobel with the coil, or a 2nd order filter with a small resistor in series with the shunt cap would probably work far better.

How are you going to address baffle step?

[Rounder]

After visiting the FSI in montreal and speaking to many pro designers i remain confident with the chosen parts. And the best systems there continue to be first order designs. So i wish to continue with the goal plan of a first order network. I have a desing for a notch to remove the impedance peak at fs of the tweeter and will also use a resistor for attenuation on top of the cap to rolloff 6 db/oct at 2000hz. As for the woofer which clearly is a harder design to work with i plan to use a zobel to fix the rising impedance despite the excellent motor. The idea of starting the rolloff on the woofer closer to 1400 to help tame the rising response came to mind. Is this an accepted strategy and if not what are the side effects. i also think this alone will not be enough to tame the woofer appropriately so a notch filter to remove the peak between 2500hz and 6000hz. Im looking for any and all comments as usual.

Thanks again

[qwad]

why if you insist on 1st order dont try a series x/over which is almost as simple tho 9db /oct rather than 6 its almost as good if not better cheers TC

[HiFiNutNut]

5th Elements has given very good advice here.

[qwad]

granted, he has, l was merely pointing out another option which
may work better for him cheers TC