hello, i am planning to replace my horn tweeter with a piezo one, i search through the internet on how they are working. until i saw the information at the Parts - Express that i need o put some 30 ohm resistor, and some info that this piezo doesn't actually need a crossover...
my question is, if i have a crossover and i want to hook my piezo in it. do i still need a resistor and a caps for that? how to do that? how can i hook up a resistor on my piezo is it in parallel or in series?
i am planning to put that piezo on the HF sign in the crossover
sorry for my bad english, any comments and suggestions are highly appreciated.!
my question is, if i have a crossover and i want to hook my piezo in it. do i still need a resistor and a caps for that? how to do that? how can i hook up a resistor on my piezo is it in parallel or in series?
i am planning to put that piezo on the HF sign in the crossover
sorry for my bad english, any comments and suggestions are highly appreciated.!
Technically, a piezo element doesn't require a crossover to function properly, as it's already a capacitive load. However, if you want it to SOUND GOOD, then yes, you NEED a crossover.
If you wire a piezo driver into an existing passive crossover, you want to imitate the impedance load that the crossover expects to see by wiring a resistor across the leads of the piezo (parallel). If the driver you're replacing was 8 Ohms, then an 8 Ohm resistor needs to replace it, etc. It isn't ideal, but it'll work in a pinch. Also, a resistance value that low will attenuate a piezo, so you might have a problem with matching levels between woofer & tweeter.
Perhaps someone with more piezo experience can chime in...?
If you wire a piezo driver into an existing passive crossover, you want to imitate the impedance load that the crossover expects to see by wiring a resistor across the leads of the piezo (parallel). If the driver you're replacing was 8 Ohms, then an 8 Ohm resistor needs to replace it, etc. It isn't ideal, but it'll work in a pinch. Also, a resistance value that low will attenuate a piezo, so you might have a problem with matching levels between woofer & tweeter.
Perhaps someone with more piezo experience can chime in...?
The piezo horn by Motorola is a capacitive device. The reason most manufacturers place a series resistor and no crossover is to limit the power, allowing for a larger speaker system to avoid cooking the piezo and they do not require a crossover due to the characteristics.
Some more expensive piezo elements have a light bulb in series with them to give a better non linear output that matches the human ear better by being low resistance when cold and high resistance when driven.
So, no crossover!
Oh yes, I forgot to add, they sound horrible!!
Some more expensive piezo elements have a light bulb in series with them to give a better non linear output that matches the human ear better by being low resistance when cold and high resistance when driven.
So, no crossover!
Oh yes, I forgot to add, they sound horrible!!
10 ohms in parallel (close enough to 8) with the piezo, on the existing crossover's output. However, regular horn tweeters tend to have more output than piezos. If resistive padding was used on the original horn, you may need to bypass it or reduce its effect, connecting directly to the high pass filter output.
If it's too complicated, just use 4.7uF in series and 10 ohms in parallel, which is what the original Motorola/CTS application notes suggested. If your woofer is in the neighborhood of 94dB/W, the levels will match close enough, unless you're using one of those super-output units with a built in transformer. Those don't need the resistor in parallel either, and do *require* a 12dB or better crossover.
If it's too complicated, just use 4.7uF in series and 10 ohms in parallel, which is what the original Motorola/CTS application notes suggested. If your woofer is in the neighborhood of 94dB/W, the levels will match close enough, unless you're using one of those super-output units with a built in transformer. Those don't need the resistor in parallel either, and do *require* a 12dB or better crossover.
The obvious question is why? there is no scenario where the piezo will perform(sound) better so why aren't you just installing a new diaphram in the existing driver or installing a different CD?
I have the original paper by the MOTOROLA guy who produced the current crop of piezo published in The JAES some 40 yrs ago, Borst (or Brost?) I think. It is very helpful. Posters are correct in saying that piezo is capacitive but that changes the traditional role of cross-over elements. The R in series does not attenuate the overall level but instead produces a high frequency roll off. On the other hand small capacitors in series do NOT reduce low frequencies but DO reduce overall levels. So use small caps as attenuators i.e. as you would a resistor in normal circumstances.
Again previous posters are correct in suggesting you bridge the unit with a resistor. That way you can use the old cross-over. At the frequencies being considered the R dominates the capacitive impedance.
There are some good sites that will fill out the details of using these units. And they do not have to sound bad. Arranging for a cross-over that does attenuate frequencies below about 5kHz is useful.
I may also have seen work by a guy called Smith that demonstrates much improved impulse response from horns when used with appropriate cross-overs. (can't quote the ref' on that but it was also JAES around 40 yrs ago or more!) That is why people do use a cross-over with these units despite its theoretical redundancy, see #2. Some people press car repair "bog" around the back of the chassis to reduce vibrations. At one stage Proac(!) even used a piezo in one of their domestic models but it was the direct radiator version of the popular horn.
If you don't like the sound take the piezo element out of the horn and use it in the ELECTOR design as the sender unit for their bass speaker motional feedback circuit ha ha! (That was around 1987 here in Australia)
Cheers Jonathan
Again previous posters are correct in suggesting you bridge the unit with a resistor. That way you can use the old cross-over. At the frequencies being considered the R dominates the capacitive impedance.
There are some good sites that will fill out the details of using these units. And they do not have to sound bad. Arranging for a cross-over that does attenuate frequencies below about 5kHz is useful.
I may also have seen work by a guy called Smith that demonstrates much improved impulse response from horns when used with appropriate cross-overs. (can't quote the ref' on that but it was also JAES around 40 yrs ago or more!) That is why people do use a cross-over with these units despite its theoretical redundancy, see #2. Some people press car repair "bog" around the back of the chassis to reduce vibrations. At one stage Proac(!) even used a piezo in one of their domestic models but it was the direct radiator version of the popular horn.
If you don't like the sound take the piezo element out of the horn and use it in the ELECTOR design as the sender unit for their bass speaker motional feedback circuit ha ha! (That was around 1987 here in Australia)
Cheers Jonathan
Last edited:
I presume it's because the driver keeps burning out. Piezos with crossovers are pretty tough. You'll blow the caps or toast the resistor first, typically. The real solution is a driver with a 2" coil, and they start around $50. In a speaker with a $25 woofer and a particle board cab, it may not be worth it. The piezo is cheaper for sure....
Jonathan R. Bost, Motorola. I used to converse with him about product improvement.
This is the better 1.8Khz piezo.
https://docs.google.com/viewer?url=patentimages.storage.googleapis.com/pdfs/US4078160.pdf
Note that it has an internal 30Ω/2W resistor in series, for the smaller 4Khz models an external resistor should be used, about 50Ω/2W (or 5W).
This is the better 1.8Khz piezo.
https://docs.google.com/viewer?url=patentimages.storage.googleapis.com/pdfs/US4078160.pdf
Note that it has an internal 30Ω/2W resistor in series, for the smaller 4Khz models an external resistor should be used, about 50Ω/2W (or 5W).
Last edited:
Interesting. So you've got series resistors to keep HF impedance out the gutter, an RL network makes the whole thing appear resistive, and then throw a 3rd order XO at it. Bulb in series to avoid lots of continuous power.
When used in that way, any idea of what sort of power handling can be expected from each piezo?
Chris
When used in that way, any idea of what sort of power handling can be expected from each piezo?
Chris
"When used in that way, any idea of what sort of power handling can be expected from each piezo?"
The longest I power tested the 0.13µF bender elements (4Khz variety) was with 240V P-P on pink noise for about 24 hours straight. Output was unchanged.
It is possible to slag them, but it takes some work (assuming you are using them in a proper fashion).
The longest I power tested the 0.13µF bender elements (4Khz variety) was with 240V P-P on pink noise for about 24 hours straight. Output was unchanged.
It is possible to slag them, but it takes some work (assuming you are using them in a proper fashion).
Just to compare/contrast with DJK's array, the BF method would have 2 banks of 5 piezos, each bank wired in parallel, then the 2 banks wired in series. Any resistor sets would be in the Xover, separate from the array. I think his Xovers are only 2nd Order? But again, plenty of power handling, not to mention a decent impulse response.
Last edited:
"most who have actually heard these ceramic piezo tweeters recommend a air gap in the wire of the XO to them for the best sound"
Uncalled for.
The original Motorola devices were fantastic (if you knew how to use them, most did not).
The current crop of off-shore cheapies have issues. I bought a bunch of units from Parts Express, and selected them. I ended up throwing away 25% of the units, and if I had ordered a bigger batch to start with I would have rejected even more.
"Any resistor sets would be in the Xover, separate from the array"
The tweeter is better off with individual series resistors for parallel connection, this tends to de-Q the bender element resonance better than one lumped resistor in the network.
Uncalled for.
The original Motorola devices were fantastic (if you knew how to use them, most did not).
The current crop of off-shore cheapies have issues. I bought a bunch of units from Parts Express, and selected them. I ended up throwing away 25% of the units, and if I had ordered a bigger batch to start with I would have rejected even more.
"Any resistor sets would be in the Xover, separate from the array"
The tweeter is better off with individual series resistors for parallel connection, this tends to de-Q the bender element resonance better than one lumped resistor in the network.
Note the use of a single Motorola KSN 1016A in front of the lower HF horn.
These RCA are 16Ω and about 110dB/W/1M, with the HF rolling off starting about 4Khz. A matching transformer was used on the tweeter with three 220Ω resistors in parallel driving the tweeter with an 18dB network at 4Khz. The response measured ±3dB from about 100hz~20Khz (at about 6' in a 30 x 50 garage). One tweeter is adequate for home use, I usually used three to five units for PA.
The RCA now reside in a community theater in Topeka, KS (originally from the Paramount Theater, Cedar Rapids, IA).
These RCA are 16Ω and about 110dB/W/1M, with the HF rolling off starting about 4Khz. A matching transformer was used on the tweeter with three 220Ω resistors in parallel driving the tweeter with an 18dB network at 4Khz. The response measured ±3dB from about 100hz~20Khz (at about 6' in a 30 x 50 garage). One tweeter is adequate for home use, I usually used three to five units for PA.
The RCA now reside in a community theater in Topeka, KS (originally from the Paramount Theater, Cedar Rapids, IA).
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.