Recycling 20 year old drivers from a Sony tabletop clock-radio

[Flint]

When @Towen7 @Zing @Babs and Christy visited me a few weeks ago we were on my rear porch for a bit and I was going to show off how I put music out there via a Amazon Echo Dot and an old Sony Tabletop Clock/Radio. I was going to show them how good that old radio sounded, but alas it would not turn on. Very disappointing.

I have made plans to build new outdoor speakers and already purchased the drivers, but I didn't want to waste the speakers in the radio since I really enjoyed how they sounded. I decided to take apart the radio, at first to fix it. When I realized I could not repair it, I took it apart to salvage whatever parts I could use later. I kept the cooling fan and all of the screws, as well as even the power supply. I then took out the speakers, tested them to ensure they were both functioning properly, then decided I needed to build a set of bookshelf speakers out of these two little drivers.

Rather than go through the tedious process of measuring the T/S parameters of the drivers, I knew they sounded good in the radio, so I merely measured the enclosures they were in with intentions to replicate the performance in stand alone enclosures.

So, off I went.


Photo 1: Stock image of the clock radio I was salvaging



Photo 2: The radio with the front grill removed


Photo 3: Drivers after removing them, note they have counter magnets to make them safe for use near an old-school CRT television or computer monitor

 

[Flint]

Design

To save money and time, I chose to build a super-simple cabinet with a standard MDF front baffle and an enclosure and port made from PVC pipe and caps. Basically, the front would look normal, but the rear would look like plumbing. I simply calculated the internal volume of the old table radio chassis speaker enclosures and applied that to the PVC pipe I felt would work best. I already had the pipe, I only needed to buy the end caps and the angle adapters for the port (see more below). So, my only expense was initially a few simple PVC parts. I did have to get a new package of epoxy to adhere the PVC to the MDF and hold and seal the ports in the enclosure section. Simple, really.


Build

I used a very simple build approach to make it easy. Since I wasn't chasing ultra-high-fidelity, I wasn't worried about baffle size or shape or enclosure dimension optimization. I just slapped it all together as I needed to make it work.

Here are some photos of the parts in process:



Photos 4 & 5: Rear Pipe Caps after Drilling for Terminals

I chose to use standard banana terminals in the rear pipe cap, so I needed to make mounting holes which were flat and square. I drilled a small set of pilot holes 3/4 inch apart, then used a Forstner bit to cut smooth and flat "countersinks" for the terminals which I recovered from my old failed main speakers. I then drilled out the pilot holes to accommodate the banana terminals.



Photo 6: The cabinet parts to be used

I drilled out a hole in the side of the large 4" PVC Pipe which would make up the main enclosure to accommodate the port tubes, but I didn't have the perfectly sized Forstner bit to fit the pipe I had. So I used the next smaller bit and then reduced the outer diameter of the port pipe to fit the drilled hole. It was tight enough that epoxy could easily seal and affix the port.



Photo 7: Front baffles after all mill work

I discovered the speakers would touch the port tubes once assembled, so I put a small ring of MDF in front of the original driver hole to extend the front edge of the driver. Worked well, but I knew it was going to cause a diffraction issue. I measured that impact and posted charts of it below.


Photo 8: Baffles after Sealing with Shellac and the rear caps with the Terminals installed

 

[Flint]

Finishing Up

I painted the baffles a burgundy color I happened to have on hand.


Photo 9: All the finished parts before assembly

I used epoxy to glue the PVC to the MDF. Then I used epoxy to fill the gaps, seal, and affix the 1" port pipe into the enclosure's 4" pipe. Finally, I pressed the end cap all the way onto the enclosure pipe, choosing not to glue it on so I can work on these in the future if I want to. I may decide to add acoustic material in the enclosure, or even make the enclosure lower volume.


Photo 10: Front view of finished speaker



Photo 11: Rear view of speaker

 

[Flint]

Measurements

It would have been nice to measure these before removing them from the table radio, but that didn't occur to me when I was disassembling the old thing. Still these measured very well.


Chart 1: On axis response at 1W/1W

This is not a bad response at all. I could rate it as 100 - 20,000 Hz +/-5dB.

It is also louder than I anticipated with a sensitivity I could rate as 77dB @ 1W/1M



Chart 2: Measured off-axis response at 1/2 Meter at 0, 15, 30, 45, and 60 degrees

The off axis response proved to be VERY good as well. The small 2.5" cone driver does a good job of dispersing the sound.


Chart 3: Close Mic measurement of driver and port

I was really happy with the enclosure loading, though I think I can improve the output below 200Hz if I make the enclosure smaller and round off the outer portion of the port, thus rendering it effectively shorter (and thus raising the tuning frequency).

That said, the speakers do well to fill the room given their size.

 

[Flint]

Listening

I hooked these up to a Parasound HCA-750 in the middle of my library in front of my drum kit to listen to them. I was immediately impressed with the sound these tiny little speakers can generate. They were clean with plenty of top end treble, though a tad soft, and decent bass which was more present than expected. Being full-range and requiring no crossover to blend two different drivers together, they had a very natural and smooth midrange. The imaging was what you'd expect. I would like a tad more bass output in their operating range (I can never expect deeper bass). I found I could turn up the bass EQ control on the mixer I was using the feed my amp by just a small amount and I liked the sound better.

At any reasonable listening level for such a small speaker they did fine. I would imagine their ultimate use would be in a bedroom or on a desk and the loudness demand would be low. I have no intention to push them hard on a daily basis.

 

[PaulyT]

Fun!

 

[Flint]

Edge Diffraction

I've ranted on and on about edge diffraction and how important it is to the performance of speakers. Since I made the quick rash decision to put these speakers dead center on a small perfectly round baffle, the worst that could happen has happened. You can see in the off axis chart above that when measured directly on axis there is a significant amount of variety in the midrange compared to all of the off axis measurements. This is because the edge diffraction (which is always present) when measured on axis is always exactly arriving at the same delay from all the edge circumference of the baffle, thus increasing the amplitude of the gain or null.

Here's a good was to show it:

I made a measurement of the speaker's response on axis then used the leftover wood from the ring cutout and some acoustic foam to reduce the diffraction to almost zero.


Photo 12: Reducing the Edge Diffraction to almost zero

I then measured the response again without moving the mic or changing the level settings or any settings on the measurement software. Essentially every aspect of the measurement process is identical other than applying stuff to cut down the diffraction to nearly zero. Here's the resulting charts:


Chart 4: On axis response with and without diffraction from baffle edge

As you can see, there is a gain in the output from about 1,800Hz to 4,000Hz, with a peak gain of over 7dB SPL. Then there's a huge dip where the diffracted sound creates a null at 4,800Hz, then the gain returns as the edge diffraction is no longer out of phase with the direct sound. This is real, and it is audible. I may decide to do something to address it more permanently in the future.

 

[Flint]

So, that's my little project.

I need to find somewhere to use these things. Maybe in my second spare bedroom, but I need a sound system to drive the speakers.



Next up is building my new patio speakers.

 

[Randy]

Nicely done again Flint. Very cool little project.

 

[Zing]

How does one determine whether they could benefit from add-on diffraction control? Is it simply if the tweeters and drivers are centered on the front baffle? Or does on/off-axis listening come into play too?

 

[Flint]

How does one determine whether they could benefit from add-on diffraction control? Is it simply if the tweeters and drivers are centered on the front baffle? Or does on/off-axis listening come into play too?

Diffraction is bad regardless of the on or off axis listening angle, but the frequency response impact is the easiest to demonstrate with a measurement.

You can look at the design of your speaker and make some determinations on how much diffraction there is by looking at how sound from the tweeter can slide along the surface of the baffle. Imagine it like the surface on water, and everywhere there's any sort of sharp break over that surface, some form of diffraction could occur. Using a waveguide can vastly reduce diffraction because, like a horn, by the time the sound is near the edge of the waveguide it has been directed away from the face of the baffle - but it depends on the depth and contour of the waveguide AND how low in frequency the tweeter operates.

I've written so many threads on diffraction, I'd hate to repeat it again, but I will make a list of things to look for on a speaker which might mean there are more diffraction issues which could be addressed.

1. Tweeter mounted onto the top of the baffle (not sunken in for flush mount)
   
2. Tweeter placed in the center of the baffle
3. Tweeter at the center of any textural or prominent ring on the speaker baffle
4. Sharp corners or edges on the baffle or baffle edge near the tweeter
5. Square, sharp edges on the grill frame, if using a grill
6. Straight, sharp lines cut into the face near the tweeter

If you have some or all of the items listed above on your speakers, maybe you want to look into adding treatment for the tweeter.

 

[Flint]

I posted this in my Rocketman speaker thread and thought I'd put it here as well.

Here's the measured harmonic distortion measurement for these little speakers:

Note that the title in incorrect and should read "84dB SPL @ 1/2M"

The distortion above 300Hz is better than 5%, which isn't terrible for a tiny speaker from the 1990s. They sound pretty cool despite the rather high distortion, though.