as each piezo approximates a 0.13uF capacitor and that five in parallel are not loud enough to keep up with a 99dB rated fullrange 12" much less my horn systems,the transformer needs to be wired to step up voltage.
Each piezo has 56 ohms in series to limit how low the total impedance may dip.
Here's the impedance of the five
KSN1016 in parallel with no transformer and no Zobel network.
When I wired the transformer with the 25 volt winding (red and white leads) as the primary - I got a step down transformer action. I guess I could try the feed to the piezo array.15 watt tap as primary and 25 volt leads to the piezo tweeter stack.
In the impedance sweep - you can see deviation from ideal capacitor behavior - I assume that's due to mechanical resonance and represents problem area issues - peaks - then a dip up higher.
A single KSN1016 (original Motorola) piezo on the 5 watt rap of a 25 volt Edcor transformer
gave ~103dB output at 5KHz at 2.8v input to the transformer. HD was around 1/4 of one percent. I've lost those transfomers.
Each piezo has 56 ohms in series to limit how low the total impedance may dip.
Here's the impedance of the five
KSN1016 in parallel with no transformer and no Zobel network.
When I wired the transformer with the 25 volt winding (red and white leads) as the primary - I got a step down transformer action. I guess I could try the feed to the piezo array.15 watt tap as primary and 25 volt leads to the piezo tweeter stack.
In the impedance sweep - you can see deviation from ideal capacitor behavior - I assume that's due to mechanical resonance and represents problem area issues - peaks - then a dip up higher.
A single KSN1016 (original Motorola) piezo on the 5 watt rap of a 25 volt Edcor transformer
gave ~103dB output at 5KHz at 2.8v input to the transformer. HD was around 1/4 of one percent. I've lost those transfomers.
Attachments
Last edited:
Stepping down voltage is what this kind of transformer is designed to do when connected normally.
If you connect it backwards, it will step up voltage, but with a capacitve load you may wind up with high frequency limitations (worse with higher step-up ratios). I'm not familiar enough with them to say whether it's an issue. Most of my transformer experience is with electrostatics.
Try red/white (25V) on the piezos and purple/gray (30 W) to the amp. If I'm thinking correctly, that should give you the least step up.
If you connect it backwards, it will step up voltage, but with a capacitve load you may wind up with high frequency limitations (worse with higher step-up ratios). I'm not familiar enough with them to say whether it's an issue. Most of my transformer experience is with electrostatics.
Try red/white (25V) on the piezos and purple/gray (30 W) to the amp. If I'm thinking correctly, that should give you the least step up.
Ran a few more numbers:
30 W and 25 V connections should give 1.62 turns ratio, so about 4 dB of voltage gain
15 W and 25 V, 2.29 turns ratio, about 7 dB voltage gain
Impedance transformation is turns ratio squared, so you can calculate that as well.
If you want to do everything as cleanly as possible, you may want to put your measured loaded transformer impedance data into a simulator. Transformers and passive crossovers can sometimes interact in strange ways. It's highly transformer and load specific though.
30 W and 25 V connections should give 1.62 turns ratio, so about 4 dB of voltage gain
15 W and 25 V, 2.29 turns ratio, about 7 dB voltage gain
Impedance transformation is turns ratio squared, so you can calculate that as well.
If you want to do everything as cleanly as possible, you may want to put your measured loaded transformer impedance data into a simulator. Transformers and passive crossovers can sometimes interact in strange ways. It's highly transformer and load specific though.
Last edited:
I have successfully used 2:1 (6dB) and 3:1 (10dB) step up transformers for very loud Piezo arrays but wound my own air core transformers.
They do double duty as part of a high pass crossover network.
Severely limiting lower frequencies hitting Piezos is good
I would never have thought using iron core EI transformers for that.
They do double duty as part of a high pass crossover network.
Severely limiting lower frequencies hitting Piezos is good
I would never have thought using iron core EI transformers for that.
That sounds like a better solution tho a decent iron (M6 core) could be OK. Is an air-core autoformer what's inside the Le Son piezo tweeter? I don't know how to make one.
I ended up running the tweeters with the 15 watt tap as the primary - sounds clean" but not "airy". The wideband is a Celestion K12H-200TC.
Z at 20KHz has dipped to 3.6 ohm using the 15 watt tap as the primary.
I'm running the piezo stack with 1.3uF to the primary (blue-gray)
What is a good modern piezo tweeter to buy for high power arrays?
.
I ended up running the tweeters with the 15 watt tap as the primary - sounds clean" but not "airy". The wideband is a Celestion K12H-200TC.
Z at 20KHz has dipped to 3.6 ohm using the 15 watt tap as the primary.
I'm running the piezo stack with 1.3uF to the primary (blue-gray)
What is a good modern piezo tweeter to buy for high power arrays?
.
Attachments
Last edited:
I like the sound of this array - no spit - good complexity- finger cymbals, hand cymbals, gongs - its great on solo acoustic guitars with various mics. I'm only about 4.5 feet away from the speaker. My other toys of interest are little horn rigs - Lordsansui's smallhorn3 and an 0.83 scale klipsch Belle. CD -are 1.4" Beyman and 1" DE250.
Here's the rest of the turns ratio info from my calculations, in case you want to play around some more:
Impedance data from here:
https://products.electrovoice.com/binary/5005, 5015, 5030, 5060 EDS.pdf
Calculations assume an 8 ohm speaker load in transformer's normal configuration.
| 70.7 V | watts | impedance | turns ratio |
| 30 | 167 | 4.57 | |
| 15 | 335 | 6.47 | |
| 10 | 500 | 7.91 | |
| 5 | 1000 | 11.18 | |
| 2.5 | 2000 | 15.81 | |
| 1.25 | 4000 | 22.36 | |
| 25V | 30 | 21 | 1.62 |
| 15 | 42 | 2.29 | |
| 10 | 63 | 2.81 | |
| 5 | 125 | 3.95 | |
| 2.5 | 250 | 5.59 | |
| 1.25 | 500 | 7.91 |
Impedance data from here:
https://products.electrovoice.com/binary/5005, 5015, 5030, 5060 EDS.pdf
Calculations assume an 8 ohm speaker load in transformer's normal configuration.
Last edited:
Please find the start of this Thread for answers to your question >
https://www.diyaudio.com/community/...tweeter-piezo-tweeter-revisited.399363/latest
https://www.diyaudio.com/community/...tweeter-piezo-tweeter-revisited.399363/latest
Last edited:
Assuming the array is around 96 dB
with all 5
Seems like a ratio around 1.5 to 1.8 would
be needed to get up to 99 / 100 dB
Looking at the great calculations done by mattstat
the 30watt or 15 watt ratios on the 25 watt tap
are gonna be the closest.
Far as modern piezo
Seems like GRS and PiezoSource
are the main producers other than maybe Goldwood.
They all basically make reproductions of CTS/Motorola
so nothing new. PiezoSource bought the CTS rights/patent
but seems to be whatever, since they sell 3 dollar speakers for
24 or more dollars.
GRS seems to really push they use true 22mm elements.
As others use smaller elements. Im no expert by far
just basically repeating their add copy.
Far as adding a little "air" or detail
with high crossover points around or above 5 to 8 kHz
Probably just move along to " Bullets" and only need
a single unit for 100 dB + systems.
Range from non piezo like typical titanium drivers
or phenolic
With good crossover / high order
JBL ST200 or Dayton ST603
Something I always wanted to play with
is Visaton HTH 8.7 as well
with all 5
Seems like a ratio around 1.5 to 1.8 would
be needed to get up to 99 / 100 dB
Looking at the great calculations done by mattstat
the 30watt or 15 watt ratios on the 25 watt tap
are gonna be the closest.
Far as modern piezo
Seems like GRS and PiezoSource
are the main producers other than maybe Goldwood.
They all basically make reproductions of CTS/Motorola
so nothing new. PiezoSource bought the CTS rights/patent
but seems to be whatever, since they sell 3 dollar speakers for
24 or more dollars.
GRS seems to really push they use true 22mm elements.
As others use smaller elements. Im no expert by far
just basically repeating their add copy.
Far as adding a little "air" or detail
with high crossover points around or above 5 to 8 kHz
Probably just move along to " Bullets" and only need
a single unit for 100 dB + systems.
Range from non piezo like typical titanium drivers
or phenolic
With good crossover / high order
JBL ST200 or Dayton ST603
Something I always wanted to play with
is Visaton HTH 8.7 as well
Seems reasonable.
If it has a ferrite core, those tend to have more issues when saturated, but given the likely cross point probably not an issue. You can always give it a test drive and replace it if you don't like it.
Again, ultimately it would be best to simulate the crossover to see what things look like, but tuning by ear is done as well.
If it has a ferrite core, those tend to have more issues when saturated, but given the likely cross point probably not an issue. You can always give it a test drive and replace it if you don't like it.
Again, ultimately it would be best to simulate the crossover to see what things look like, but tuning by ear is done as well.
I'm not sure if you're asking a new question or if I confused you with the toroidal inductor comment. That's what I was referring to in post 16 - that the toroidal inductor might have a ferrite core. But even if it does, it'll probably be OK given the use case here.
Given the power and frequency response ratings of the University Sound 5030, you should be far, far away from any core saturation issues with it since you're only using it to drive tweeters and you intend to use a passive crossover also.
Designing/spec'ing audio transformers involves a lot of compromises and specifications. Something off-the-shelf is typically preferable unless you have very specific needs.
Given the power and frequency response ratings of the University Sound 5030, you should be far, far away from any core saturation issues with it since you're only using it to drive tweeters and you intend to use a passive crossover also.
Designing/spec'ing audio transformers involves a lot of compromises and specifications. Something off-the-shelf is typically preferable unless you have very specific needs.
Yes, that's where I got the idea from.
It actually turns out much easier than using EI iron which in a general purpose transformer is needed to reach low, say 50 Hz ... what for?
We need nothing below 2kHz anyway.
Not sure about your particular needs, but personally very happy with the plain vanilla 3.5" one, the CTS/Motorola clone.
Very loud if driven hard above 3500 or 4000Hz, the 12dB/oct "free" crossover inherent to the step-up transformer avoids the nasty ugly 3 kHz peak which makes Piezos unpopular.