March 18, 2012

Spray painting over butt joins

Spraying over butt joins, especially with MDF, has always been a problem. The typical result is a seam or crack at the join which is quite unsightly. An internet search reveals some quite involved attempts to get around the problem. The most reliable solution seems to be to avoid them completely and veneer over the join, but it is unfortunate to go to the effort and expense of adding veneer only to paint it. Further, when one wants a roundover on the edges, veneer is not a real solution.

Understanding the problem

The seam is caused by the edge of MDF soaking up more than the surface, resulting in expanding after painting.

My first trial involved an MDF butt join where I wiped the surface of the join with polyurethane glue, then added two part bog filler over the top. I then applied another layer of each before brushing on undercoat then spraying with a can. When a seam came up I applied more layers or glue and bog, but it kept appearing. The seam was quite a bit more subtle than if no treatment were used, but still I'm aiming for a higher standard.

Trials in progress:

1. MDF with PVA glue - this is the base sample I'm doing so that I can compare the results step by step.

2. MDF with Polyurethane glue - does a stronger glue reduce the problem?
3. MDF with a groove cut with a router bit to hide the seam.

4. MDF with fibreglass reinforced filler over the seam, auto bog over.
5. MDF with a groove cut, then filled with fibreglass reinforced filler.

All of these were then repeated with ply. This example is with ply #5.


March 17, 2012

Spray painting system

For some time now I've been putting together a spray painting system and it's now complete.

2.5 hp compressor:

This compressor has a maximum air delivery of 280 L/min. I've also installed a moisture filter, seen below:

Moisture is an issue with home compressors. Commercial systems have expensive air drying systems. Obviously any moisture in the air is a problem. I may put together a more elaborate system with long runs of copper pipe to allow the air to cool so that water will condense in the pipes before going out the gun. 

I've made a device to hold the gun while mixing in the paint. You can also see my filtering mask. It's a 3M and not the same as those used for dust - special filters are needed. This one also fits very well and is comfortable.

I've discovered spraying outdoors can be frustrating, wind can be a problem. I'm considering constructing a spray booth to allow spraying in the workshop.

Here are some samples where I'm testing different methods of dealing with butt joins. The problem is to avoid a seam when painting over. 

March 16, 2012

Let's talk about woofer speed

Woofer speed is an entirely hairy topic, but I'm going to discuss here one aspect briefly - transient response. It is commonly thought that driver moving mass and motor strength are critical aspect of the time it takes for a woofer to respond to a transient signal. Some take a narrow view and look only at moving mass. Others will argue that this can be overcome by a powerful motor. It seems in reality neither are directly related to transient response.

In the past, Dan Wiggins of Adire Audio had an article on this topic on his website, which is no longer online, but can be found on the Way Back Machine:

Woofer speed by Dan Wiggins >

This post is essentially a summary of that paper - the short version.

We start with the basic equation that we all learn in Physics class:

F = ma

which means

Force = mass x acceleration

The force is provided by the driver motor and we know it as BL factor multiplied by the current i. The mass in question is the moving mass of the driver (mms). So we can now rewrite the equation in speaker terms:

BL x i = mms x a

BL is time invariant - that is, it doesn't change with time
Current (i) on the other hand, does change with time as it carries the audio signal
mms is also a constant and does not vary with time

So we see that both BL and mms are time invariant and don't change with time, hence they are not real factors in the equation. The two critical factors are in fact current and acceleration which we are seeking to investigate.

So we see that the ability of a driver to respond quickly to a transient signal is proportional to the current supplied. If we could make the current increase infinitely fast, then the driver would respond in kind. However, we can't increase the current infinitely fast. The reason is that the voice coil is in fact an inductor, and it turns out that the inductor is the bottleneck in this equation. Inductors don't like the current flowing through them to change. The bigger the inductor, the longer it will hold onto the current. So in a nutshell, we now have an answer - the critical determinant of how quickly a woofer will respond to a transient signal is the voice coil inductance.

So if you want to know something about the transient response of a woofer or subwoofer, consider the voice coil inductance.

In a real room, the lack of bass damping will often be a more sigificant factor, but once this is dealt with via treatment, this aspect is something to consider. In a recent comparison of an Acoustic Elegance TD18 woofer and my Eminance Magnum 18" woofers I noticed an obvious difference in the perceived transient response. I used some EQ to match the frequency response and there was a clear improvement with the TD18H which has a much lower inductance.

So what happens if the mms is made very very heavy? The fs is pushed down and the driver becomes very inefficient. Taken too far it will perform quite poorly for reasons other than transient response.

March 14, 2012

Will this driver work down-firing?

Some drivers are fine in a downfiring orientation, others aren't suitable. It depends on the stiffness of the suspension. A downfiring orientation can cause the cone to sag, resulting in a loss of excursion and output.

Sag = Cms x mms x 9.8

mms is in kg

The sag should be less than 5% of driver xmax

March 12, 2012

Bathurst 2011 - the audio event of the year

There is an audio event that happens every year in Bathurst that I call the audio event of the year. I call it an audio event but in many ways it's more about a group of mates getting together who just happen to also be audio enthusiasts who love music. This year we even had some live music (shown above). 

The host:

If you ever get the chance to meet him, "Jerry Tones" as he is sometimes called, is quite a character. The kind of person you don't meet every day. 

The event takes place near the Bathurst race event, although most don't actually go to see the races on this extended weekend. 

Le Chateau Jones:

Audio has a way of attracting people who are also into photography. One photographic enthusiast shared some tips with me in a tutorial. This gave me a little inspiration to take a few more shots.

Jerry Tones is in the process of renovating, a very ambitious project. Here is one room that is complete:

Weston Room

Earle Weston of Weston acoustics is a regular and set up a room with some Tannoy Cheviot speakers and some of his amps, including one given away very generously as a door prize. 

Sources include a Marantz player and a Zune media player.

These are beautifully crafted amps and my shots don't come close to doing justice to them.

Earle demonstrated one particular amp with the ported boxes. It was quite a lesson. The amp in question had an ultra low damping factor and as a result, the woofer had massive excursion and you could see it flop around as if playing heavy 5 Hz bass content. I was sitting there with John Reiki (HifiZine editor and owner) and we were pondering the issue when I decided to try an experiment. It occured to me that the lack of amplifier damping meant that the driver had (almost) nothing to stop it moving, especially below tuning. So we blocked the port with a thick Harry Potter book (useful for port blocking and other magic tricks). Immediately the problem was gone, the box now providing the needed damping. The bass was a little more than we might have expected, probably due to the amp lending a slightly bloated quality to the bass. So what we had was a mis-match with the ported box. The other valve amps did not have this issue. 

Andrew Ward of Aslan acoustics also had a speaker there although we didn't end up hearing this one.

The Franks

Originally the Franks lived in this box:

This explains the name, Frank being short for Frankenstein. The box now houses a sealed Maelstrom 18" sub driver.

The Franks are high sensitivity speakers with a 4 way digital active crossover (DEQX) with dual subwoofers. 

1" Morel Supreme dome tweeter
6.5"  + 10" PHL mids
18" PHL woofer

PHL make high end pro drivers that are used in Genelec studio monitors. They are expensive but very good and among the best available.

You may notice the room has a lot of fibreglass insulation. Terry has commented that he found the imaging and sound stage to improve dramatically as a result. Early side wall reflections are damped and absorption has been placed on all walls. There is also a bass trap at the rear of the room. Terry is a big believer in acoustic treatment.

2D QRD diffuser:

There are a number of these around the room. Terry has also made 1D versions which aren't in use. 

The Franks have unusual construction. The enclosure consists of an MDF skeletal frame which forms the basis for a concrete like mix made up of a fine aggregate with plasterboard base coat. The box is then wrapped in vinyl.

So how do they sound?

Many who hear this system describe it as the best that they have heard. It certainly deserves the acclaim. What is particularly unique about the sound is the sound stage and imaging. The sound stage is very wide and deep, probably the biggest I have heard. Ordinarily this requires some kind of unusual speaker like an omni or dipole, however these speakers lack image focus. By contrast, Terry's system also has pinpoint imaging. To achieve both at the same time is an impressive feat. 

DEQX demo

A highlight this year was a demo of the capabilities of DEQX. This came about from discussions of my active crossover listening comparisons, in which a small group could not hear any improvement with DEQX. Terry argued that we had dumbed down the DEQX and prevented it from showing what it can do. This is certainly true, we wanted to test sound quality only and in that regard found no reason to spend the extra compared to cheaper options. However, Terry set up a demo in which two profiles were created on DEQX. One was limited to the processing power of MiniDSP and DCX. The other allowed DEQX to strut its stuff. In particular, it was allowed to correct for phase and group delay. We then blind tested this with instant switching, not knowing what was being heard. I was the first to sit in the chair and do the demo and quite soon I didn't need to be told which was which, because the difference was obvious. 

Changes noticed with DEQX:

  • much tighter bass
  • wider and deeper sound stage (quite dramatic)
Both had a basic level of time alignment with digital delays. Both were matched in level and in response closely. These differences were related to the group delay correction. Without it, the sound was flat and almost lifeless in comparison. 

I then watched as others sat through the demo, each person noticing the same differences, differing only in the amount of time taken before declaring what they heard. 

Terry and I have discussed about another demo down the track. He has expressed concern that people don't simply buy the DEQX as a result of these comments. It isn't a trivial matter to design an speaker and there is always the risk after spending a considerable amount, that the result is quite poor. So I put this here as a warning. Don't simply dive in. Remember you take on projects at your own risk. 

More about the Franks

Their sound is clean and neutral. Clearly the system is well sorted and well balanced. They are capable of very high output and it surprises me that dome tweeters were able to keep up. This is easily one of the best systems that I have heard. 

Linkwitz Orion system measurement

The Linkwizt Orion needs little introduction amongst DIY enthusiasts. This project uses the Linkwitz crossover but the baffles are a DIY project that have been well executed. Greg is a graphic designer and the his feel for design is evident in his home as well as in how this speaker has been put together. I prefer this version to the original in terms of its form. 

The Orion is quite interesting regarding its placement. Linkwitz recommends an unusual placement out into the room further than one might typically consider viable. We experimented with various placements and noted their impact on the imaging and sound stage. 

The room has a high ceiling (sloped) and a mezzanine above. It also has many windows and openings onto other spaces. Acoustically it is a relatively complex space.

Various Pass labs power amps:

All of them remain hot even when not powered up.

Temporary acoustic treatment on the rear wall is also seen.


In running sweeps, high levels of reverb were evident. In this regard a swept sine wave can be surprisingly revealing, even more so than music. 

To give some perspective, a relatively dead sounding room would have a RT of 0.2s. This would be typical in a home cinema system. As you can see here, the RT is low in the lower midrange, but above 200 Hz it is 0.5s and in high frequencies it almost reaches 0.7s which is quite high. There is some reduction above 5k but it still remains quite high. A reverb time of around 0.4s might be considered a good ballpark figure for a two channel system. 

Frequency response in the listening position - farfield including room effects:

The bass typically runs about 10 dB higher than the midrange and the treble is slightly reduced.

Bass range only:

You can see a relatively broad suck out centred around 40 Hz. This results in the bass being slightly thin and lacking body and fullness. Later we included the subwoofer and worked on reducing this effect. The fullness of the bass was then improved.

Waterfall response. 

In this case the decay plot turns out to be more revealing:

The red line represents the frequency response, while the purple line shows the decayed version 150 ms later. In my bass integration guide, I use this as an indicator of bass buildup in a room. A well damped room will see the response decay at all frequencies by around 20 dB or more. So the line has been drawn on to show the decay that meets that target. The regions where the rule is not met are highlighted in magenta. You will note that I'm not concerned with below 40 Hz. The area that needs the most work is about 48 - 75 Hz. Above this region the target is barely missed.

Immediately this points to pressure traps as a good solution. They are lower in profile and can easily cover this narrow range without taking up much space.

Over the course of the day, many other measurements were taken. Various sub positions were measured, along with various settings on the crossover and we also tested the EQ system. The crossover on the Orion was also tested to confirm things were as they should be, along with channel balances. For the sake of brevity, these won't be shown.

We also experimented with a little treatment on the front wall. I suggested to the owner the use of narrow band pressure traps in addition to a mix of diffusion and absorption on the front wall, with experimentation to determine the desired amount. Absorption in particular tends to dial back the characteristic open baffle sound and there is some personal taste involved. Our quick trial indicated that with a little absorption, it was desirable to have the speakers closer to the wall.

This system sounded very good indeed. When the JL audio Fathom sub was added, the bass had more weight and fullness and added what was missing. Running test tones down to 10 Hz either with or without the high pass filter in place resulted in extreme excursions on the woofers. The high pass appears to be set around 28 Hz but in my view it should be set higher, more like 40 Hz, allowing for augmentation from a subwoofer. Otherwise, the maximum output is restricted.

I can be a little critical of the lack of focus in the image from dipole speakers - all suffer from this problem, especially in home cinema applications where human vocals are stretched artifically. However, this aspect can be dialed back with treatment. There is a trade off between the sense of sound stage depth, and the sharpness of the image. This will be a matter for the owner to decide where he will set the balance.

The sound as it stands at the moment is very detailed and refined. Clearly, this is an exceptional speaker. The JL subs were clean and articulate and added to the Orions very well.

March 2, 2012

T20 Rythmik dual opposed tapped horns - measurements

T20 has now been measured:

Red and blue lines represent the output of each driver individually, black is combined. You can see that the horn is much less effective with just one driver and the difference is far more than expected. Running together they smooth each other out. The difference suggests that the passive driver when switched off, actually undermines the horn. The plot above shows the response unfiltered.

Blue: Lower crossover

Compared to simulations:

Simulation shown in red. You can see the peak is higher in reality and has more damping. The real version rolls off earlier on the bottom end. With room gain however, the response is flat to well below 20 Hz, even with high pass filters in place. In short, this sub is now a beast!

March 1, 2012

S3 simulation - akabak vs hornresp

Simulation of Se in both Hornresp (green) and Akabak (blue). Past experience suggests that Hornresp is probably right up to to 2k and then the result is likely to be more like the blue trace.