First time OB builder

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Tordenguden see pics for wiring. Instead of just replacing the loopback cable for another one, the Help says it is more accurate to use a Y adaptor.

I hope you are not planning to use an Y-adapter, to mix the loop-back signal and the output for the DCN28 - because that is what it looks like on the picture ?!?

The Quad Capture "only" has 2 channels running full duplex. And on channel, is used for reference in the software, so that more accurate result can be obtained. You may use an Y-adapter at Output 1, to simultaneously send signals to both left and right input on the DCN28 - but thats a different story.

The way I do it, could be wrong, but doubt it, since I'm getting pretty good results :D
Microphone goes to Input 1.
Loop-back cable goes between Input 2 and Output 2.
Output 1 is for DCN28.

How did you adjust the buttons - both front and back?

Fair enough that you want to experiment with room-cancelation. But I would advice to measure between tweeter and midrange, at 1 meter distance, straight on - to begin with. When you master this - it'll be way easier to play with other possibilities thereafter ;)
 
Last edited:
Yes, yes, yes Silent. The slight downward tilt into the highs at the listening position is good! You do not want flat at your position. It will be overly bright.

Still with the wider range on the mids?

Yes still crossing over at 400Hz from mids to mids woofers. I actually thought the fuller (deeper) vocal sound of the bigger area drivers would need to be present all the way through the vocal range, but to my surprise they still seem to give a fullish sound crossed over at only 400Hz.

I could bump that up a little bit to add a bit more bass to the vocals, but I am quite happy with where they are for now. The 18WUs did exactly what I was hoping they would do. The temporary 10" Vifa I had in that 3rd speaker spot before is like chalk and cheese.

Still keen to develop the woofer cabinet a bit more, perhaps after Christmas...
 
I hope you are not planning to use an Y-adapter, to mix the loop-back signal and the output for the DCN28 - because that is what it looks like on the picture ?!?

Yes that is exactly what I was planning to do as that is the preferred option according to the Help.

The Quad Capture "only" has 2 channels running full duplex. And on channel, is used for reference in the software, so that more accurate result can be obtained. You may use an Y-adapter at Output 1, to simultaneously send signals to both left and right input on the DCN28 - but thats a different story.

The way I do it, could be wrong, but doubt it, since I'm getting pretty good results :D
Microphone goes to Input 1.
Loop-back cable goes between Input 2 and Output 2.
Output 1 is for DCN28.

That is how I currently have it set up but instead of buying a TRS to TRS cable I bought a TRS to XLR cable not knowing that it was not the same thing electrically.

How did you adjust the buttons - both front and back?

Back buttons:
Ground Lift = Normal
Phantom = 48v
Hi-Z (input 1) = Off

Front buttons: see pic


Fair enough that you want to experiment with room-cancelation. But I would advice to measure between tweeter and midrange, at 1 meter distance, straight on - to begin with. When you master this - it'll be way easier to play with other possibilities thereafter ;)

Had a bit more of a play with it last night the only way to get any meaningful measurements for equalisation is near field. Perhaps down the track I will find another way but for now that seems to be te best way.
 

Attachments

  • Quad Capture.jpg
    Quad Capture.jpg
    129.8 KB · Views: 242
Finally got around to doing some comparative testing of 2 x 18WUs (8 ohm) vs 4 x 18WUs (4ohm).

I know some view using 4 drivers for a mid woofer as a bit of a waste, but I think the test results show a more preferable response curve using the 4 drivers.

I plan to build a U frame similar to how they are laid out on the test baffle, and I may wind up picking up a bit more bottom end SPL, but that will just give me a bit more headroom to possible crossover slighly lower (~300Hz)

Already have an idea what the new speaker might look like...

The four 18WUs in a U frame at the top, with some kind of legs to hold it up, and provide slight angle changes.

Dangling bellow that the RAAL and the two 12MUs

Sitting on the bottom either the Pro 5100 in a H baffle, or two Dayton 15" reference series drivers in a slot loaded dipole.
 

Attachments

  • 4 vs 2 Mid Woofers.jpg
    4 vs 2 Mid Woofers.jpg
    317.5 KB · Views: 227
Last edited:
Hi Silent Screamer,

some commenst from my side:

1. There is a long long way to go for you... ;). I am now "digging in the dirt -of measurements and their interpretation" with my DCN28 & the XOver software since almost 2 years...
2. I am also using the 2nd output of the soundcard for the loop-back signal to the 2nd input. It´s precise enough, there is no need to use an Y-cable (right now, maybe 1.000 hours of measurements later). Other topics are much more important before considering about differences out of the 2 outputs.
3. Please consider the directivity index of your speakers first. It could save plenty of your time. Target is a constant directivity across all the speakers. The target curve at the listening position mostly depends on the directivity index.
4. Measuring positions: 1. fixed measuring on axis near field close to the speakers (~40cm) to measure the acustical phase. It is hard to get the correct phase out of the XOver software. Minimum phase is calculated and not reliable enough. You can get the correct measured phase when you move the red cursor to the peak of the impulse response and switch of "add propagation delay" in the FFT menu. Only with the correct phase definition a correct crossover prediction can be done in the XOver. Compare the phase measurement of XOver and Arta attached. Almost the same (above 600Hz) 2. Vertical middle between the two speakers to be crossed over. (check if the phase remains the same) to define the delay of the speakers. Note: filters and EQ´s will shift the phase sometimes more than 360°. The SPL then looks ok but the phase is not correct and the impuse responde / step respnse will be terrible - no punch/kick. Do not use 36dB, 24dB is more than enough. The higher the filter, the worse the impulse behaviour. 3. Wagged measuring at the listening position with > 15 averaged measurements finally gives you a good impression of the reality. A stationaire measurement at the listening position is useless. You can not measure the phase at the listening position because of the room, so don´t worry about the phase there.
5. Start to experiment with filters (24dB L/R). Then adjust the level as Robert already explained. Then the delay to get a good impulse response. EQ´s only at the end of the road after you found out the behaviour of your system.
6. Windowing with multiwin is not really usefull. A window of 2ms means a resolution of 500Hz (nothing below can be shown & 500Hz between each data point)
7. Toooooooooo much stuff to explain in one post and short time... sorry
 

Attachments

  • phase.jpg
    phase.jpg
    355.8 KB · Views: 186
  • phase arta.jpg
    phase arta.jpg
    143.2 KB · Views: 178
Attached an example: Prediction of the crossover function. Same speakers, same measurement. You can move the green cursor to adjust the takeover frequencies. Only if you have the correct phase information you can see what happens with the spl.

Recommendation: Use a second software to make control meaurements.

Try to measure the correct phase of your speakers. Try slightly different positions (+/- 20cm) to be sure to know te effect. I´ll go to sleep now.
 

Attachments

  • spl prediction with correct phase 1.jpg
    spl prediction with correct phase 1.jpg
    436.5 KB · Views: 168
  • spl prediction with correct phase 2.jpg
    spl prediction with correct phase 2.jpg
    440.4 KB · Views: 45
1. Probably even further than you think ;O)

2. Still no joy on the cable from the U.S. Might just forget about the one I have coming and order from someone else. Since the Y adaptor wasn't going to be long enough, I might get a male to male TRS and a 2F-1M TRS adaptor instead. That way I can test both ways.

3. Sorry not familiar with the term "directivity index" what is that in layman's terms?

4.1 Robert initally suggested I should start around .75m, from trying to EQ from listening position, I come to realise that I can't get usable measurements from so far back and have started using much closer measurement around .15m can probably measure a bit further back, just playing with distance.

I have yet to understand what phase is and what it does, I have been aware of it for many years, but still don't know where it fits into the scheme of things. Would like to buy a book that explains the basics, but I find most stuff online delves into things too deeply before explaining the concept of it.

It sounds like you are using the software to auto pick a crossover point. So far I have been following the advice of Greg and a few others along with some trial and error to manually set the crossover points.

4.2 Not yet up to the point of using the delay functionality in the crossover, as I am still get the basics sorted out. Robert mentioned measuring from between Tweeter and Mids to catch both.

Interesting what you say about EQ and filters and the effect on phase, I will be mindful of that once I understand what the phase should look like and learn which is more important level SPL or phase.

I'm using 24dB on one side and 36dB on the other on all of the crossover points, it gives me the flattest line to start off with before correction.

Impulse and stepped response are to do with phase?

4.3 Ok phase cant be measured at listening position effectively.

5. Ok delay before EQ got it.

6. So are you saying use DuoWin instead of MultiWin? I still dont really know what difference there is between them.

7. No worries the path to learning is sometimes slow and there is a lot I have yet to learn. One day all of these things will make perfect sense but I need to work up to that point. Thanks for all your advice and pointers :)
 
Having just returned from a long trip to my mothers place (~2000 klms round trip) I brought back with me a sub woofer I made back in the late 90's.

I did some measurements on it with and with the the top ports in place. With the ports in place it acted more like the double band pass (Isobaric push - pull) it is, but with the top plate removed it measured very similar to the Tc Sounds Pro 5100 18".

I just used the same woofer crossover point and amplifier so I could do a back to back comparision. There was still no real bass to speak of which has me wondering if the pre amp in the DCN 28 is putting out enough voltage to really drive the 1000w per channel amp properly.

Since neither speaker is really moving the cone a great deal I have to start looking at something to explain the lack of cone movement.
 

Attachments

  • IMAG1240-1.jpg
    IMAG1240-1.jpg
    151 KB · Views: 70
  • IMAG1237-1.jpg
    IMAG1237-1.jpg
    96.3 KB · Views: 61
  • IMAG1231-1.jpg
    IMAG1231-1.jpg
    122 KB · Views: 72
  • IMAG1219-1.jpg
    IMAG1219-1.jpg
    119 KB · Views: 82
  • IMAG1215-1.jpg
    IMAG1215-1.jpg
    124.9 KB · Views: 67
  • IMAG1214-1.jpg
    IMAG1214-1.jpg
    147.4 KB · Views: 67
  • 2 x 12.jpg
    2 x 12.jpg
    311.6 KB · Views: 48
I tried both ported and sealed with a single 12" driver and there is no apparent difference between the old sub and the 18". Currently I am running one of both to do some listening tests.

I moved the port to the outside to make it easier to change between ported and sealed (port hole blocked off)
 

Attachments

  • IMAG1247-1.jpg
    IMAG1247-1.jpg
    134.6 KB · Views: 94
1. Probably even further than you think ;O)

Not really, believe me. You reached the border of your town on the trip around the world. It was the same for me, I always thought "I got it now..." ;) Let´s go together (if you like to - I think I can also learn much more)

2. Still no joy on the cable from the U.S. Might just forget about the one I have coming and order from someone else. Since the Y adaptor wasn't going to be long enough, I might get a male to male TRS and a 2F-1M TRS adaptor instead. That way I can test both ways.

As I said. You will not recogize a difference - yet. But be careful and always think twice before you plug and play, the inputs of the DCN are sensitive. Easiest way is to use the 2nd channel as reference, as I wrote.

3. Sorry not familiar with the term "directivity index" what is that in layman's terms?
Directivity means how the speaker distributes the sound. A horn for example has a high directivity. 15° out of it´s axis the sound level is almost zero. A subwoofer does not bundle the sound and distibutes it almost 360°. At the listening position, on axis, you might have a flat SPL curve. But 15° or more out of axis you might have a very inhomogenius sound field. The first sound that arrives at your ear (on axis) is only responsible for locating the source. The quality of the sound itself is mostly created by the reflexions - more or less bundled by the directivity of the individual speakers. It´s never bad, but you have to know it to take care of the correct target curve at the listening position. The target curve correlates with the directivity level to compensate it. Maybe the attached drawings might help to understand it.[/QUOTE]

4.1 Robert initally suggested I should start around .75m, from trying to EQ from listening position, I come to realise that I can't get usable measurements from so far back and have started using much closer measurement around .15m can probably measure a bit further back, just playing with distance.

0.15m is way to short for the mid & high range. You will not see the boundary effects of the speakers. Rule is ~1.5*width of the baffle to also get the baffle step into the measurement. Play around with the distance. You will recognize that EQ of small dips and peaks makes no sense because 5cm farther it´s completeley different. EQ in the near field makes sense to define rough trends or eliminate resonances.

I have yet to understand what phase is and what it does, I have been aware of it for many years, but still don't know where it fits into the scheme of things. Would like to buy a book that explains the basics, but I find most stuff online delves into things too deeply before explaining the concept of it.

Try out the following: Make a blank XOver e. g. with two output channels. No measurements. Define filters with the and play around with the quality of filters, reversing, delay, and so on. You will see that every action shifts the phase more or less. See what 24db & 36db leads to - I tried it out. A dip in the SPL and a wave in the phase.[/QUOTE]

It sounds like you are using the software to auto pick a crossover point. So far I have been following the advice of Greg and a few others along with some trial and error to manually set the crossover points.

No!!! Not at all, I am doing it comletely manually. The software helps to predict the result.

4.2 Not yet up to the point of using the delay functionality in the crossover, as I am still get the basics sorted out. Robert mentioned measuring from between Tweeter and Mids to catch both.
Depends on the distance and the directivity of the speakers. If the middle between mid and high is on ear axis it works well. Otherwise not. At a distance of 0.15 from the baffle it will def. not work.

Interesting what you say about EQ and filters and the effect on phase, I will be mindful of that once I understand what the phase should look like and learn which is more important level SPL or phase.
SPL is more important but without a correct phase it will never sound perfect.

Impulse and stepped response are to do with phase?
Everything correlates. Impulse and Step gives information about the time behaviour, shows the delay that you need to set to have all the drivers "on time".

6. So are you saying use DuoWin instead of MultiWin? I still dont really know what difference there is between them.
None of them.

7. No worries the path to learning is sometimes slow and there is a lot I have yet to learn. One day all of these things will make perfect sense but I need to work up to that point. Thanks for all your advice and pointers :)

Why haven´t you choosen a vertical line up of all chassis B-M-H-M-B which leads to a cylindric wave front? Might be better with less elimination and better localisation?

Best regards and a good new year!
 

Attachments

  • bmaß ungleichmäßig korr zielkurve.jpg
    bmaß ungleichmäßig korr zielkurve.jpg
    17.1 KB · Views: 47
  • bmaß ungleichmäßig lin zielkurve.jpg
    bmaß ungleichmäßig lin zielkurve.jpg
    16.3 KB · Views: 56
  • bmaß gleichmäßig lin zielkurve.jpg
    bmaß gleichmäßig lin zielkurve.jpg
    13.8 KB · Views: 84
  • 24db & 36db.jpg
    24db & 36db.jpg
    400.6 KB · Views: 51
I haven’t fallen off my perch I have just stop to reflect a bit…
To move forward I truly need to understand in much better detail what effect BSC has on my bass.
For me that means writing a program… yeah I know that sounds strange, but to get right down to the nitty gritty of it, I need to get into the code and formulas.
Only then will I be able to fully appreciate what the baffle and the room is doing to the speakers.
I can keep throwing speakers at it until the cows come home, but common sense tells me I need to stop and understand the fundamental better before trying any more test baffles.

I have started up a new thread here http://www.diyaudio.com/forums/software-tools/227205-bsc-calculator.html to develop an application that will help illuminate the situation.
I have deliberately not used a spreadsheet or other software to calculate this so when I finally get to the point of entering the parameters into my application it will have one of those holy cow moments.
Knowing myself I will then automatically want to try different baffle shapes to get things looking better. So I am going to beat myself to the punch and write the app before I enter any real data.

The second reason I am writing this app is I need to practice my coding skills, and if I don’t write something of interest it will wind up in the draw with a number of other unfinished apps I have started on.
 
Been a while since I last posted in this thread but I haven't finished tinkering just yet. With the weather starting to warm back up I felt inspired to go out into the shed today and trim up the test baffle I have been using for some time now.

I definitely like the more open and less shrouded sound of the other test baffle, so I have trimmed up around the tweeter and the mids, and after putting everything back together and testing I had everything wired up properly, I accidently dropped a lead from the output of the active xover to the amp.

Talk about bad luck the center tip of the lead touched one of the output terminals of the only amp I had left on blowing one of the channels. Lucky the xover was off as was the amp I was hooking the lead up to.

Not sure if I will bother fixing since I plan to replace the amp down the track. So here are a couple of pics of the latest rendition of the baffle.
 

Attachments

  • IMG_0198.jpg
    IMG_0198.jpg
    484.1 KB · Views: 232
  • IMG_0200.jpg
    IMG_0200.jpg
    451.9 KB · Views: 246
Just revisiting and completely re-reading this thread from start to finish. What has changed? Absolutely nothing... I am a slack ***. I never did get around to building those H or U boxes I promised myself.

5 years have passed and if anything I think I know less now than I did back then (if that is humanly possible). I grabbed the external sound card out of the box for the first time in years a few days back, and I have no clue where I went with which cables.

I am hoping a diagram a few pages (post #370) back might lead to some kind of epiphany. Kind of feel a bit like Indiana Jones on a quest to discover the lost wiring diagram that once worked. I left myself a clue that Robert from GroundSound was helping me understand it, so that means somewhere buried deep in numerous emails might be some clues.

I have forgotten what little I did know about the Xover software, so hopefully our resident gurus have worked it out by now and are still lurking in the forum. If I am really lucky Robert got around to finishing that Help file he was working on.

Ok, what I have actually managed to achieve in the last 5 years is build myself a fairly sizable CNC machine so should I ever get off my bum and get back into building speakers the job will be much simpler.
 

Attachments

  • IMG20170726150735.jpg
    IMG20170726150735.jpg
    998.7 KB · Views: 131
Last edited:
Tell me about it. I have an electric tilting trike I started about 9 years ago. The CNC has progressively been getting better and better for about the last 3 years. The plan was to build the CNC so I could finally use up all the drivers I have kicking around. So far all I have done is upgrade the CNC.

But I can feel a surge of speaker building coming on, it's been building up for the last 6 months.
 

Attachments

  • IMG20170908171036.jpg
    IMG20170908171036.jpg
    972.6 KB · Views: 107
  • IMG20170908171019.jpg
    IMG20170908171019.jpg
    1,003.8 KB · Views: 113
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.