I thought that I would look at making some loudspeakers to be used with a computer. I have four mid range /upper bass drive units I bought to be used with the 8 inch drive units I'm using in the loudspeaker for a valve amplifier, but they are 8 ohms, and two in series or parallel wouldn't fit too well with the flat impedance I want for that loudspeaker, so computer speakers they will become. And as their frequency response is good up to 6kHz, I can cross them over to some planar magnetic tweeters I bought from Australia a year ago, going cheap!
I don't need a lot of bass, so I was looking at how small a box I can get away with, with two drive units in parallel in each box. Well, it seems that a box of only 0.45 cubic feet or 12 litres will give a response down to 100 Hz (3dB down) with a Qts of 0.7071, which has no bump in the bass output around the F3 point. This is, perhaps, not to everyone's liking, as a little bass boost here is usual with most designs of this kind. The box is sealed, as I prefer them to any boxes where the output from the back of the drive unit can interfere with output from the front, causing phase problems. It is interesting to note that it has a similar bass response to the lab speakers design, with a F3 also at 100 Hz, but with a Qts of 1.0. It will be interesting to compare the two.
The upper bass/mid range units are again from Audax, designated AP100M0, and they have a response from 200 Hz up to 6 or 7 kHz, where they can be crossed over to the planar magnetic tweeters, I think made by Foster. Because they will be, by use, near field, they may need special attention with regards to response shaping, but the crossover should be quite simple. The planar magnetic tweeter won't be pushed below a 6 kHz crossover frequency, so I don't have to think too much about it. The mid range units are 6 ohm impedance, and only of about 85.4 dW/m, so two should produce about 88dB, and as each will accept 30 watts (so the manufacturers say) that means significant output. I hope the tweeter can keep up, but it only has to deal with a tiny fraction of the total output, so it might.
The frequency response of the bass/midrange unit is here:
Notice the very smooth response up to 6 kHz, where the crossover will be, but the (lower) so-called waterfall plot looks strange, it has a very long overhang at 3 kHz, and others (though less) at either side of 10 kHz, fortunately above the crossover point. I'll have to work out what is vibrating, and see if I can sort it.
Here are the drive unit details:
And here's the planar magnetic tweeter, by Foster:
and some info about it:
This is to be a pseudo active design, by which I mean that there will be a dedicated amplifier built into the enclosure, with a full passive crossover, but with an external power supply that will supply ±32v DC to each of the loudspeakers. I have already ordered the amplifier parts including the pcb's, and have the transformer, diode bridge and smoothing caps, and suitable heat sinks in the project bag. I'm going to use pine boxes, similar to the ones for the 'valve loudspeaker project', although these will be used edge on, so the baffle will be only 190mm wide.The box is made from 12mm thick real pine, but they will be modified with facings of other materials to get a non-resonant panel. They will need side cheeks, and I intend to use chipboard and either hardboard or mdf facings, forming a sandwich.
The amplifiers are the same as those in my 'lab speaker project', the TDA 7294, producing a good output power of about 50 watts, which is about ideal for the drive units. The heat sinks will be mounted on the back of the loudspeaker box, with connectors for ±32 volts and signal in, maybe with an led on the front to indicate power on.