Project 4000 Speaker Renovation

[Speedorhaste]

Project 4000

After attending the Greg Osborne GTG a few weeks ago and pondering the apparent lack of bass in my existing speakers, I thought I would launch a minor renovation program to see what can be done. If you guys are agreeable, I thought I would bring this out of the closet and make it a community project. (That way you lot can do the intellectual stuff and I can man the screw driver, router, soldering iron, what ever it may take.)

For ease of reference, I’ll refer to the speakers as “The 4000’sâ€. That was their project name at inception as explained later.

Assumption

I am assuming that these speakers are worth the effort. They are all in good working order, of reputable pedigree, the box is as solid as a rock, so if I can add anything to their performance at reasonable cost and effort, I’ll give it a go. If you think I am wasting my time, speak up and I’ll go shopping!

The history of the speakers

Way back in 1981, a guy by the name of David Tilbrook (chief technical boffin working for electronics magazine Electronics Today International) designed a 4 way speaker that aimed to deliver a high quality product that challenged (in both price and quality) commercially available units (targeted at those costing 3 to 4 times the cost of this kit). Many kits for this speaker design were sold – the kit reference was ETI 4000/1. A smaller 4000/2 version was also published which was only a three way design (dropped the mid woofer). Very little data was published with these designs. At the time I bought a Philips book on speaker design and construction which had most of the drivers included, but even that book did not publish any technical specs for the drivers.

The configuration

The design uses all Philips drivers, a whopper of a cross over (or so I thought at the time), with both woofers housed in their own sealed enclosures. The make up of the speaker is as follows:

Model and cross over frequency

Tweeter AD 01610/t8 3000 hz

Mid range AD 02161/sq8 750 hz

Mid woofer AD 70601/w8 150 hz

Woofer AD 12250/w8

Cross over is second order (12db slope). I’ll post the circuit diagram for the crossover, together with a picture of it, in a subsequent post.

The mid woofer has its own sealed enclosure of 26 litres and the woofer a sealed enclosure of 80 litres. I built the boxes with two layers of 19mm particle board (MDF did not exist in those days). Lots of glue and screws. Real timber veneer all round and black glass inserts on top (have served well as beer props for many years).

The problem

Well, none that I was aware of until I heard a more recent interpretation of loudspeaker design at Greg Osborn’s. The 4000’s have always seemed such a neutral speaker that covered the full range really well. High’s are not overly bright, mid range is not overpowering, bass is very tight (may be too tight). In comparison to any of the Osborn’s, it is a sleepy speaker. Not highly dynamic and lacking in depth of bass.

The Sunday After

Payback for attending the GTG on the Saturday meant I had to spend the following Sunday in the ceiling dragging down boxes of old books, university notes, stuffed toys etc. Time for a clean up under the supervision of my other half. And this was at the expense of having to turn down an invitation to visit Kevin in Torquay to check out his gear. I’ll get there eventually, Kevin.

Well, in amongst the “treasures†I found the original magazine with the design of the 4000’s. Which got me thinking. As I said, no details were published of the spec’s of the speakers. These babies precede the internet (I think) but I gave it a go. I found a document written in Dutch which references the woofer in a comparison of subwoofer designs. While I can’t understand a word of the article, the author had run the spec’s through some speaker design software, screen shots of which were pasted into the document. So, I now have the vital stat’s for box design.

The opportunities

There are a few areas I think need to be considered:

•The cross over - populated with electrolytics and ferrite core inductors

•Box design – sealed enclosure seems to have natural limitations – converting to bass reflex partly addresses these issues, but not entirely.

•Bi-amping (possible with the Yamaha 2600 driving these speakers but not sure of crossover implications

•Replacing the woofer – not convinced this is cost effective, but can’t rule it out

I’ll deal with each of these opportunities separately in subsequent posts. Any input at this stage is more than welcome.

[JA]

Any input at this stage is more than welcome.

[/b]

Hello SpeedorHaste,

Nice to meet. Welcome to the forum. As I undrerstand it you would like to improve these speakers as much as possible- including the bass.

One option as you have already suggested is to port the boxes. Maybe this would be good. Would need to see Thiele Small data for the woofer to go further on this.

What are the box internal dimensions for the woofer - depth inparticular for port length considerations?

Can you post a picture of the crossover board in the speakers. Improving the components in the capacitor would have some benefit. Better caps (not electrolytic) and heavier gauge inductors (possibly air core types??$$$)?

Internal bracing improvemnents (Although 2 layers of particleboard is already quite good).

Adding Sound dampening material (acrylic fibre)

HTH

Best

JA

EDIT:

Have you considered using a subwoofer and rolling of your main speakers at say 50-60Hz?

[Speedorhaste]

Hello SpeedorHaste,

Nice to meet. Welcome to the forum. As I undrerstand it you would like to improve these speakers as much as possible- including the bass.

One option as you have already suggested is to port the boxes. Maybe this would be good. Would need to see Thiele Small data for the woofer to go further on this.

What are the box internal dimensions for the woofer - depth inparticular for port length considerations?

Can you post a picture of the crossover board in the speakers. Improving the components in the capacitor would have some benefit. Better caps (not electrolytic) and heavier gauge inductors (possibly air core types??$$$)?

Internal bracing improvemnents (Although 2 layers of particleboard is already quite good).

Adding Sound dampening material (acrylic fibre)

HTH

Best

JA

EDIT:

Have you considered using a subwoofer and rolling of your main speakers at say 50-60Hz?

[/b]

Hi JA,

Thanks for your reponse. I'll post additional info as I get my act together, including cross over deisgn, bass driver spec's etc. You are right re subby - this could be the easiest solution of all!

[JA]

Hi JA,

Thanks for your reponse. I'll post additional info as I get my act together, including cross over deisgn, bass driver spec's etc. You are right re subby - this could be the easiest solution of all!

[/b]

Yes, it depends on how much extra bass you want.

In the meantime I have found an article in German on the AD12250 with t/s specs- not sure about their accuracy since it is not from Phillips, but it was enough for me to conclude that your SEALED boxes should be -3dB at 50Hz and -9dB at 30Hz using software.

Based on the t/s parameters it appears your driver could be used as a sealed or vented box driver. If you added a 75mm port 18cm long this would tune the box into the 20-30 Hz range and your NEW VENTED boxes would then be -3dB at 40Hz and -6dB at 30Hz. You would certainly notice this. Of course going from sealed boxes to vented increases the group delay yadah yadah yadah........... For the cost of a port or 2 it's worth a try. You can always plug the hole if you prefer sealed bass. Jaycar has various size ports with flanges to make installation easy.

Best

JA

[Speedorhaste]

OK, here's my second post -

The cross over

The crossover circuit diagram is attached to this post. Being a 4 way design and a first cross over point of 150hz, it requires some high levels of inductance. With 12.6 and 3 mh inductors, these were constructed with ferrite cores (I think – might be iron) in order to keep the size and cost under control. Resistors are regular wire wound (in ceramic cases) and the caps were bipolar electrolytics.

Attached is a scanned image of the cross over. Apologies for the quality

Opportunities

• Replace the caps

• Replace the resistors

• Replace the inductors with air core

• Redesign (god forbid)

Replace the caps

As mentioned in post 1, the caps are mostly bi-polar electrolytics. All I have read on the forum in this respect suggests they be replaced . What I need is guidance on brand, supplier and specs to look for.

This appears to be a relatively cost-effective step, provided replaced caps is going to add to the sound quality equation.

Replace the resistors

Now I must admit I haven’t seen much commentary on this forum about quality and selection of resistors. Those installed are 10W wire wound ceramic encased units. Is there any merit in upgrading these?

The only time I have had problems with resistor selection was in the first amp I built. The designer warned to use low inductance resistors, otherwise the amp could become unstable. Well, it was unstable when first built. The supplier of the parts kit included poorly chosen output resistors.

Replace the inductors with air core

I suspect everyone (anyone?) reading at this point is nodding furiously, thinking “Yes, the inductors are critical to the overall quality of the crossover, and not being air cored, they are going to be introducing all sorts of anomalies into the speaker performanceâ€.

I need to know if this is true or false. The largest sized inductors are 12.6 and 3.0 mh (2 of each). You know what that means, don’t you? Air cored 12.6 mh inductors turn out to be about 20cm in internal diameter, 5 cm tall, 150 metres of wire, therefore very heavy, expensive (I presume) and somewhat difficult to fit into the boot of your car, let alone tucked inside the speaker box (the cross over is fitted inside a cavity in the bottom of the box).

Are the existing inductors so bad? Can I keep them and not compromise any other efforts to improve these speakers? Or is their replacement an imperative?

Redesign

Not too keen on this AT ALL, but need to consider it as an option. As I don’t have much in the way of specs for the drivers other than the woofer, I’d be guessing a bit and doubt I could do a better job than the original designer. Being four way the cross over is relatively complex. No way would I go with anything more than 2nd order (think I would need another box for them if that was the case). So unless someone out there can convince me, the spec’s are what they are for a good reason.

Next

I 'll post what I have been able to do re modelling the bass driver and a modified enclosure

[JA]

cAN YOU POST ATACHMENTS AS JPEGS- THESE pdfs won'r open for me.

Best

JA

[Speedorhaste]

Yes, it depends on how much extra bass you want.

In the meantime I have found an article in German on the AD12250 with t/s specs- not sure about their accuracy since it is not from Phillips, but it was enough for me to conclude that your SEALED boxes should be -3dB at 50Hz and -9dB at 30Hz using software.

Based on the t/s parameters it appears your driver could be used as a sealed or vented box driver. If you added a 75mm port 18cm long this would tune the box into the 20-30 Hz range and your NEW VENTED boxes would then be -3dB at 40Hz and -6dB at 30Hz. You would certainly notice this. Of course going from sealed boxes to vented increases the group delay yadah yadah yadah........... For the cost of a port or 2 it's worth a try. You can always plug the hole if you prefer sealed bass. Jaycar has various size ports with flanges to make installation easy.

Best

JA

[/b]

JA,

You're too fast for me!! Yeh, I found an article on the internet in Dutch. Fortunately, I have someone at work who is fluent in Dutch, so have been able now to understand what the writer was doing - also got the t/s stats and ran it through some software and I think we have similar outcomes. I'll post separatley what I got in the way of modelled outcomes. It is very helpful, thought, to get this confirmed by others like yourself who have "been there before".

Thanks

Speedorhaste

[JA]

Hello Speedorhaste,

Leave the resistors alone.

Change any inductors if you want to air core types. I did this on Magnepan MG2Bs using Solen Flat copper Foil inductors and got wonderful results- the inductor was big- close to the size you said!! Can't remeber the value now, but is was nowhere near 13mH. (could be 1.6mH- cloudy on this)

Change the capacitors to plastic film types, "Audiophile" types if you want to spend some $ (not). Iwould certainly do this for all the electrolytic capacitors say 20uF and below. Above this u decide- it just gets more expensive, but you can use metallised polypropylene motor start/run caps in the 10 to 50uF+ range.

Post a photo JPEG of the crossover board to see component layout.

Best

JA

[Drizt]

cAN YOU POST ATACHMENTS AS JPEGS- THESE pdfs won'r open for me.

Best

JA [/b]

hehe, sounds like you need to download adobe acrobat

[JA]

hehe, sounds like you need to download adobe acrobat

[/b]

Hello Drizt,

Maybe...But I have other pdfs that open perfectly. Not sure perhaps I will check the current version. Still good practice would suggest it be backwards compatible- unless my copy is corrupt.

Best

JA

[Drizt]

Hello Drizt,

Maybe...But I have other pdfs that open perfectly. Not sure perhaps I will check the current version. Still good practice would suggest it be backwards compatible- unless my copy is corrupt.

Best

JA [/b]

Not sure, but it works fine on mine

Maybe just try downloading and installing the latest version and see how you go.

[Speedorhaste]

Here's the files in JPG. Scanner is mis-behaving - they took a while to come up and quality of picture is a bit dodgy. Hope this is better than the PDFs.

[JA]

Here's the files in JPG. Scanner is mis-behaving - they took a while to come up and quality of picture is a bit dodgy. Hope this is better than the PDFs.

[/b]

Ok Speedor haste,

Changing the capacitors to plastic film caps should be OK- your highest value is 47uF (you can use several smaller caps or 1 large 40uF (motor start)+6.8uF cap. Sometimes you will even find a 45uF value in these. They are a lot bigger and a 40uF cap will be probably 5 to 10 times larger- depending on its voltage rating.

Also the two smaller air core inductors for the tweeter and the dome mid will couple inductively with both of them aligned the way they are pictured. Rotate one of them through 90 degrees to stop this. The high pass on the tweeter and dome mid could be simplied to 6dB slopes, but that is a redesign issue etc.....

Any way that should give you enough work for a few hours. :biggrin:

Best

JA

[Speedorhaste]

OK, JA is onto me with respect to the bass driver. Here's what I have been able to do so far with the spec's I found:

The speaker box

As mentioned, this is a sealed enclosure design. Having obtained the woofer specs, I ran them through WinISD and graphed the predicted response for three box designs:

• Existing sealed enclosure

• Ideal vented enclosure

• Vented enclosure constrained to the existing box size

These three outcomes are plotted on the attached screen shot. I think you can work out which is which.

In its current configuration, the -3db point is around 55hz. Not too impressed about that! The ideal vented design requires 113 litres of space and gives a -3db point of around 31hz. The compromise, using existing box but vented, looks to deliver a meaningful improvement with -3db at 36hz.

This is as far as I have progressed with my research so far. Will absorb responses to date and summarise most plausible solutions.

Cheers

[JA]

The compromise, using existing box but vented, looks to deliver a meaningful improvement with -3db at 36hz.

[/b]

Do it- cut the hole.

Best

JA

[Speedorhaste]

Ok Speedor haste,

Changing the capacitors to plastic film caps should be OK- your highest value is 47uF (you can use several smaller caps or 1 large 40uF (motor start)+6.8uF cap. Sometimes you will even find a 45uF value in these. They are a lot bigger and a 40uF cap will be probably 5 to 10 times larger- depending on its voltage rating.

Also the two smaller air core inductors for the tweeter and the dome mid will couple inductively with both of them aligned the way they are pictured. Rotate one of them through 90 degrees to stop this. The high pass on the tweeter and dome mid could be simplied to 6dB slopes, but that is a redesign issue etc.....

Any way that should give you enough work for a few hours. :biggrin:

Best

JA

[/b]

Thanks JA,

Plenty to ponder over at this point. Re one of your earlier queries, the boxes were lined with carpet underfelt, not wadding stuffed inside. 25 years ago this was all that was avaialble to me. Presumably going vented would warrant installation of some additional sound absortion material. Not sure on the quantities in that regard.

Speedorhaste

[Speedorhaste]

Do it- cut the hole.

Best

JA

[/b]

I'm convinced! All I need is a fanfare and to clear the room of witnesses. You recommended a port dimension of 75mm by 18cm. The software I ran gave different dimensions. At 75mm diameter, it suggests 13.43 cm length. The difference is obviously in the tuning frequency. My calc's were at 29.55hz, which seems to give the best-fit response curve relative to the ideal box size curve. Is selecting the tuning frequency an art or is there more science in this decision?

[joz]

I'm just sitting back and watching all this unfold.

It's all Greek to me

Go on boys maybe I'll start to pick this up,...maybe not?

[Aslan]

This looks like a great project speedorhaste...............It sent me diving through my old literature to look for my "Building Hi-Fi Speaker Systems" book published by Phillips themselves in 1979.

Alas, no mention of the 12250..........the largest woofer listed is the 12100, so the 250watt version must have come out between 79 and 81.

JA has you on the correct path, change the smaller value caps first for poly ones, cut the bloody hole and stick in a tube, then look at the inductors in series with the drivers (IMHO components across the drivers have less sonic impact than those in series.............JA may have a differing view).

This is an interesting project as I remember Tilbrook's work well from my youth.

A/

[Speedorhaste]

This looks like a great project speedorhaste...............It sent me diving through my old literature to look for my "Building Hi-Fi Speaker Systems" book published by Phillips themselves in 1979.

Alas, no mention of the 12250..........the largest woofer listed is the 12100, so the 250watt version must have come out between 79 and 81.

JA has you on the correct path, change the smaller value caps first for poly ones, cut the bloody hole and stick in a tube, then look at the inductors in series with the drivers (IMHO components across the drivers have less sonic impact than those in series.............JA may have a differing view).

This is an interesting project as I remember Tilbrook's work well from my youth.

A/

[/b]

Thanks for your encouragement. I have been thinking about this for nearly a month and so did a fair bit of hunting around before I started the post. I have the same Philips book. Hasn't been opened for probably 20 years but found it on the book shelf waiting for this day. As you suggest, it comes up short on this particular driver. Does the "250" in the 12250 part number denote its power rating? I always thought it was only a 100w driver. And yes, I am pretty sure the driver was released in 1981. It was summer of 1981 that I built these speakers - in the back yard with no cover. Obviously it didn't rain that year either.

Re "the hole" - I just need to finalise the calc's for this and then source the right size units. I note Jaycar's website does not list 75mm sizes that JA mentions. I was thinking of rolling my own with 60mm or 90mm PVC pipe - diameter will depend on the available landscape around the driver. Because of limited options with speaker placement in the room, I am going to have to front mount the ports - hence space might be at a premium.

With inductors, apart from having to buy a better multimeter (I have inherited three but none can measure inductance), I feel inclined to wind my own, again subject to what I can find available. All recommendations will be kindly received.

Regards

[JA]

Hello Speedor haste,

You can addd damping and measure etcc....But all I would suggest is to staple down the existing wadding and then staple 25mm+ or so acrylic fibre insulation on top. Or try it with and without.

Best

JA

[JA]

I'm convinced! All I need is a fanfare and to clear the room of witnesses. You recommended a port dimension of 75mm by 18cm. The software I ran gave different dimensions. At 75mm diameter, it suggests 13.43 cm length. The difference is obviously in the tuning frequency. My calc's were at 29.55hz, which seems to give the best-fit response curve relative to the ideal box size curve. Is selecting the tuning frequency an art or is there more science in this decision?

[/b]

Hello Speedorhaste

The design I specified was tuned lower- cant recall exactly probably around 24-25Hz. If you start with a longer vent you can always trim it. Selecting the tuning freq. is a tradeoff between LF extension, LF linearity above resonance, group delay considerations, power handling, increased distortion at lower frequencies etc...

I recall that fs=25Hz, Vas=200 l (I think), Qes=.43, Pe=180Watts, Xmax not specified, cant remember the others and looking for the pdf article I mentioned. How do these values compare with yours?

Best

JA

[Aslan]

Speedorhaste................Jaycar have some economical air wound inductors, as well as metalised poly caps which will yield an improvement whilst keeping costs down.

If you cant buy off the shelf ports to suit, PVC pipes are excellent and do come in an 80mm size as well as 60 and 90.................port length will need to be adjusted accordingly. Just make sure you have at least 50mm clear space around the back of the port.

[JA]

I note Jaycar's website does not list 75mm sizes that JA mentions. I was thinking of rolling my own with 60mm or 90mm PVC pipe - diameter will depend on the available landscape around the driver. Because of limited options with speaker placement in the room, I am going to have to front mount the ports - hence space might be at a premium.

[/b]

Hello Speedorhaste,

Front mounting the ports is better. I'm glad you bought up the 75mm issue, just use whatever is convenient from Jaycar or Speakerbits etc,.... If you go for 90mm the duct will need to be longer for the same tuning frequency though. Make sure you allow sufficient space inside the enclosure between the end of the duct and the enclosure wall. I would say 2x the duct diameter would be an adequate clearance.

Best

JA

[Speedorhaste]

Hello Speedorhaste

The design I specified was tuned lower- cant recall exactly probably around 24-25Hz. If you start with a longer vent you can always trim it. Selecting the tuning freq. is a tradeoff between LF extension, LF linearity above resonance, group delay considerations, power handling, increased distortion at lower frequencies etc...

I recall that fs=25Hz, Vas=200 l (I think), Qes=.43, Pe=180Watts, Xmax not specified, cant remember the others and looking for the pdf article I mentioned. How do these values compare with yours?

Best

JA

[/b]

JA,

I think we must have found the same article on the driver. My specs are the same, though the print was a bit fuzzy and couldn't work out if Pe was 100 or 180. Not sure if that makes a difference or not. The specs I used are shown in the screen shots on page the attached file. The software I used fills in the missing spec's based on what is known. I see what you mean about the LF trade-offs - playing with the tuned frequency subtlely changes the shape of the curve.

Regards