Full range drivers with piezoelectric tweeters


Back in the late 50's while experimenting with optimization of my first system I tried using a RadioShack piezoelectric tweeter to enhance upper frequency reproduction. From my perspective today I'm sure I would find that setup marginal at best but the enhancement of upper frequencies was significant.

Since then, piezoelectrics haven't crossed my mind. However, over the years I've read and heard discussed the advantages of full range speakers based, I suppose, on the lack of a requirement for crossover circuitry. This makes sense to me and I decided that I want to experiment with this concept in a small setup in my garage. Based on information that I received during at least one discussion, though, it seems that not all "full range" drivers are very upper frequency capable.

This is where the question of piezoelectrics comes in. Based on my understanding that, by their nature, they cannot reproduce low frequencies (what these limits are I still have to determine), a crossover is not necessary which would give me more latitude in choosing a full range, or woofer in this case, driver.

For the purpose of this experiment I have an old but decent pair of bookshelf speakers that that I would remove the tweeters and crossovers from and parallel connect the piezos across the input terminals. This is the way I connected it in my original system over 50 years ago.

The concerns that I have are related to impedance issues and the division of frequencies between the two. In effect, the woofer would always be reproducing all frequencies that it is capable of and the piezo would always be self limited to frequencies above a certain cutoff. So there would be, indeed, a potential gap or overlap with no slope control.

I'm not asking for a detailed analysis or driver recommendations; I just want some idea of whether it is conceptually viable and if this approach is ultimately inconsistent with the goal is the full range speaker idea.
128x128broadstone
That makes sense, Timlub. My next project is to find a decent 6.5 inch full range driver and I'll use your advice in choosing a proper piezoelectric transducer. I'll let you know how it goes.
Broadstone, replacing a ribbon is ideal, one of the very few drivers that have a flat consistent impedance curve... so when you put the right resistor on the piezo, it will drop right in and work with the current crossover
Timlub, thanks for the explanation. With that in mind, I'm going to pursue this with a little more confidence that it can be made to work in the situation I described. I'm pretty familiar with crossover design and building so if I'm not able to get results that I'm happy with, it'll be fairly easy to take another direction even if it would be to reinstall the ribbons with the new woofers. It's way too easy and inexpensive not to try it and if that's still not the answer, I'll just slap together something else.
Hi Broadstone,
Keep in mind that I found consistent results with the Motorola & CTS Piezo's. We tried several, some had very low sensitivity, others had very cheap elements with sensitivity as low as 87... The Motorola & CTS are dual element designs and are fairly consistent. When you take these apart, you'll see a small separated area with a small resister in parallel, if I recall directly, it was 30 ohms. We tested these with and without and it just sounded better without it. We liked it best with a 10 ohm resistor and no crossover, but with a crossover, they will cross accordingly with whatever resistor you use on them.
Schubert wrote: "Most Gallo speakers do this and are among best on market."

This is true. However, Gallo speakers were specifically engineered to have precisely matching crossover slopes. That's one of the main reasons I chose all Gallo Reference speakers for my surround sound setup.

However, you cannot simply slap some drivers together without a crossover and expect to get any semblance of decent sound. A lot of care and engineering is used when choosing speaker components for a speaker system. If you take shortcuts the sound will suffer, quite dramatically in most cases...

-RW-
Yes, Elevick, if I were to use ribbons, a crossover would be necessary; in my experience they don't hold up to delivery of low frequencies except at Low volumes. As a matter of coincidence, the tweeters installed in these enclosures are EMIT ribbons. The use of piezoelectrics, though, are a somewhat different story assuming that my experience with them is representative of their character across the board. Back in those days I was a teenager and used to listen at fairly high volumes and never destroyed one.

Your point, though, concerning the volume difference reminds me that when I used a piezo, it seemed a bit overpowering. If I were to attenuate anything, then, it would be the tweeters and maybe adjustable resistance to them would be in order.

The impedance issue that you mentioned remains a concern but to answer your question, the amp that I'd be using, although of mediocre quality, is very high current capable.

Tim, your answer is exactly what I wanted to hear. At this point I have neither driver that I would use and the ones currently installed will be removed along with the adjustable X-over. So I'll be starting from the ground except for the enclosures which are the old Phase Tech PC 60's. Your description of the mods is very interesting and reassuring especially in light of my original experience where I found the them to be effective but a little "scratchy" sounding. The physical attenuation you describe, if I understand it, would not interfere with those frequencies in the audible range as long as the amount of added material was controlled to avoid too much attenuation. I assume that this determination was based on some amount of trial and error. BTW, I was surprised to see that the roll off was as low as 1800 Hz. Also, when you put them back together, did you replace the resistor and, if not, why was it removed?
to more directly answer your question rather than ask questions, here are some basics.
Assuming that you piezo's are old, hopefully they are motorolas...
Most common, a circular load in square mount about a 3.5 inch, these have a natural rolloff of about 3500.Sensitivity around 92. Even though we made them sound better, we never got great results from them...
Round 3" mount expedential horn, Natural rolloff about 3200 hz. sensitivity around 91. With the right mods, these became fairly good fast like a ribbon, would get edgy without a lot of work.
2x5 never liked em, so didn't experiment
2x6... overall, we got the best results from these... took them apart, put a coat of latex damping on the cone, removed the resistor, filled the internal cavity. Put them back together, wrapped them in butyl tape, added appropriate resistor and they were surprisingly good. These had a natural rolloff of about 1800hz. Sensitivity around 94.
An amplifier see's these as a capacitive load, do not hook these up and test them alone, use a resistor if possible. You could pop an output if you aren't careful.
I hope this helps,
Tim
Back in my old days, we did a ton of work with piezo's... some are crap others pretty good, Most people think they are terrible, but I've gotten some great results with some old Motorola's or the current CTS brands....I haven't found any others that I would use. to get outstanding results, it took a lot of work... What is the sensitivity, impedance and crossover point of the tweeter that you want to replace?
It's way harder than you think. I tried it with a mint pair of Pioneer HPM ribbon tweeters with my Merrill drivers. As it turned out, after going through a crossover for the tweeters, the output level was so much lower than the main driver that it wasn't worth it.
I wasn't about to attenuate the Merrills to fit the tweeters. The crossover is necessary not only to protect the tweeters but also to minimize any impedance irregularities.
Be careful to check what your impedance will drop to-can you amp handle it? Also, be careful with connections. A loose wire could damage the amp.