On Electrostatic Tweeters

[Edward Huggins]

Between 1957 and 1960 there seemed to be a bit of a fad to fit Electrostatic Tweeters into UK built audio equipment. These were quite low cost units made by Grundig and Isophon. Certainly Pye, Bush and Philco used them.

At the same tine UK manufacturers such as Elac, Goodmans and Celestion were already making quite adequate cone tweeters (typically 3" units) for a similar price.

By the early 1960s these imported drive units were rather quickly abandoned, but I can never quite see what their initial attraction was?

Do any Members' memories go back that far to help shed some light on this?

I always preferred the smoother sound of a coned tweeter, but that's just my taste.

[mhennessy]

Perhaps obvious, but the change to transistor amps would have made driving them more difficult. I'm sure there are many other reasons, but that seems like a compelling one to me at least.

I've never heard one in good condition, so can't comment on their quality. But every 3" cone tweeter I've heard has been pretty rough by comparison to the small dome types that I'm used to. Yet some larger full-range units can actually have a pretty reasonable HF response.

The rudimentary crossovers that were common (often just a single cap) couldn't have helped, and I suspect that manufacturers tended to run the tweeters a bit "hot" to remind customers that they paid extra to get the tweeter. With any multiway speaker, the crossover design makes a huge difference to the sound quality - it's easy to forget (or underestimate the effect of) this.

So a valid comparison between the two tweeter types would have to be set up before any conclusions could be made about the respective technologies - e.g. the same woofer with each, and careful measurements to ensure the levels and frequency responses are matched in both cases. Could be a fascinating experiment!

My take is that on a valve amplifier the bias voltage is already there, they don't need the 'constant charge' of real (full range) electrostatic speakers because the movement is very small, nice advertising copy and cheaper to make than a moving coil 'speaker.

[G6Tanuki]

Yes, connect the electrostatic mini-tweeter from the output-valve's anode to ground and you get DC bias as well as using the output-transformer as a "HF choke" to provide the 'crossover'.

Then boast in the ads about "Binaural tone" or something similarly nebulous-but-impressive-sounding, and add at least twenty quid to the sale-price.

[mhennessy]

The ones I've seen use small capacitors between valve anodes and tweeter, with resistors from HT to provide the bias. Obviously these resistors and capacitors form the high-pass filter.

I can't immediately see how you'd get away with a direct connection between anode and ground (unless the tweeter unit itself incorporated some sort of filter components?). I'd be interested to see circuits that did that - the great thing about audio is that there is always something new to learn.

[G6Tanuki]

Can't remember where I saw it, but the tweeter-direct-to-the-anode is definitely something I've seen - at the time I thought to myself "cor, that's ingenious!".

unless the tweeter unit itself incorporated some sort of filter components

The mechanics of the tweeter air interface does that.

[mhennessy]

I agree that the acoustic output of the tweeter will fall at low frequencies - it only has a small diaphragm.

But what stops the movement of the diaphragm at lower frequencies?

Obviously there will be a limit imposed by the stiffness of the diaphragm and any suspension - just like a conventional loudspeaker - but once you're down at these frequencies, there is a real risk of distortion or even damage from over-excursion. And quite apart from that, it will no-doubt sound horrible at those lower frequencies - especially if the primary resonance has a high Q.

I've been trawling the sheets on Paul's DVD, and have yet to find an example of a radio that uses an electrostatic crossover with no crossover. It's slow going - perhaps someone with more knowledge of sets from this era can help? Having identified a tweeter that can run without a crossover, it would be fascinating to look at how they've got around all the problems - and to ask if it might just have been much simpler to include a cap and resistor

[Radio Wrangler]

As electrostatic drivers look essentially capacitive, a simple series-C attempt at a crossover would act as a broadband attenuator.

There would need to be a significant resistive or inductive term to the driver to give any highpass characteristic.

David

[mhennessy]

Yes, I had a similar thought. But of course, the bias resistor takes care of that

The sets I've seen use bias resistors that sit in the 22 to 47k region, and series caps in the 470p to 5n area. Ignoring the capacitance of the tweeter - which I guess is fairly low, given the small size and reasonable spacing between diaphragm and plate - these all give quite reasonable crossover frequencies in their own right. But I'm painfully aware that I've only found a relatively small sample of all the sets that must have been designed.

There's some impressive DIY tweeters over on DIYAudio.com - for example: https://www.diyaudio.com/forums/plan...-midrange.html