Piezo XO
Hi
I want to build a pair of very inexpensive loudspeakers. I will be using a piezo tweeter with a crossover like the one posted by Carlosfm, but I need to obtain 15 dB of attenuation (piezo tweeter driven by 1W outputs 105 dB, woofer 90 dB). Will the crossover still work if I increase the value of the series resistor to say 330 Ohm?
Hi
I want to build a pair of very inexpensive loudspeakers. I will be using a piezo tweeter with a crossover like the one posted by Carlosfm, but I need to obtain 15 dB of attenuation (piezo tweeter driven by 1W outputs 105 dB, woofer 90 dB). Will the crossover still work if I increase the value of the series resistor to say 330 Ohm?
Re: Piezo XO
Dropping the value of the parallel resistor to say 8 ohms (no lower) will help a lot( needs to be high wattage, parallel a couple of 16ohm 10w), need to raise the cap value though, depending on x-o point aimed at. So I would use 8 ohms across, say 50 ohms parallel, around 3 to 6uF cap.. buy a few electros and try different values.
btw, I actually drew that circuit diagram Carlo is using !!!
Athon said:
Dropping the value of the parallel resistor to say 8 ohms (no lower) will help a lot( needs to be high wattage, parallel a couple of 16ohm 10w), need to raise the cap value though, depending on x-o point aimed at. So I would use 8 ohms across, say 50 ohms parallel, around 3 to 6uF cap.. buy a few electros and try different values.
btw, I actually drew that circuit diagram Carlo is using !!!
Andy,
sorry for not giving you credit for that crossover circuit and thank you for the information.
Lacking the required components (for the moment) I did a simulation of the circuit using Microcap 8. The tweeter is modelled as a 0.12 uF capacitor. I was originally thinking of crossing at around 2.5 k (because there would be a lot of attenuation on the tweeter and the woofer is quite large - 7 inch, and I do not need too much power - 30 WPC will be sufficient). However people using piezos keep telling me that it does horrible stuff in the midrange so i'll set the crossover point higher. Using a 3 uF cap, 50 Ohm series and 8 Ohm parallel resistors results in a crossover point around 5 kHz, which is good, but there's only a slight attenuation (2.6 dB). Is there any solution for a 15 dB attenuation?... besides biamping? L-Pad? But for what impedance?
sorry for not giving you credit for that crossover circuit and thank you for the information.
Lacking the required components (for the moment) I did a simulation of the circuit using Microcap 8. The tweeter is modelled as a 0.12 uF capacitor. I was originally thinking of crossing at around 2.5 k (because there would be a lot of attenuation on the tweeter and the woofer is quite large - 7 inch, and I do not need too much power - 30 WPC will be sufficient). However people using piezos keep telling me that it does horrible stuff in the midrange so i'll set the crossover point higher. Using a 3 uF cap, 50 Ohm series and 8 Ohm parallel resistors results in a crossover point around 5 kHz, which is good, but there's only a slight attenuation (2.6 dB). Is there any solution for a 15 dB attenuation?... besides biamping? L-Pad? But for what impedance?
If you shunt the piezo horn with an 8 ohm resistor you should be able to design a conventional 2nd or 3rd order x-over using inductors and capacitors. This would allow you to precisely control the cross-over point and slope. It would be a good idea to include the piezo capacitance in the model. (shunt across that 8 ohm resistor.) This approach will also allow you to use an 8 ohm LPAD if you need to attenuate the tweeter output to match your woofer.
THANK YOU KEVIN!
Using 8 Ohms across I can indeed control slope and crossover point. As a plus the parallel resistors from the L-pad and the 8 Ohms across can be reduced to a single equivalent parallel resistor, reducing the parts count. I will be doing tests on the crossover next week but I'm pretty confident it will work properly.
Using 8 Ohms across I can indeed control slope and crossover point. As a plus the parallel resistors from the L-pad and the 8 Ohms across can be reduced to a single equivalent parallel resistor, reducing the parts count. I will be doing tests on the crossover next week but I'm pretty confident it will work properly.
This is a bit of an old thread, but here goes. The main advantage of piezo tweeters is their rising impedance at low frequencies, making it possible to use them without any phase-lag or phase-lead inducing crossover components. I'm looking to build a low wattage, high efficiency system, and a piezo tweeter simplifies my design. My main concern is the difficulty in finding one with anywhere near a flat response to 20KHz.
Piezo tweeters are not renowned for a flat response to 20k. Incidentally, the last time I messed around with them I followed advice and put an 8 ohm resistor across the terminals. Then I could use an 8 ohm variable Lpad, (and also any standard crossover design), because the tweeters impedance is so high that the amplifier sees only about 8 ohm.
This is the hot tip with piezos. A series + parallel resistor smooths it out even more, and also provides necessary attenuation.
You can scale the resistors up, but not too much. Then you have to adjust the crossover accordingly. Consider that the impedance of the tweeter is quite variable with frequency. Nominal impedance is typically around 120 ohms.
An Lpad, variable or not, is already a series resistor and a parallel resistor, and the tip of using an additional parallel resistor after the Lpad has the advantage that one is working with a known 8 ohm load, rather than one that is "variable with frequency".
Exactly.
You can dispense with the L-pad and just use series and parallel resistors if you are confident in your design. L-pad + 8 ohm resistor in parallel with the piezo tweeter works great though, and will give correct rolloff with a basically off the shelf crossover.
It's not hard to understand why this smooths out the response. And many "cookbook" speaker builders probably don't understand that piezo tweeters simply don't work with off the shelf crossovers without adjustments. You could say the same about a lot of conventional drivers too, to a lesser degree perhaps.
Well, all this is useful and interesting. I'm looking to marry a KSN1016A with 2 x 6" full range speakers in what will probably be a TQWP cabinet. I see there is no getting away from the need of a proper crossover, and the hints on taming the peaks in the piezo response will be considered. Let you all know how it turns out, if we live long enough ...
KSN1176A and KSN1036A from PiezoSource Piezoelectric Speakers might be the right candidates:
http://69.64.35.188/piezo/files/KSN-1177A-Datasheet.pdf
http://69.64.35.188/piezo/files/KSN-1034A-1036A-1036B-1039A-Datasheet.pdf
Graphs seems to be heavily smoothed, though.
Last edited:
Oh come on. If someone wanted to do it piezos would be the ideal supertweeter. The truth is a lot of money was made by mass producing them quickly and cheaply, so they are overlooked. They are about half designed and still have enormous potential for supertweeter use probably using Danley coupling tech, yes HORN loading multiples.
Motorola made some fine tweeters, the "replacements" not so much.
I took a Toa tweeter horn apart and they glued the paper of a piezo down like a speaker and it made a huge difference. Most likely due to the woofer not shaking the whole tweeter cone. I did that to a pair of older motorola's and they did sound much clearer but it was extremely hard since the wire is so fine soldering is almost impossible.
On the PA monitor speakers I make I use a 6 ohm 3 watt resistor in line with an 1157 automotive light bulb. I solder across both terminals, then use a capacitor for my crossover frequency (usually aim for 3k-5k depending on woofer since 2k is a squeal frequency in live music) People love them and they don't burn out.
Never had any luck with the 8 ohm resistor, they always broke no matter how I mounted them or burned up. Nothing is harder on gear then musicians.
I took a Toa tweeter horn apart and they glued the paper of a piezo down like a speaker and it made a huge difference. Most likely due to the woofer not shaking the whole tweeter cone. I did that to a pair of older motorola's and they did sound much clearer but it was extremely hard since the wire is so fine soldering is almost impossible.
On the PA monitor speakers I make I use a 6 ohm 3 watt resistor in line with an 1157 automotive light bulb. I solder across both terminals, then use a capacitor for my crossover frequency (usually aim for 3k-5k depending on woofer since 2k is a squeal frequency in live music) People love them and they don't burn out.
Never had any luck with the 8 ohm resistor, they always broke no matter how I mounted them or burned up. Nothing is harder on gear then musicians.
Active crossover on piezos?
Hmm. You may need a series resistor unless using powerline Motorola or CTS units as piezos exhibit falling impedance with frequency. Powerline have series impedance in the form of a light bulb and PTC already.
Don't use anything except genuine Motorola or CTS units. Everything else is garbage.
Hmm. You may need a series resistor unless using powerline Motorola or CTS units as piezos exhibit falling impedance with frequency. Powerline have series impedance in the form of a light bulb and PTC already.
Don't use anything except genuine Motorola or CTS units. Everything else is garbage.
KSN-1188 with active crossover
The beauty of using active crossovers is that the speakers can be coupled directly to the output of an amplifier, if the amplifier input is tailored to the driver response. The KSN-1188 go down to 500 Hz, but looking at the graph response I would not push them below 2 KHz.
There seems to be some consensus that wiring a high-wattage 8 ohm resistor in parallel with the drivers will smooth their response.
The beauty of using active crossovers is that the speakers can be coupled directly to the output of an amplifier, if the amplifier input is tailored to the driver response. The KSN-1188 go down to 500 Hz, but looking at the graph response I would not push them below 2 KHz.
There seems to be some consensus that wiring a high-wattage 8 ohm resistor in parallel with the drivers will smooth their response.
Attachments
I got these:Active crossover on piezos?
Hmm. You may need a series resistor unless using powerline Motorola or CTS units as piezos exhibit falling impedance with frequency. Powerline have series impedance in the form of a light bulb and PTC already.
Don't use anything except genuine Motorola or CTS units. Everything else is garbage.
GRS PZ1188 Piezo Horn Driver Similar to KSN1188A
GRS claim these are the exact replica and specs from the CTS but with a larger ceramic disc. I open it up and saw some sort of small cap or resistor soldered inside...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.