Test run of 2-way Voigt Pipe

[Timn8ter]

I finally got a pair running and spent some time listening today. I started with the idea that two extended bandwidth drivers with a first order crossover should blend better than limited bandwidth drivers with a high order crossover (please comment on that assumption if you so desire). I also wanted to do this with relatively low-cost drivers. I selected the Pioneer B20 8" "full-range" and a new tweeter from GR Research, the GR-T6. I removed the whizzer from the Pioneer, treated the cone with Dammar and damped the basket to try and tame some annoying upper-end problems. The GR-T6 is a wonderful tweeter. I picked it up at the introductory price of $29.95 a pair (matched pairs). I've used a system BDSC network and a electrical XO point of ~5500 Hz and an L-Pad on the tweeter to compensate for it's 96 db efficiency. It doesn't sound bad at all. Tons of smooth bass and nice clear highs. Jazz and Reggae sound great. I'm going to try some different XO configurations to satisfy my curiosity even though that would be deviating from my goal. The front baffle is not glued on yet and testing continues.

[Nuuk]

Very nice concept all round. There is also something 'just right' about that shape (even with the blue tape ).

Where did you get that huge port?

[sreten]

Originally posted by Timn8ter
I started with the idea that two extended bandwidth drivers with a first order crossover should blend better than limited bandwidth drivers with a high order crossover (please comment on that assumption if you so desire)

Hi Tim,

by system BSDC I assume you mean line level passive ?

Just note that series networks have some advantages over
parrallel networks for first order applications, basically not
needing to worry about driver impedance compensation.

see :

http://sound.westhost.com/parallel-series.htm#1.0

A variation is this :

http://www.lcaudio.dk/com/dele.htm

This is overdamped second order on the treble, and
will work fine with the L-pad connected as normal.

Should have lower distortion of the treble unit
below the c/o frequency.

Also see :

http://melhuish.org/audio/response.html#pioneer

sreten.

[Timn8ter]

Originally posted by Nuuk
Very nice concept all round. There is also something 'just right' about that shape (even with the blue tape ).

Where did you get that huge port?

Thanks. Blue and birch do go togother don't they? The port is the flared end of a 3" precision port. I just happened to have two of them and they were the correct length.

"by system BSDC I assume you mean line level passive ?"

Yep, thanks for the terminology correction.

"Just note that series networks have some advantages over
parrallel networks for first order applications, basically not
needing to worry about driver impedance compensation."

Excellent suggestions. That XO from LC Tech is interesting. Thanks for the links. I've got more things to try now.

[Timn8ter]

After looking things over I ripped my wiring all apart and started over. I selected a solid copper core 14 AWG and ditched the L-pad for a series first-order XO at close to 6000Hz and a BDSC network at the terminal cup. The series XO cleaned up some stuff on the Pioneer. One song is kinda giving me fits. Tierney Sutton's "Alone Together" which is a string bass with Tierney's vocals. The bass is being plucked and is very loud. At certain points with the volume up high the slap of the strings along with her voice causes quite a bit of distortion. I'm not sure if it's my amp clipping or the speaker and only happens at high volume. I'll have to try a different amp. That song is a hell of a test track.

[sreten]

and a BDSC network at the terminal cup.

Thats not line level passive, at all. Its at speaker level.

Your losing headroom and the impedance above the baffle
step will not suit driving a crossover designed for voltage drive.

How about active BSDC ?

Find the feed back resistors in the power amp, typical values
are between ~ 220K with 10K and ~ 22K with 1K.

Double the value of the large feedback capacitor to give +6DB
and then add the same value with a series capacitor to bring
HF gain back down to 0dB, with + 6dB in the bass.

For 2.5K and 56K :
replace 56K with 100K and add 100K with series capacitor
in parrallel, Z Capacitor at centre frequency ~ = 100K.

There's no reason to ditch the L-Pad for a series crossover.

sreten.

[Timn8ter]

Thats not line level passive, at all. Its at speaker level.

Yeah, I realized that after I had responded. I'm not thrilled with the idea of modifying the amp as I switch speakers around often these days. I'll try different values with the L-pad. I misunderstood about the "self-correcting" ability of the first-order series XO then. Am I going completely the wrong direction here? Are you saying that my using speaker-level BSC with a first-order XO is not a good combo?

[sreten]

I'm just saying if the BSC circuit precedes the crossover you'll
have half current and half voltage drive above the BSC frequency.

A simpler way of BSC without losing headroom is between pre
amplifier and power amplifier or via a tape loop, i.e. line level.

I don't think your going in the wrong direction, I haven't really
thought about the implications of using BSC attenuation to also
attenuate the tweeter, but it will interfere with the L-Pad and
the calculated crossover components, I think !?

P.S. the crossover I linked to has no BSC, I presumed line level BSC.

sreten.

[Timn8ter]

It's my amp. I hooked up to a different amp it plays great although it's a little bright.

[Bricolo]

Originally posted by sreten


Thats not line level passive, at all. Its at speaker level.

Your losing headroom and the impedance above the baffle
step will not suit driving a crossover designed for voltage drive.

How about active BSDC ?

Find the feed back resistors in the power amp, typical values
are between ~ 220K with 10K and ~ 22K with 1K.

Double the value of the large feedback capacitor to give +6DB
and then add the same value with a series capacitor to bring
HF gain back down to 0dB, with + 6dB in the bass.

For 2.5K and 56K :
replace 56K with 100K and add 100K with series capacitor
in parrallel, Z Capacitor at centre frequency ~ = 100K.

There's no reason to ditch the L-Pad for a series crossover.

sreten.

And how good is this solution?
http://sound.westhost.com/bafflestep.htm#var-equaliser