Pair of Hi-Vi bookshelfs for about $100 in parts (SS6N and Q1R)

[owdi]

My two previous experiences with Hi-Vi products have been positive (trying the B3S, and owning the M200s), so I decided to use them for my first loudspeaker project involving a passive crossover. My budget for speakers and crossover was only $100, since this is a learning project.

I got this idea after seeing the buyout of the Hi-Vi SS6N at Parts Express. Out of curiosity I poked around for a matching tweeter. My criteria was low fs and ferrofluid for easy crossover design, and shielded or neo for placement near a TV or computer monitor. This left a very short list of tweeters within my budget, and I chose the Hi-Vi Q1R for the nearly flat impedance curve.

Some may think I'm taking a risk purchasing drivers for which I haven't seen any reviews or independent tests. Those same people would think I'm crazy for winging the crossover design and buying the components based on guestimates :-)

Below is the parts list. I made the assumption that the SS6N will have flat frequency response to 3khz, and no major break up modes that require a notch filter. Should be OK since it's not a metal cone woofer. The single inductor on the woofer will serve double duty as a lowpass crossover and baffle step compensation circuit. This design requires a narrow baffle to avoid a bump in midrange response. If I end up with a midrange dip, I will unwind the coil to lower inductance.

For the Q1R tweeter I chose an inductor and cap which should give me a 2nd order highpass filter at around 2500hz. I'm not sure about the amount of padding, so I bought three 16 ohm non-inductive resistors per tweeter. Either 2 or 3 wired in parallel should do the trick.

If anyone has experience with the SS6N or Q1R, please post. Also, I'm not sure what effect mixing 1st order and 2nd order filters will have on phase. If someone can offer some input, please post.



HI-VI SS6N 6" SHIELDED POLY CONE WOOFER
HI-VI Q1R 1-1/8" TEXTILE DOME TWEETER
JANTZEN 2.70mH 18 GA AIR CORE INDUCTOR
JANTZEN .60mH 20 GA AIR CORE INDUCTOR
DAYTON DMPC-6.8 6.8uF-250V POLYPROPYLENE CAPACITOR
DAYTON DNR-16 16 OHM 10W NON-INDUCTIVE RESISTOR
DAYTON BPGS-25G BINDING POST w/1" THREAD 2 RED/2 BLK

Dan

[owdi]

Some progress...

I decided to build floorstanders instead of bookshelfs. The extra height allowed me to make the baffes as narrow as possible without sacrificing internal volume. They will be 8" wide, 5.5" deep, and 36" tall. Net volume after port displacement is about half a cubic foot. According to WinISD, tuning the port to 44hz will give a -3db at 55hz. Below is the modeled low end response, and a picture of the baffles. Tomorrow it's finally time to glue!







They are cut from pine, and will be reinforced with MDF. Not the best choice, but I like the grain.

Dan

[sreten]

Hi,

using just series resistors for your tweeter will play havoc
with the 2nd order crossover frequency and alignment.

The amount of tweeter attenuation you need is also the amount of BSC.
I reccommend you use an L-pad arrangement fvor the resistors.

see http://www.mhsoft.nl/spk_calc.asp

your values are pretty near L/R 5ohm 2.5Khz for the tweeter
and you should L-pad towards achieving an effective 5R load.

Typically for your c/o arrangement tweeter polarity is reversed.

/sreten.

[xstephanx]

This is the true essence of speaker design! although if you reinforce that pine baffle with 3/4 MDF, id say its a more than adequate baffle assuming you brace it normally.

im not a crossover expert, but what you might wnat to do, is make the first order to where it soudns good, but if youre getting too much breakup, calculate the f3 of your crossover and replace with a 2nd order.

also you can always find tonnes of broken electronics and get resistors/capacitors/inductors even transformers from them, and if you have the room to store this stuff, it can be very handy!

[owdi]

Originally posted by sreten
Hi,

using just series resistors for your tweeter will play havoc
with the 2nd order crossover frequency and alignment.

The amount of tweeter attenuation you need is also the amount of BSC.
I reccommend you use an L-pad arrangement fvor the resistors.

see http://www.mhsoft.nl/spk_calc.asp

your values are pretty near L/R 5ohm 2.5Khz for the tweeter
and you should L-pad towards achieving an effective 5R load.

Typically for your c/o arrangement tweeter polarity is reversed.

/sreten.

Agh, you're right, thanks. Six months ago when I ordered the parts I thought I could pad the tweeter by wiring a resistor in parallel only. Once the glue in my enclosures dries I will take some measurements and order more non-inductive resistors.

xstephanx - I definately do not have the room to store all that stuff. I'll consider a 1st order network after I get in box measurements. I suspect it will not be feasible without adding a seperate BSC circuit.

Here is a screenshot of the modeled baffle step, with sharp corners. Tweeter is offset to flatten response as best as possibe (after 4 hours of tweaking).







Here is the screenshot with .75 inch edges.






The rounded edges really clean up the tweeters response, especially off axis. Response stays within 1db above 3khz up to 30 degrees off axis.

The modeled off axis performance between 1 and 3khz is all over the place for the tweeter, don't know what I will do about it.

-Dan

[owdi]

I made an lpad out of three 16 ohm non-inductive resistors, wired everything up, and here is what happened.








That crossover definately needs some work. The 4khz peak of the SS6N demands special attention, a single inductor just isn't going to do it.

Dan

[owdi]

Revision 2 of my filter network. I changed the lpad on the tweeter from a 8ohm and 16ohm resistor to a 3ohm and 6ohm resistor, which raised the crossover frequency. I initially added a zobel to the woofer, but found a 2nd order electrical filter was more effective at knocking out the 4khz peak. I think it's a bessel, with an f3 of around 800hz. Despite the low f3 there isn't much of a hole in response, and that 4khz peak is -12db instead of -3db.



There are 4 curves on here, which is a bit confusing. The individual response for the tweeter and woofer are pretty easy to spot. The system response is in the brigher green, and the reverse null is below it.

I changed the gain on the microphone between measuring individual and combined response, to make it easier to see what is going on. Levels are not comparable to my earlier chart.

From comparing the reverse null to the system response it's obvious I have some phase issues. Right now I'm not confident it's a real problem, since it could be a result of the location of my microphone.

Dan

[owdi]

Here are some system pictures, with my little 10" sub in the background.







Dan

[69stingray]

Great write-up, keep it coming. Those pine front bafflers look great.

[Dryseals]

Good work.
I purchased four of those SS6N during the parts buyout. I wish I had purchased four more. I've played with them in several different configurations and settled on a small ported enclosure, 3 way set up for my sons room. I bought several of the small Vifa midranges when they were dumping them at $1.98 a piece and added in a Dayton tweeter. I'm not in the shop right now but I'll get the numbers later.
I'm finishing the front panels today and should start final testing, I'm still working through SoundEasy, but should be able to post some findings soon.