Dual Subs with active external crossovers

[Will Meyer 2,374]

I recently posted on a thread from @Dadvw regarding a blown plate amp and I mentioned a project @jeffsand I were working on for an active crossover that, while it is based on a Dahlquist DQ-LP1 design, incorporated a modern power supply, modern multi layer board with a ground plane, newer components, and a revised circuitry. 

 

Some background, I've been using the Dahlquist DQ-LP1 for years. There is a lot of info on the internet if you want to grab a copy of the manual and schematic. Like the C-1, it has a loyal following including hobbists that have documented numerous improvments over the years. I've collected a lot of info and can post it if desired. I drive 2 subs, a left channel and a right channel. Common practice is to use a mono sub as "low frequency is non-directional".  I was watching a Paul McGowen video on multi subs and he addressed that the frequency of a sub goes from non-directional to directional - he strongly advocated dual subs.  I researched the internet and you can't get a straight story, one of the prevalent therories is that LF sound is non-directionsl under 400hz since that is equivilent to the spacing between the average human ears. Sounds hokey to me.

 

Here's what I know, I can run a low frequency signal and I can point directly to the sub. Maybe I'm just weird and have a thin head.

 

IMHO, for 2 channel music, if you feel you need a sub then you want ideally 2 subs. One by the Left speaker and the other by the Right, each being fed by their respective independent stereo channel.

 

Picture to follow of my layout (room size and layout is not perfect but you work with what you have). You'll see a set of Dahlquist DQ-SW1 subs, with JBL L100t perched on top of them.  I use the preamp signal (4000t), and feed it into the Dahlquist DQ-LP1. The low pass signal (independent L and R) feeds a TFM-24, which feeds the SW1s. The HP goes into a TFM-42 and feeds the JBLs. 

 

The DQ-LP1 is an old unit. @jeffs and I (mostly Jeff) redesigned the DQ-LP1 (MY-LP1) to improve it. This was not and is not intended as a commercial project, just an interesting personel project. I can post more on the project if interested.

 

 

 

 

[Will Meyer 2,374]

[Ar9Jim 6,514]

Very cool topic, Will.

The directionality of subwoofers is interesting. The first info I had read was in the AR9 owners manual from 79'. It said the point of non-directionality was 200Hz that allowed  bass drivers placement on the sides of the enclosers. The drivers were symmetrical right and left, and the 200Hz crossover point to the woofers worked.

 

In more recent years, when using single subwoofers located in non-symmetrical areas, the 200Hz crossover point has proven too high. Some harmonics produced seem to be above 200Hz and directionality still becomes an issue. 

 

In most recent times, I view the sub from the bottom up.. If the mains produce well and only roll off below say 50 Hz,  I'll start blending the sub with a low pass filter of 50Hz and tweak from there. 

 

In the current set-up the low pass filter on the sub is at 40Hz, using the sub to fill the bottom where the mains roll-off. The sub covers only a narrow bandwidth 

 

Some vintage music has almost no content down that low. Using a sub as a supplement to fill in the bottom of the mains, is usually way below directionality comes into play and a single sub may sufficient in many room areas.

 

If using a sub above 150Hz or so, placement and having 2 in symmetrical positions becomes of increasing benefit. 

 

It's a lot of fun finding that blend, when the subwoofer is invisible and supplements the mains just right.

 

 

 

 

 

 

Edited July 24, 2023 by Ar9Jim

[jeffs 3,242]

My history with the DQLP1.

Will and I have had several DQLP1 discussions for a few years now. I'm not unhappy with my setup with has a miniDSP doing the job of actively splitting out the LF and HF signals. I've threatened a DQLP1 purchase a few times, but never pulled the trigger.

This was an interesting project. With the exception of a new power supply and updated components, the original schematic was pretty much copied with the exception of splitting out the switch functions. The original unit has one switch per side which turns off the LF signal at the same time bypassing the HF filter cap, sending full signal to the mains.

From my side of the design, I wanted the finished product to fit in a 1U space. Found some nice transformers that allowed this to happen, while also being board mounted.

 

Mistakes:

The frequency adjust requires a triple ganged pot. Will found some quad ganged that would do the trick. Based on the non-linearity of the DQLP1 frequency dial, I assumed these were audio taper. As I tested the first unit, I realized this was a mistake.  Not all that bad except that 100K ohm linear quad ganged pots were not available. I decided to go with 50kohm. This meant that the caps in the LP filter sections all had to double. Note the doubling up of the caps (orange, green, and red pairs). I had allowed an extra spot for a common modification to the original DQLP1.

 

Build notes:

Build was easy. The concept of the enclosure was simple, the implementation was considerable effort. I used a blank 1u plate for front and rear. I used solid cherry sides to pull everything together. Top and bottom were cut out of an old cover. I have minimal tools, especially for metal work. Each wood side has 7 holes with brass inserts. Next Mouser order, I need a few RCA.

 

This worked well on the bench. I have yet to use it in my office setup. My initial thought was to A/B using 2 amps, one with the miniDSP and one with the LP1. I had all of the pieces to do this but sold an amp last week. OCCD has mostly left me, for which I am relieved.

The LP1 50k ganged pots have a center detent. This corresponds to 70Hz, which works well for me. Interestingly enough, When I plugged the miniDSP into the computer, I realized I had it set to 70Hz.

I know what all of the knobs do. Everything is mirrored around the center, the inside pots are L and R freq adjust, the end pots are L and R level adjust, middle offer some equalization at low frequency.

 

 

[4krow 6,630]

  Man,

  That is a strikingly clean looking build there. I get frustrated when running wires all over the place. A circuit board can clean it all up. 

[xavionics 2,091]

Wow, Purdy!

 

Very nice build, well done! 👍👍

[4krow 6,630]

Of course, you have a BOM available. The board surprised me in the mail today. So, more out of curiosity at this point. For a moment, I thought that I was looking a LP preamp board, but then saw the other outputs.

[MRHP 868]

I have two DQ-LP1's and use them for all the time. They are very handy. Mark Lucas did his magic and repurposed the center channel outputs so that I can plug in different crossovers that he made up for me. Super handy!

[MRHP 868]

[Will Meyer 2,374]

I've not seem that upgrade before but it looks like the center RCAs and accomplanied mini-circuit cards are wired into the passive hi-pass filter.

 

So changing the card will change the high-pass threashold?

For example: you can plug in a card and set the threashold at 100Hz, or switch it for another card and lower it to 80Hz?

 

Neat idea!

 

Edited August 2, 2023 by Will Meyer

[MRHP 868]

Yes sir that's exactly how it works. If for instance you plug in the 100hz card it sends 100hz and up to the high side and 100hz and below to the subs. Vacuum tube amps crossover at a different point than solid state amps. Something to do with their speaker output taps. A vacuum Tube amp is around 100k ohms and SS is around 10k to 33k ohms. This changes the crossover point. Example: one of my cards crosses over at 78hz with a Vacuum tube amp and the same card crosses over at 350hz with a solid state amp. So Mark made me several cards to cover vac vs ss.

[Will Meyer 2,374]

Got it.

 

The passive High Pass filter uses the amplifier's input impedance in conjuntion with capacitors to set the high pass threshold and that input impedance varies depending upon the amp you use.

 

For example, a TFM-35 has an input impedance of 47k ohms. 

A tube amp could be significantly different.

 

The DQ-LP1's design requires changing the caps by removing the cover and soldering different values in place.  Not condusive to rotating amps or speakers. I like the module idea - I'll use it in my next build.  Thx 

[4krow 6,630]

Getting to the final soldering of little stuff, but the transformer will be next and probably a part here and there that I overlooked. This was all very easy and I had a lot of the parts on hand, as well as Will to the rescue for many that I did not have on hand. Want to get the right transformer though. If I get to completion, I plan to put it up for sale and then figure a way to do some kind of karma with the proceeds. OR I would be willing to put this up as Karma if someone were interested. Dunno. It just popped into my mind this moment.

[jeffs 3,242]

9 hours ago, MRHP said:

Mark Lucas did his magic and repurposed the center channel outputs so that I can plug in different crossovers that he made up for me. Super handy!

This is a really good idea.

I wish I knew these existed to match the connector spacing.  I spaced RCA without measuring the actual unit, only reviewing pics.  Can't imagine my spacing lines up. It was the intent that the front of MYLP1 would face mount due to pots, but the RCA would be panel mpunt and wired. IN that case, best to know this spacing ..... or you could make them yourself.

 

Awesome tweak on the DQLP1. Simple and effective.

Edited August 2, 2023 by jeffs

[4krow 6,630]

Sorry, not great photos but this gives a bit of an idea of where I am at so far. Fingers crossed, I may have a suitable transformer downstairs. Crap, I just checked, and the tranny isn't the one needed for this project.

 

 Jeffs, Note the 3.3uf output capacitors instead of 2.2 It is what I had and figured that it would not be a problem, but am checking with you just to be sure.

 

 

Edited August 3, 2023 by 4krow
added info

[4krow 6,630]

Only one more switch to install. Next, chassis decision.