My Utopia Build

**JonMarsh** · 23 January 2014, 21:00 Thursday

Interesting concept! When do you expect to start cutting wood? By all means, please keep us up to date- we love pictures as well as technical details!

**benthe8track** · 24 January 2014, 00:02 Friday

Very cool!

**5th element** · 24 January 2014, 10:48 Friday

You say that you find the dynamics are missing, I wonder if you've ever tried a waveguide loaded tweeter as they tend to help give a dynamic presentation.

**harryharry** · 24 January 2014, 13:45 Friday

I do not think it is mainly in the tweeter frequency range I have been missing the dynamics. This comment was mainly about the Accuton C173-6-191E which I find to play every sound precise and correct but I am missing some dynamics. By all means it is a very good driver. The problem was that I did not immediately start to thumb my foot when listening to some of my favorite tunes. Then something is wrong. I would very much like to try the Accuton C173-6-090 which I believe is a different beast.

But I have been experimenting with waveguides on the tweeters with good results. This design is in fact using a waveguide/horn on the TPL150. I do not use the horn version but have instead created my own waveguide which is a bit more shallow than the standard Beyma waveguide.

I also have a Scan Speak 6600 and a Monacor WG300 I am planning on testing in a some future project. Like this: http://www.troelsgravesen.dk/7100-waveguide.htm

**harryharry** · 24 January 2014, 14:00 Friday

Originally posted by JonMarsh

Interesting concept! When do you expect to start cutting wood? By all means, please keep us up to date- we love pictures as well as technical details!

I just need to mount all the pieces on the plates. I am planning to use waterproof MDF which has a bit harder core and moves less when painted. The other option is to use baltic birch plywood or just plain MDF.

If I am lucky I will start with the cutting next weekend.

In the design the sides of the speaker are a bit curved. I am planning to glue a couple of plates together and then machine the profile before sanding. The other option is to bend several layers of thinner MDF. I do not know what will be the best method. The most common method is to bend the MDF. Any input?

**TEK** · 24 January 2014, 15:39 Friday

I think that you will have to do a lot more work if you are going to machine it and sand it (I assume that your CNC cannot do that?) You will also miss internal volume.
So I think that bending and gluing several layers will be easier and give a better result as well. Even if the CNC can do the work.

**A9X** · 23 April 2014, 00:44 Wednesday

I'm interested to see how this comes out. After I finish the subs, I'm building my surrounds in a similar fashion.
Below is a rough sketchup of what it's going to look like.

AE TD15, TD10M B&C DE250 + Ewave WG. All active.

**Face** · 23 April 2014, 06:10 Wednesday

Which TD15? Why not a SEOS 10 or SEOS 12 WG?

**A9X** · 23 April 2014, 17:13 Wednesday

X.
No SEOS because even after being in the initial order it seems that someone was not able to handle the difficult task of posting overseas.

**Face** · 23 April 2014, 18:27 Wednesday

I believe his reasoning behind that is due to being misquoted and taking a loss at the PO.

What size enclosures for the X? I have a pair here that will be going in sealed 3.25cf enclosures.

**A9X** · 23 April 2014, 18:44 Wednesday

Originally posted by Face

I believe his reasoning behind that is due to being misquoted and taking a loss at the PO.

He knew where the heck I lived when he took my money and sat on it for months. I was also in no hurry and repeatedly said he could take his time to post and that I would pay all the expenses. The USPS online calculator is accurate: I had a decent business for many years buying and selling audio and MI gear predominantly with the US and I used it all the time.

Originally posted by Face

What size enclosures for the X? I have a pair here that will be going in sealed 3.25cf enclosures.

60L net, sealed.

**benthe8track** · 23 April 2014, 19:18 Wednesday

Originally posted by A9X

He knew where the heck I lived when he took my money and sat on it for months. I was also in no hurry and repeatedly said he could take his time to post and that I would pay all the expenses. The USPS online calculator is accurate: I had a decent business for many years buying and selling audio and MI gear predominantly with the US and I used it all the time.

I've ran into that from American vendors in the past were reluctant to just ship to Canada. I think a lot of stems from confusion. I've had vendors tell me they don't know how to, "fill out the shipping documents". News flash: There aren't any, you put it in a box and take it to the post office...

**harryharry** · 04 August 2014, 17:21 Monday

Some woodworking

I have finally had some time to start this build. I had planned to get the parts CNC machined but in the end I cut everything by hand. I have simplified a bit. The side walls will not be curved. The speakers are going to be the main speakers in my movie theatre which is a bat cave so the finish does not matter too much. I will spray paint the speakers. I planned to go for a piano finish but I will not. It is too much work. Instead I will use a semi shiny black finish.

I have build some test enclosures for the drivers and this test was very successful so I have high expectations.

Some pictures:

**harryharry** · 04 August 2014, 17:42 Monday

The mid driver (PHL 2530) is going to be in a sealed box. How much does the internal shape of the box affect the sound? I will fill the box with damping material.

PHL recommend 3-4 liters closed box for this driver which is very small for an 8" driver. I know people who have used a 6 liter box with success with this driver. Datasheet: http://www.phlaudio.com/datasheets/20_pdf/2520_2530.pdf

To cut down on volume I have these options. I think option C is the best to avoid standing waves and keep the volume down. I plan to use 22 mm (0.9") MDF. The driver is going to be used from 350 HZ to 2500 Hz. Should I add some extra brasings or fill the extra volume by sand? Since this driver is not going to play bass it might not be necessary?

**5th element** · 04 August 2014, 17:50 Monday

Hmm the pictures aren't working. I'm guessing the image host doesn't allow direct hotlinking of files that they are hosting. I've looked at your post and there's nothing wrong with the tags, they just don't work. If you can please use the HT guide forum as a a host for the pictures, by attaching them. This also means that if your host goes down the pictures are still here for all to see in years to come.

**harryharry** · 04 August 2014, 18:13 Monday

Originally posted by 5th element

Hmm the pictures aren't working. I'm guessing the image host doesn't allow direct hotlinking of files that they are hosting. I've looked at your post and there's nothing wrong with the tags, they just don't work. If you can please use the HT guide forum as a a host for the pictures, by attaching them. This also means that if your host goes down the pictures are still here for all to see in years to come.

Fixed

**5th element** · 04 August 2014, 18:53 Monday

Those are coming on quite nicely. I like B the best. What you don't want, imo, is to have the back panel too close to the driver, or have it angled like in B, but you also want to make sure that there is plenty of breathing room around the sides of the driver too. A keeps the sides free, but has the back too close, C has a nice shape to it, with the taper going backwards, but the sides of the cone are far too close to the sides of the driver. B would be the best compromise.

**harryharry** · 05 August 2014, 16:01 Tuesday

I went with option C. It was easier to cut and to keep the volume down.

The volume is slightly larger than 4 liters + driver. There are no parallel walls.

I can fill the back camber with sand. However I do not know if this would improve the sound. It will add mass. However this construction is allready heavy. It will stiffen/damp the walls. I do not know if this will matter a lot since this mid driver is going to be used from 350 hz to 2500 hz. I use 22 mm (0.9") mdf covered with 5 mm bitumen sheets. What do you think?

**5th element** · 05 August 2014, 16:40 Tuesday

Hold on a second you have the option to use the full volume of that cabinet for the midrange? That is by far the best option and just heavily stuff the entire thing with wadding. This will require crossover modifications to account for the increase in volume, but using the full cabinet volume will be much better at absorbing the back wave.

You do have two options though. The first would be to use the small enclosure and to load the space with sand. This will definitely make the cabinet dead and give advantages in that regard. Or you could use the entire cabinet volume and have a better cabinet for absorbing the back wave and preventing any reflections from marring the performance.

**harryharry** · 05 August 2014, 17:19 Tuesday

Wouldn't it be much easier to reach xmax with the use of the full volume of the cabinet. This driver has a very low xmax. It is intended to be used with sealed encolsures from 3-4 liters.

**5th element** · 05 August 2014, 18:24 Tuesday

In a sealed cabinet, the SPL that a driver produces is directly related to the amount of air it moves, the cabinet size does not matter.

With a large cabinet the resistive force of the air pressure is relatively low so you reach xmax at a given SPL at low input powers. With small cabinets the opposite is true and you will require a lot more power to reach the same excursion value, the SPL though, for a given excursion, will be identical. All you're doing by increasing the cabinet size is reducing the system Q and resonance but you are also increasing the overall efficiency.

With a mid range in a three way you will be tailoring the highpass filter to reach a specific acoustic goal, be it 2nd order LW at 500Hz, or 4th order LW at 300Hz etc, the filter makes up for what the cabinet does not do, so it does not matter what box size you use.

**harryharry** · 01 September 2014, 15:36 Monday

Some progress. Cabinets painted and drivers mounted. This is the biggest speaker I've ever made.

Now I will start on the crossover. I have tested the drivers earlier in similar test cabinets so I will probably only need to tweek the crossover I used earlier. Time will show.

**JonMarsh** · 02 September 2014, 13:39 Tuesday

OK- Rock and Roll! You know what they say, it's not real until we have pics, and now we have a pic! :T

How similar are the cabinets you previously used? Same baffle width? Same height and driver spacing? If so, you're probably good to go, but final test and listen will tell all...

**5th element** · 02 September 2014, 16:16 Tuesday

Those should have some serious SPL capabilities!

**harryharry** · 04 September 2014, 11:15 Thursday

Max SPL is roughly 115 dB per speaker. That was however not the reason why I chose these drivers. The drivers deliver high dynamics with low distortion and high sensitivety. The overal sensitivety of the speaker is about 93 dB/2,83V/m.

**harryharry** · 07 September 2014, 17:46 Sunday

First attemt on crossover. First I had planned to go active on the bass section however now I've tried to simulate a passive cross over.

The frequency looks ok however the phase is not good. Impedance graph do have some peaks at 10 ohm but stays above 4 ohms.

I am no expert. Do you see any obvious mistakes I've made?

**harryharry** · 10 September 2014, 08:43 Wednesday

Prototype crossover. I did not have all the values in stock so I had to improvise a bit...

This speaker have some scary dynamics.

Now it is time for some fine tuning by ear.

Is it ok to use electrolyts in the bass section or should I only use polyprops?

I also need some large low resistance inductors (8.2mH and 2.7mH). Can I use Iron cored inductors or should I use transformer type or something like Jantzen C-coil (Toroidal coils)?

**Face** · 10 September 2014, 10:47 Wednesday

Iron core or C-coil will be fine for the woofer section, I'd avoid solid core on the mid or hf sections though.

IMO, I try to stay away from electrolytic caps unless it's a ridiculously large value or a budget build. Why use something with a finite life in an all out speaker?

**JonMarsh** · 10 September 2014, 11:41 Wednesday

Originally posted by Face

Iron core or C-coil will be fine for the woofer section, I'd avoid solid core on the mid or hf sections though.

IMO, I try to stay away from electrolytic caps unless it's a ridiculously large value or a budget build. Why use something with a finite life in an all out speaker?

+1 on this advice; the 500W Erse do the job pretty well, and don't have the peculiar difference in inductance measurement for different meters and conditions that the Jantzen C coils have had, in my experience. The speakers with them seemed to work fine, but that kind of nonlinear behavior bugs me...

And if this is an all out system build, consider Clarity MR or Jantzen Superior Z for the tweeter networks.

**JonMarsh** · 10 September 2014, 11:47 Wednesday

Originally posted by harryharry

First attemt on crossover. First I had planned to go active on the bass section however now I've tried to simulate a passive cross over.

The frequency looks ok however the phase is not good. Impedance graph do have some peaks at 10 ohm but stays above 4 ohms.

I am no expert. Do you see any obvious mistakes I've made?
[ATTACH=CONFIG]23436[/ATTACH]
[ATTACH=CONFIG]23437[/ATTACH]
[ATTACH=CONFIG]23438[/ATTACH]

Phase doesn't necessarily look bad- how did you determine acoustic offsets (Z plane- front to back) This is critical for crossover behavior. One way to do that is measuring all the drivers simultaneously, then combining the individual measurements in the crossover program (LspCAD in your case) and adjust offsets based on reasonable starting physical values until the simulated response looks like the measured response- keeping in mind you must position the drivers correctly in LspCAD in the Y plane, also.

Then, as a check on your crossover design, take a look at the out of phase nulls- what was your real transfer function target? Midrange low pass looks like 3rd order, but is it a B3 or quasi LR3? If B3, the null won't be present out of phase, and the levels at crossover should be about -3 dB. I don't recommend B3 except for special cases like when you want phase quadrature on a MTM driver setup.

**harryharry** · 10 September 2014, 15:31 Wednesday

Yes, I measured all drivers, combining the individual measurements in the crossover program and adjust the offsets (z-axes). I got an almost 100% identical result.

The inroom far field measurements is quite good. However I had to adjust the bass network to get a shallower curv below 100 Hz. With the first prototype crossover I ended up with to much room gain below 100 Hz.

I plan to use Jantzen Superior Z in the tweeter network. I try to convince me that I should go all in with the CalrityCap MR however the big values I use in the tweeter network is very expensive (8.2uF and 6.8uF)

**5th element** · 10 September 2014, 17:59 Wednesday

Yeah, the fact that those C-core's don't measure correctly on certain devices worries me a little, it says that they have linearity issues, either at certain drive levels, or frequencies, neither being things I'd want near any loudspeaker of mine.

Air cores are always preferential to cored inductors. All you have to watch out for is the DC resistance within the series connected inductors. Usually all you need to do is make sure that the DC resistance is roughly between 5-10% of the woofers DC resistance and you're good to go. Any inductor in a shunt position also needs to have its DC resistance monitored, but it isn't usually a problem.

What exactly do L9, C8 and R9 do? The same goes for R14 C11 and L11. Although not large in value I can only assume that they are for impedance compensation. Now unless you're using a valve amp, or a finicky class D amp, this is probably not necessary.

As to electrolytics, I would have no problem in using them to start out with. Once you've got the crossover tuned properly, going to poly caps would be a good idea.

**harryharry** · 11 September 2014, 16:27 Thursday

Second attemt on the crossover

Second attemt on the crossover.

This time I have tried to model a LR2 at 400 Hz and LR4 at 2000Hz. I have tried to implement a slightly downward SPL slope in line with the Brüel&Kjær recommendation. :http://www.bksv.com/doc/17-197.pdf page 4

Inverse polarity mid driver:

Phase

Cone excursion simulation at 30V of PHL2530 mid driver. First time I've tried this. I am a bit worried using first order electrical high pass for PHL2530. Xdam for the PHL2530 is 4mm so I seem to be ok if I have done everyting correctly.

**harryharry** · 11 September 2014, 16:40 Thursday

The L9, C8, R9 and R7 C7 and L8 are impedance correction circuits. I am going to use a Hypex Ncore to drive the speaker. Could I skip these?

Impedance with circuits in place

Impedance with no impedance correction circuits in place

**harryharry** · 12 September 2014, 09:58 Friday

Third version. Small adjustment on low pass on the bass driver. I do not know if this is better than the second version (post #34). However when I reverse the pole on the middriver the nulls are more symetric. What do you think?

Transfer Function:

**JonMarsh** · 12 September 2014, 11:30 Friday

I think that's a good change- the voicing is reminiscent of the Danish speaker line, Peak Consult, but you'll certainly find it listenable and musical, something that isn't always the case with "flat", especially with typical digital sources.

**harryharry** · 13 September 2014, 18:54 Saturday

Semi final cross over version

Very far field (6 ft / 2 m)with long window measurement. The suckouts at 100Hz and 250Hz are caused by room interaction.

**harryharry** · 13 September 2014, 19:07 Saturday

Impedance correction

Can/Should I skip the impedance correction ciruit? I will use a Hypex NCore NC400 to drive the speaker.

With some impedance flatening cicuit C9, R8, L9 in place.

No imp correction:

**Paul W** · 14 September 2014, 13:23 Sunday

Overall shape of the FR curve should be very listenable...but I'd get rid of that hump ~800Hz.

**5th element** · 14 September 2014, 15:32 Sunday

Impedance compensation comes in two varieties.

The first is the type that goes after the driver, but before the crossover and, if correctly chosen in value, has a direct impact on the frequency response and is absolutely necessary for the crossover to work correctly.

The second is the type that is usually placed after the driver and the crossover and is used entirely for impedance linearisation.

The first type does not give you a flat impedance loudspeaker, more it adjusts the impedance and allows the crossover to do its job properly. This may or may not result in a flatter end loudspeaker impedance, but it is not it's primary goal. The second type does absolutely nothing to affect the loudspeakers performance, but in linearising the impedance, it will improve the interaction between the loudspeaker and the amplifier. Some amplifiers will appreciate this, where as most will not. The ncore is one that has a very low output impedance, has a load invariant frequency response and is capable of driving difficult loads without any issue. Impedance compensation is not required with this amplifier.

**harryharry** · 14 September 2014, 18:44 Sunday

I will follow the advise of 5th element and remove the impedance correction.

I have revised the cross over a bit to remove the 800Hz hump. I have also improved the bass low pass a bit.

Mid driver inversed

Phase:

This time I voiced the SPL flat. I might voice it slightly downward by ear later.

My Utopia Build

My Utopia Build

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment