Home Theater Forum and Systems banner

1 - 15 of 15 Posts

·
Registered
Joined
·
16 Posts
Discussion Starter #1
Some advice and experience is needed on making decisions about how to piece together a 2 channel, 3 way system, for a 10' x 12' workshop.

There are 4 small monitors from a 5.1 system that I'm thinking of stacking two on each side to act as the midrange/tweeter. Can this work well or is it a dead end. Right now the top speaker is upside down to keep the 2 tweeters closer together. Would staying with a single speaker be simpler and better.

This weekend I finished assembling a pair of cabinets each with a single 12" driver to act as the woofer. A quick test and a few hours of operating time for exercise went well and the new speakers, in the cabinets, were tested and show an Fs of 49 Hz using a DATS. With very little knowledge of REW a few tests were run but I really don't know what I'm doing here yet.

Research on 4th order 3 way crossover design starts to bring up many questions.

1st is selecting the crossover point to use for the 12" driver.

Details- General 4" midranges like the one here are very flat down to about 200 Hz. The 12" is flat up to 1,500 Hz so there is a big overlap here. The polycarbonate dome tweeter is said to be crossed over at 5,500 Hz in the literature but I don't trust it. The small monitor crossover is a one way design with a 4 uF cap and 0.5 ohm resistor in series. Running a calculator says 5,000 Hz is the xover. These little monitors have a ported cabinet and sound very good.

Any suggestions on crossover points and whether using the pair of stacked monitors is a good or bad idea.
 

·
Banned
Joined
·
4,838 Posts
Would staying with a single speaker be simpler and better.
Yes, generally speaking, but we are talking about a workshop here, not a critical listening room. Volume and coverage might be higher priorities, and then your setup might be just fine. If you are looking for truly high-quality sound, then one pair of speakers would be the best approach.

I am not a crossover expert - jtalden, where are you??? - but my suggestion is 500Hz and 5kHz.
 

·
Registered
Joined
·
16 Posts
Discussion Starter #3 (Edited)
Thank you for the input. After reading the speaker evaluation done by the forum here and many comments made about imaging and a couple about millimeter accurate speaker placement your advice is helpful. My thinking behind the 2 pair setup is the point you bring up, one of coverage. High freq. distribution can be evened out somewhat at the expense of definition.

Doing a bit more research on crossover design I don't think this forum post may be the best place to take up HTS members time and space. Not that the knowledge isn't here but much of the ground I find already covered over at the DIYaudio forum. They have some good threads and heads over there. Crossover design is clearly a very technical and geeky area when you get into it and I'd like to jump in the pool and thrash around a bit.

On the critical listening end HTS is the place. When looking at musical instrument frequency ranges you find bass runs 40-200, vocals about 200-700, tenor sax 110-630. Have there been recommendations over time not to transition drivers within certain ranges.
 

·
Registered
Joined
·
168 Posts
To answer someone will need to know if the monitors be stacked directly on top of the 12in woofers or a few feet away? Stacked is like a 3-way design where the XO points is a trade-off between power handling/excursion/directivity whereas if they are far away you cannot XO as high (i.e. more like a sub/sat situation).
 

·
Registered
Joined
·
3,697 Posts
Stacking with the top speaker inverted to close the two tweeter spacing helps prevent or minimize lobing issues.

It sounds like you have already invested money and materials into this project. I'd run with what you have and call it good, not a whole lot of room for improvement without investing a lot of effort for what could possibly be little net gain, or worse. Knowing exactly what components we are discussing might help us guide the process.

If you are open to alternative suggestions, a two-way coincident or waveguided pair of mains would give the widest coverage and pattern control. Add dual subs and the majority of the shop floor will be covered evenly.
 

·
Registered
Joined
·
3,697 Posts
Fourth order 3-way XOs are "interesting", textbook is not going to give anywhere near the real world response one expects.

Just another reason why simple two-way mains actively XO'ed to subs that can be dialed into time domain with the mains electrically and physically can be simpler and easier to implement than a same baffle 3-way design. Trying to force obviously disparate drivers (woofer, midrange, tweeter) into line with like XO topologies can be an exercise in futility.
 

·
Registered
Joined
·
16 Posts
Discussion Starter #7
Interesting comment about the 2 way waveguide + dual subs. This stirs the pot a bit. As good as 3 way systems can sound the 20th century was a while ago.

Doing a listening comparison between two pair vs one pair from the earlier suggestion the single pair won the race.

At this point not too much has been invested. A couple knock down cabinets from parts express and 12” Dayton DC300-8 speakers. Parts on hand were the monitors (4” mids/tweeters from the 90's), a pair of SEAS 27TFFNC tweeters and SEAS crossovers. A trip to the local radio shack (I'm talking about a real radio shack, one where surplus parts are stacked to the ceiling, not the one using that name who changed their business model) for crossover parts. Less than a hundred bucks for a box of inductors and capacitors, some with very good audio grade.

As suggested above, a starting crossover at 500 Hz was tried with a 2nd order circuit. Results were random. Not bad but lacking. Evaluation listening had the Hammond B3 organ split between the 12” and 4”. Moving the XO to about 230 was much smoother with the B3 all in the 4”. Next up was removing the polycarbonate tweeters and 1st order crossover in the monitors. The SEAS parts lying around were put in and tested. This is now a listenable system. With XO's at 230 and 2300 HZ the balance is good. Cymbals have much better articulation with the strike of the sticks having more character and less hiss from the hi-hat. The original tweeters were very well matched to the mids but component quality was just not to the level of the parts now used.

In assembling the XO for the midrange it was initially set up without the low end blocked in a band pass arrangement. Sound quality was good but after adding the parts to block the lower frequencies from the mid there was a huge overall improvement. One simple detail and a night and day difference. Bass was punchier and more musical at the same time. Bass lines and solos were more easily heard and had a satisfying feel.

As good as the system now sounds for what its intended purpose is, I'm not sure if the best approach may be just to build a cleaned up 2nd order system. There are a lot of moving parts in a 4th order crossover so I think I understand your comment about 4th order XOs and the challenge of making the math equal the real world physics.
 

·
Registered
Joined
·
21 Posts
Fourth order 3-way XOs are "interesting", textbook is not going to give anywhere near the real world response one expects.

Just another reason why simple two-way mains actively XO'ed to subs..
The parts cost to build excellent passive crossover-speakers systems is greater than what it costs to build an equally well performing active system. But the already built ones are even cheaper.

My present system uses KRK 10-3 speakers (tri-amp active crossover) that I got on sale for 360 bucks each.

If I had to buy a new system now I would go for JBL active monitors with the 5" woofer and a pair of PartsExpress 18" Ultimax subs with DIY box. Then use an old Harmon Kardon "high current" amp to drive the subs and a receiver with preamp outs and set its internal crossover at 120Hz.

BTW: The JBL would perform better at louder levels if crossed at 160-200Hz but you might have to go to a 15 or even a 12" sub to get that high. I did audition the JBL with 8" woofer and didn't like it as much as the ones with the 5" woofer.
 

·
Registered
Joined
·
16 Posts
Discussion Starter #9
Good point about the expense of decent crossovers. It turns out they can easily cost as much as the drivers. For this project, fortunately, budget hasn't been a big issue. One of the primary objectives from the start has been to get a handle on the fundamentals of audiophile speaker design.

Building and testing multiple crossover iterations along with the use of different drivers has been indispensable in learning what does what and developing a deeper understanding of recommendations found here in the forum.

After Tesseract's suggestion on the idea of using active XOs and having read a large number of posts on the subject I have now walked that path a short distance. But only after spending extremely useful time experimenting with analog crossovers. REW along with a DATS have received much use since the earlier posts in this thread. Many pages of 1st hand notes have been generated as a result.

On the active front three Crown XLS1500s were added as they seemed to have the best value out there. Huge overkill and destruction potential for the tweeters but the cost/quality made sense to use the same amp for all drivers. A mini DSP has been added to manage crossover duties after giving some consideration to Behringer before making a decision. Ability to tune and adjust parameters met the R & D nature of the the project.
 

·
Registered
Joined
·
21 Posts
My first active crossover system was a four-way with LR 24dB/octave crossovers and whatever hifi amplifiers I could scrounge. The tweeter was a 1" dome, hign mids were dual 5", low mids were 10", and subs were 12" sealed. This was around 1990 and the crossovers were all analog.

One thing to watch is the large signal analysis. Be sure you keep the voice coils in their magnetic gaps. I played it safe and calculated the excursion for each driver based on the spl and low side crossover frequency without taking into account the -6dB crossover roll off for each driver. Those calculations showed me I needed dual 5" drivers. That system was the most neutral and effortless sounding stereo I have ever heard.

A second important aspect that you have alluded too is good frequency response beyond the crossover frequency. This is important for keeping the phase relationships of the drivers correct around the crossover points. The LR 24dB crossovers are less sensitive to this than some other types but it is still significant.
 

·
Registered
Joined
·
16 Posts
Discussion Starter #11
Here is where the project is so far:

SEAS tweeters 27TFFNC used in small 1990's Pinnacle AC 400 ported cabinets with original 4 1/2” coated paper drivers. Power handling is not particularly high but it gets loud enough. Woofers are 12” Dayton Audio DC300-8 in sealed boxes as mentioned.

Some testing was done with AMT 40-4 tweeters with a recommended crossover of 3500 Hz. These were good but the mid/tweet crossover point after trials sounded better in the 4000 Hz range. Coupled with the 12” woofers crossed at 250 the mids were being asked to do a lot over a wide range and did not sound so good. On the other hand, the SEAS 1” dome tweeters when crossed at 2500 on the high side were much better with the 4 1/2” and the impulse response was reasonably flat through the crossover point.

Next target is to try phase aligning the woofer and mid around 250 Hz and see if the sound improves.

The 12” driver, in a sealed box, does not have much output through the 1st octave but with the system being designed for audio vs home theater that should not be too critical as long as it gets with the program by about 40 Hz which the impulse response says it does.

Parametric EQ has been added to the tweeter to flatten the overall impulse response.

I'm still left guessing how you guys get the high end graphs so flat in http://www.hometheatershack.com/forums/two-channel-audio/69421-official-1-000-speaker-evaluation-home-audition-event-results.html
 

·
Registered
Joined
·
21 Posts
A little trick I learned from my experiments was to use relatively large sealed cabinets for all drivers and then measure the impedance. Then use the frequency where the impedance minimum above system resonance occurs as the low frequency crossover point. This also seemed to work with the mids in an open baffle configuration.

My experiments with the tweeter crossover point were similar to yours. Pushing the crossover frequency above 3k seemed to clean up the sound of the 1" aluminum dome tweeters I was using.

If you have a few extra bucks, look at Eton for your mids. I only have experience from what they were making 20 years ago but they were excellent performers. For example:
https://www.madisoundspeakerstore.com/approx-5-woofers/eton-5-200/a8-symphony-5-mid/bass/

Use two of those in an M-T-M arrangement. I wish I could have afforded a set of Eton hexcell but they were horrendously expensive back then.

BTW in case you are wondering, my system used Vifa AG35-06 tweeters, Vifa P13WH (5" poly cone) for mid high, Vifa 10" paper cone (M26?) for mid low, and some experimental 12" diameter 15mm xmax, f0=17Hz woofers in 4ft^3 sealed cabinets.
 

·
Registered
Joined
·
31 Posts
Just curious, what tools do you have for designing a passive crossover? It's not mentioned in your initial post. To do it anywhere close to "correct" you will need some sort of measurement system (REW?) and crossover design software. Active will be much easier because at least the crossover points will be close to where you put them and it will be easier to tweak. I've designed many many passive crossovers and even with a wall full of caps, resistors and coils it still takes a long time. And I still simulate with DSP first. Here's what I consider to be minimum equipment to design good passive crossovers. Calibrated microphone, system for acoustic and impedance measurements, crossover design software and an accurate LCR meter. You would be surprised how far off parts measure from their printed value. Of course you can do it all by ear and if it sounds OK, great. Have fun.
 
1 - 15 of 15 Posts
Top