Matching Drive Unit Levels

[Richard]

Hi All,

Here's the xover for my speakers,



R1, 2, 4, 5 are series resistors used for matching drive unit levels. I use 6R8 mid and 4R7 treble.

My queries are 1) whether there is a downside to this way of matching levels and 2) is there a better way, such as L pads between xover and drive units?

[petercom]

This is the conventional way of attenuating drive units. However it does have a side effect.

If you think of the crossover elements as actually being a variable resistance proportional to frequency then you can see that the series resistance has an effect that is proportional to the driving impedance and the driver impedance.

Hopefully the driving impedance is low and so is not a factor (though it may be a factor with valve amps with no loop feedback and limited bandwidth output transformers).

The driver impedance, however, is also variable with frequency. For example the treble unit voice coil inductance rapidly rises beyond 7-8KHz. As a result the series resistance and crossover network may attenuate correctly at the crossover point but will have less effect at high frequencies.

More accurate attenuation can therefore be had from an Lpad between the drivers and the crossover. This can also allow the crossover to behave as it was designed to in theory. However the Lpad also provides extra electrical damping of the drive unit which may, or may not, be a good thing.

If you have been following my articles in HFW on speaker design you will see that I usually use a series resistance before and after the crossover. There is no hard and fast rule here, but using LspCAD you can easily see the effect on the response and it gives you some values to listen to.

[Wieslaw Lipowski]

I have found a piece of interesting information here:

http://sound.westhost.com/lr-passive.htm#s6.0


Regards
W.L.

[Richard]

Many thanks for the info Peter and Wieslaw.

It was Wieslaw posting me in a pm which got me thinking about how these units are matched. After Peter's explanation, I can see there are pros and cons.

I'd like to try L pads. Would you recommend the wire-wound variable types or is there an easy way to calculate a similar level of attenuation to that used now using fixed resistors? If so, I could then adjust resistors around those values by ear.

[petercom]

You could run a trial using the variable wirewound pots and then replace them with fixed value resistors when you have reached the desired balance.

[Richard]

Yes I suppose trial and error would be best if there are other issues besides simply lessening the signal level. Maplin used to do L pads but no more, does anyone know a reasonable source?

[alnewall]

Richard.

I've got four L-pads made by Violet. Model No. AT-40x. Marked L-8ohm. Vintage items but still ok AFAIK.
Supposed to be reasonable quality and should do for experimenting.
If you think they might be some use, PM your address.
They were destined for the bin anyway.

[Richard]

Very kind Al, will pm you .

I'll post the results of any listening differences. It should be fairly easy to do as the xovers are out of the cabs in external boxes.

[alnewall]

No problem.

[Richard]

Hi Al, They arrived today, so a big public thank you and I'll let you know how it goes. atb, Rich.