Syntrillium Backup Forums :: View topic

teflon20 · Location: Australia

Hello.
I have 20+ year old plessey 3-way speakers which use a cheap way of speaker crossover network.(2 capacitors).

The question is should buy a proper crossover network? (inductor type)
I have ordered better capacitors to put in to see if there's any difference, but I'm not sure wheather a proper crossover network will make a big difference with these old no so great not so bad speakers. Any suggestions.

jonrose · Location: USA

My first bit of advice would be to watch those crossover frequencies - and thereby, your parts values. Hopefully you've ordered capacitors whose values which are at least close to, if not exactly, what the originals purported to be; and matched or exceeded their stated working voltage (pretty important Wink

); and made sure they were bi-polar units ( Wink

).

With regard to replacing the original caps with a crossover network, which employs inductors (as well as capacitors - but I'm guessing that's what you meant to say), I would again watch the values closely, as you might need to stay within bounds of the originally designed crossover points to avoid hi-frequency driver damage (from choosing too low a crossover point - despite the steeper curve the LC network might afford you, you can't go too low without potential damage).

Keep in mind that this may indeed change the overall speaker sound, as the rolloffs will be somewhat sharper with an LC network. If that's desirable, then that's fine, but you won't be able to really compare to what it used to sound like, as the old capacitors have undoubtedly changed value over time.

The main point is that with steeper rolloffs, and possibly different crossover points, these could sound better or worse, depending on how the designers thought out the interactions of response of each driver, the crossover points, and the filter slopes, the box design, driver quality, etc.

Just some thoughts.

teflon20 · Location: Australia

Thanks for your reply jonrose.
What does driver mean? Is that the actual speaker? I'm not sure.
What I was refering to in inductor type was wheather I should buy a ready made crossover network board that includes inductors.

Yes I did get bi-polar, same values and higher voltage caps which are audio grade. The previous caps were made by elna.
I haven't got a meter that measures capacitance, so how much do you think the values might be off in the old caps.
I have a 100W RMS receiver which my techi says that the previous 50V rated caps weren't enough and that higher 100V caps are needed.
I feel the speakers sound like they have lost there low and high frequency power and quality, that's what made me look at and change the caps.

Well buying new caps is certainly cheaper than getting a ready made or even a custom made crossover speaker network board to fit.
I will be getting the new caps in about 3 weeks.
I'll try them out and tell you the results.
Thanks bye.

jonrose · Location: USA

Yes, drivers = speakers. You've got at least two drivers per speaker cabinet, obviously, perhaps more.

It sounds as if you've ordered quite adequate replacement parts, so unless you're a tinkerer that likes to burn money, perhaps the best bet is to just install those and be happy.

No telling what the current values are for your old capacitors. That depends on a lot of environmental and usage factors, as well as the original materials and design/construction. If you know a shop that has a "Z" (impedance measuring) meter, and you're really that curious, then maybe you could cajole the folks there into telling you what values these things test-out at. And, er... Beer usually works pretty well for this kind of thing...
Wink

teflon20 · Location: Australia

Beer is good but I don't like it. Maybe you mean to give them a beer oh right.:)

I had a look at my speakers again and it inludes 'by foster' Made in japan. These speakers came with an old pioneer system.

Do you know about foster? (no not the beer) Big Grin

Do you know where I can look up specs on these speakers?

SteveG · Location: United Kingdom

I don't know whether there's older drive unit info on it, but the company website is here. Somewhere, I have one of their old dynamic mics - they were actually pretty well built considering their price. We are not talking about the posh end of the market here, but the company iteslf has been around a long time, so they must be doing something right...

You may notice a difference in your speaker sound if the original caps are old dried-out electrolytics back to back in a single container (the so-called 'non-polarised' ones). The best replacements are sealed paper ones, if you can get the correct values, but these are rather less critical than you might think - the tolerance values on all large value caps is quite considerable anyway.

teflon20 · Location: Australia

Thanks for the info steve.
The value of the capacitors are both 4.7 microfarads.(I couldn't be bothered including the greek m in micro) he he.

djwayne · Location: USA

You might want to look into a variable crossover to fine tune your sound. I have a variable filter on my sub-woofer that can be fine tuned from 60 hertz to 180 hertz. You can really clean a sound up if it sounds too muddy, and get more of a punchy sound out of it.

SteveG · Location: United Kingdom

djwayne · Location: USA

Exactly. With a three-way crossover system, he'd have more control of each speaker, and could fine tune it to his liking. Older speakers would perform diffently from new ones, as I'm sure the age of paper cones and coils, plays a part in reproducing sound.

Thirty years ago, I had friends in two different bands which had different PA systems. One band had the old Shure Vocal Master PA with Kustom PA cabinets with 4- 8" to 10" speakers in each cabinet, the other band had a new-fangled three-way amplification system with the three way crossover network, horns mids, and sub cabinets. Which one do you think sounded better ??.......It depended on how drunk the sound man was. Some of the sound men would wire it up backwards and blow out speakers or horns. hehehehe !!!

jonrose · Location: USA

Since tef noted that the caps were "both" 4.7 microfarad units, I'm guessing these are just two-ways, though...
Tef? Confirmation?
Wink

teflon20 · Location: Australia

I think it's 3-way.
There's a tweeter, midrange and a subwoofer.
Wheather there are wired as 2-way I'm not sure.
But they definetely sound like 3-way.
Unfortunetly these speaker system don't have a model number or the model number has come off, so I can't check for specs.

SteveG · Location: United Kingdom

teflon20 · Location: Australia

Thanks for input men.

But as I said I will insert the new caps and judge for myself.
If the speakers don't sound good even after the new caps I think I might give Foster a call for a new pair all together. Wink

I think it's time for an upgrade anyway.

teflon20 · Location: Australia

My techi called and I got my caps early and have installed them.
The way the speakers are connected - the sub is connected straight and the the midrange and tweeter are both connected via seperate caps both using the same value.

As bad timing goes I have a middle ear infection!

As far as I can tell.. things sound better in bass and midrange but I can't really tell properly for the high's because of my ear infection.
Also I found a setting on my receiver to cutt off frequencies below 100hz which reduces distortion in the subs.
Overall I'm pretty happy with a A$5 upgrade.

alofoz · Location: Australia

The tri-amping idea is definitely a good one. I've recently done exactly that using a Linkwitz-Riley Active Crossover design. And I used Cool Edit to help me determine the best crossover frequencies for my drivers as well as the correct gain for each amp.

Tef, if you decide to do this (recommended) it'll be easier for you now than it was for me. Jaycar have a $50 active crossover kit that will do exactly what you want. Of course you'll also need 3 stereo amps, no bells & whistles but good performance. And write off the $5 for those caps!

For background info here's another Aussie site. Explore the site a bit there's more relevant stuff there. The Jaycar kit acknowledges this site.

I think the site explains why you might want to use this design, so all I'll say is that the Linkwitz-Riley design consists of two Sallen-Key active filters in cascade for each crossover each with a 2nd order Butterworth response. You can easily simulate this in Cool Edit.

My system is different from yours but here's a brief description. My intention was to have a smallish 2-way plus a sub-woofer. So I designed a 25 litre system using a C program I wrote based on Thiele-Small parameters. This is fairly respectable on its own with passive crossovers, but the active crossovers & sub certainly make a difference. The "woofer" in the 25 litre system is a 6.5" model.

Before picking up a soldering iron I used CEP to simulate the design. I did this in mono for simplicity - stereo isn't needed here. Here's the procedure:-

1. I made mono recordings of music I'm familiar with and with different characteristics. 4 copies of each.

2. I applied scientific filters to 3 of these (details below*) so that I had bass, mid, treble and the original wideband versions of each musical piece.

3. I inserted these into multitrack. Each of the tracks was routed to a different output on my sound card, then to 2 stereo amps which provided 4 amps. I had previously set the amps' volume controls to have the same level which I then left unchanged.

4. The amp outputs for tracks 1-3 were fed to tweeter, mid range (i.e. the 6.5" "woofer") and the sub woofer. The 4th output (track 4) was fed to a reference speaker.

5. I muted the track feeding the reference speaker.

6. I played the music, adjusting the relative levels of the bass, mid & treble.

7. I clicked the solo button on track 4 to A-B with the reference speaker. Doing this cuts out tracks 1-3 at the same time cutting in track 4.

8. Repeat steps 6 & 7 until satisfied.

9. Repeat steps 1 to 9 for different crossover frequencies until very satisfied.

10. Now that you have found the best crossover freqencies, calculate the R and C values needed. as a rough guide here, keep the resistor values between 10k and 33k. Choose a preferred value C, calculate R, choose the closest preferred value, recalculate f. That should be close enough. Build the crossover.

Once the crossover is built you can adjust the levels for each frequency range. Use the mixer fader settings from step 6 as your starting point.

* To simulate the Linkwitz-Riley response use the scientific filter with a 2nd order butterworth response. Choose the cutoff frequency and Low pass or High pass. Click OK. When finished press F3 to repeat the process. This gives the proper 4th order response for a Linkwitz-Riley filter. Note that it must be 6 dB down overall at the cutoff (i.e.crossover) frequency. For a LPF the corresponding HPF filter must have the same cutoff frequency and vice versa.

To state the obvious: For the bass file you'll need a LPF, for the mid file you'll need the corresponding HPF and a LPF at a higher frequency. For the treble file you'll need a HPF to complement the mid file's LPF.

The Jaycar kit's details were published in The Jan 2003 issue of Silicon chip. The formula given in that article to calculate the Rs & Cs is incorrect. It should be:

f = 1 / (2 * pi * R * C * sqrt(2))

teflon20 · Location: Australia

Thankyou very much alofoz, I will look into this.
Does it cost $50 + $50, for Left and Right speakers.

alofoz · Location: Australia

No, that's for both channels. A complete unit will also require a case & power supply and you might find the total coming to rather more than $100. Worth doing though.

motorhead6 · Posted - Mon May 26, 2003 12:46 am

I didnt know there was such a thing as a 2 capacitor no inductor crossover.

motorhead6 · Posted - Mon May 26, 2003 12:52 am

BTW 4.7 microfarads is a pretty large value of capacitance which means its going to be a fairly low frequency cutoff point. The lower the value of capacitance the higher frequency the cuttoff will be. 4.7 sounds low but a low value would be in picofarads. I still dont see how two capacitors would work. Maybe a capacitor and a resistor, although it wouldnt be that great but I dont see how 2 capacitors would do it. Im not doubting it but I just dont understand how.

SteveG · Location: United Kingdom

motorhead6 · Posted - Thu May 29, 2003 3:50 am

{f=1/2*pi*RC }

Im hip to that.If you use an inductor and a capacitor you can replace that RC with the square root of LC. L being inductance.