How do I read this transformer datasheet?

[keith1200rs]

Xyz is most likely the volume of the magnetic material - it should be in the data sheet. I cannot check at the moment.

Keith

 

[ants]

If I read the datasheet correctly the volume Ve = 1050mm³, I have just measured the core, it is hard to estimate volume of magnetic material with a ruler but it looks less than 1050mm³, actually more like 650mm³. At a guess I'd say the volume is 1050mm³ including the space inside the core for the windings.

Thanks,

Ant.

 

[keith1200rs]

Are you including both halves of the core in you measurements? The figure given is likely to be the combined value for two mated halves.

Keith

 

[ants]

Yes you are right, it says Magnetic characteristics per set. I will start looking for a more appropriate core now and rebuild with more windings. I'll be putting 60Vpp on the primary and 180Vpp on the secondary. I could use the 0.2mm wire I have or switch to an RM4 that I already have, it is 2-3 times bigger and might do the job. The RM6 I have is a bit on the small size. I can rethink tomorrow. Thanks.

Ant.

 

[ants]

I built the transformer with an RM6 ungapped core although I now have some EFD15 gapped cores for later use. It isn't the best but it does work in that I put 100Vpp across the piezo.

The problem I have is more fundamental. I designed the circuit for a piezo of 41KHz and <100 ohms. I didn't actually have that piezo till last Friday, and it is actually 41KHz and ~1.2K ohms. So I'm getting a trifling amout of power out of my circuit onto the piezo, probably 0.2W.

The problem is the piezo, in that it just doesn't work as I'd hoped. It has 2 layers of pzt glued together and I need a piezo with many more layers of pzt, probably 10 or more.

I'd say this comes back to a comment by FvM earlier, in that the circuit has to be designed for the piezo and this particular piezo needs many more volts and current than I can supply from my own personal built circuit.

Some thoughts / questions.

If I remember correctly the current draw of a piezo is proportional to the number of layers. So if I go from 2 to 20 layers it is similar electrically from going from 2 to 20 capacitors. Each needs to be 'driven' so the current and thus power goes up for the same volts. Is that right?

I'd like to get some value out of these piezo's by driving them with an audio amp that I have set up with a mains power supply. The audio amp needs a load of 4 or 8 ohms. I'm not sure exactly how to build the transformer for 4 ohms impedance. Do I calculate what the inductance needs to be for 4 ohms at 41KHz, where does the reflected load impedance of 1.2K ohms come into this? The piezo capacitance is 1.5nF.

Thanks,

Ant.

 

[FvM]

I didn't actually have that piezo till last Friday, and it is actually 41KHz and ~1.2K ohms.

If the capacitance is also 1.5 nF, the resonance Q is unusually low.

Using a higher power audio isn't a bad idea. You should be sure, however, that it's power bandwidth is above 41 kHz and that the output stage isn't two sensitive to inductive loads. An output series resistance as a precaution may be reasonable.

 

[ants]

If the capacitance is also 1.5 nF, the resonance Q is unusually low.

I'm not sure how to calculate resonance Q but I have a few and they all need tuning fairly accurately. It is possible their impedance is lower, when I tested it I might not have been putting enough volts across them to ring/resonate, the transofmer I built doesn't seem to be doing enough it is 1:3. They are bimorphs and are supposed to have fairly large displacement and shouldn't go over 100Vpp.

Using a higher power audio isn't a bad idea. You should be sure, however, that it's power bandwidth is above 41 kHz and that the output stage isn't two sensitive to inductive loads. An output series resistance as a precaution may be reasonable

I have just been testing with an LM3886, it is fairly decent with a good GBP and slew rate. I have gone back to using the transformer I intended to replace at the start of this thread it is 1:19. I have put a series resistor to protect the amp. With this set up I have put a few hundred volts across the piezos and I do get noticeable displacement but its not a practical working voltage.

I'm thinking I should make another transformer at a wiring ratio of 1:6 and use the LM3886 to put 300V across them to complete testing. With a powerful audio amp I probably don't have to worry too much about the accuracy of transformer.

Thanks,

Ant.

 

[keith1200rs]

What make/part number are the piezos? As FvM said, that is a very low Q. Be careful about putting too much voltage across them. I have vaporised a few trying get too much power out of them.

Keith

 

[ants]

I had them made for me, so they don't have a datasheet but I have plenty of them, they are 2 pzt rings glued together and I need to glue them to a metal diaphragm myself, a bit like the buzzers you find commonly but I haven't done that yet. I don't have much experience of making piezos so its not a surprise these don't work how I want, they do vibrate without being on the diaphragm so I will get a bit more displacement when they are on the metal. But still I don't think it'll be enough vibration. I geared the circuit up for about 100Vpp and 2 amps not 300Vpp and xyz amps. The Q is ~ 20, the pzt is a soft material that isn't so common, it vibrates more but also has to be tuned more finely. I'm not sure what the next step is now.


Thanks,

Ant.

 

[keith1200rs]

What is the material? Obviously not PZT4. When you bond them to a metal disc you will have some new resonances to be concerned with. If the piezos are very thin then they won't take a lot of voltage - around 200V/mm for PZT4 I think, but you material will be different.

Most of my high power ultrasonic work have been with 3mm thick PZT4, 50mm diameter although I have used 2mm which as well as 25mm discs. 50W into a 50mm disc bonded to a metal plate will dance around the room at resonance so if you are planning on 100W you will have some serious motional energy if you get it right.

Mind you, it depends on what you are doing as to what frequency works best. For my application 80kHz did the job better than 40kHz even though at 40kHz the physical effects looked far more vigorous. I think it depends on the properties of the material or liquid you are trying to effect.

Keith

 

[ants]

The material is pzt 5H, it can be easily squeezed by hand and the capacitance/voltage changes but I'm uncertain of its full potential. The target frequency is 100KHz, so the ones I have at 40KHz I already knew weren't exactly what I needed, they are more of a stepping stone to see if I can get a decent amount of ultrasonic motion. It is essentially an electro acoustic disc, the difference is that buzzers have one layer of pzt and typically brass forms the other layer. With this setup it is 2 layers of pzt that vibrate, I'm not sure how much the metal will alter the resonant frequency. I think these types of piezo are used in mobile phones commonly as a speaker, but I'm not completely sure.

I'm building a handheld/portable nebuliser so I'm looking to shake the liquid as much as possible. I think I need to get more current to the pzt and try to find out the impedance much more accurately.

Thanks,

Ant.

---------- Post added at 14:30 ---------- Previous post was at 14:22 ----------

I mentioned more layers of pzt because the displacement is additive, in that 20 layers produces ten times the displacement of 2 layers (in theory). I couldn't even hazard a guess of the impedance of a multi layer device. I was hoping I wouldn't need them.

As a matter of interest how did you source your pzt?

 

[keith1200rs]

Initially I bought piezos from Morgan Electroceramics and then source some from Taiwan. I did get involved in a couple of nebulizer projects as well. Is it a clinical or simply therapeutic one?

I might have to continue this discussion privately.

Keith

 

[ants]

I spoke to Morgan a while back but haven't ordered from them. I have not long ago glued the pzt to the metal, but there isn't much displacement, I will have to make more for test but I think I'm going to need a multi layer piezo. I will order some off the shelf parts from Morgan or Noliac and see where they are at. Its a therapeutic one, I have no problem with a private message.

Thanks,

Ant.

 

[keith1200rs]

When I am back in work tomorrow I will see what I have. If it is public domain information I will post it here. I have a complete nebulizer circuit and may even have one or two assembled circuits with piezo so I could send you one. I just need to check who owns the intellectual property,

Keith

 

[ants]

Hi Keith,

That'd be great, thanks!

Ultimately my intention is to build the prototype, take a patent and pull in funding to set up a company. I am attempting to copy the technology of a previous employer, unfortunately while I worked with their atomiser I didn't build it and it was confidential so I didn't have it for long. When I look ahead at how daunting all of the technical tasks are and how long it will take I can be quite fazed. I've spent 18 months part time already and it will last that long again I'd say. I wouldn't invest that amount of time if there weren't potential rewards, their atomiser is great, but failure is very probable. Part one of the project was container building, part 2 the electronics, part 3 the vibrating element and part 4 making the vibrating element atomise. Probably the closest commercial product to what I want to make is the Glade Wisp or its derivative the Glade Candle. It is available in most supermarkets and uses an electro acoustic disc to atomise small amounts of air freshener. I couldn't hope to get close to the electronics of that, ultimately I need something 'good enough', a container which is ok, electronics which work, atomiser that works, optimising it would take more time again.

Thanks for the help.

Ant.