How to find an audio transformer, step up from 8ohm to 1Kohm

Status
Not open for further replies.
Re: How to find an audio transformer, step up from 8ohm to 1

Power ultrasonic drivers, as far as CW operation is intended (e.g. ultrasonic cleaning or "welding" of thermoplastics) are usually resonant circuits, simple ones often self oscillating designs. The transducer capacitance can be compensated by an inductor to reduce the reactant currents to be carried by the driver. If a series resonant circuit is used, a 1:10 voltage step up ratio can be possibly achieved without a transformer.

The discussed Apex and similar high-voltage power amplifiers are good for wideband and pulse applications, but not actually needed for a simple CW generator.

In any case, you should know the exact transducer specification. Apart from impedances and maximum electrical ratings, there are possibly also mechanical requirements, e.g. an initial force for safe operation. This is due to the fact, that the tensile strength of piezo material is limited and can be easily exceeded in resonant operation.
 

Re: How to find an audio transformer, step up from 8ohm to 1

For circuit idea look at the US Patent 4319716 or try to recycle discarded Piezoelectric Humidifier.
 

Re: How to find an audio transformer, step up from 8ohm to 1

Thanks for the replies!

I have explored building a circuit from scratch and I have bought parts to do so. I have altered the parts on one board I have, it is a humidifier circuit that resonates at 1.67MHz and got the resonant frequency down to around 100KHz.

I think [[signal generator ---> LM3886 ----> audio transformer ---> piezo]] is going to be the best route because i'm going to be building the piezo atomiser myself.

Essentially I'm going to be doing some home research on atomiser building. I am buying pzt at 1.67MHz then attaching it to a metal plate, then I'll be damping the frequency down to the required 100 KHz and reorder pzt and plates if necessary. So right now I can't know the power consumption, capacitance or impedance of the transducer/ piezo, I can only estimate and it is a little imprecise :)

The advantage of using the signal generator/LM3886 and audio transformer is that it will allow me to change frequencies without having to change components while I create the transducer/atomiser that I want. The next stage when I have the finished piezo is to build a complete circuit. This will then need to be optimised.

I have found a RadioShack audio transformer which does exactly what I want.

www.radioshack.com/product/index.jsp?productId=2103254

Unfortunately they don't ship to the UK where I live and both UK suppliers I found are out of stock. How or where could I get these parts?

Thanks for your suggestions.

Ant.
 

Re: How to find an audio transformer, step up from 8ohm to 1

If the said radio shack transformer actually has 1:10 windings ratio (unfortunately I don't see a specification at Radio Shack) it most likely doesn't work in your application because of it's expectable high leak inductance. But you should try yourself. Simply consider an about 10 to 20 kHz cut off frequency for usual cheap audio transformers at nominal load. So you can expect a leak inductance resonancy somewhere in the 10 kHz range.

It's generally no problem to design a transformer for 100 or 500 kHz. You most likely find a series ferrite transformer intended for switching mode power supplies with suitable parameters, if you don't know, how to make it yourself. (I can't exactly understand why, because I'm used to build transformers since my highschool days.) Coilcraft and other manufacturers have some of them.

I basically keep my opinion that a specially designed driver in a resonant circuit would be the best final design solution, but I consider, that a power amplifier with a step-up transformer can be useful for test and transducer evaluation. (In this case, direct drive by a high voltage amplifier, as previously suggested by other contributors, would be superior).

The first step in building transformers is understanding the principle, the second is calculation, based on it. There are some empirical points, e.g. how leak inductance depends on windings geometry. But you'll find out by test easily.

As a final comment, also piezo transducer electromechanical properties are open to modelling and calculations. Particularly longitudinal waves and respective thickness vibrations of a piezo disc and attached objects doesn't involve complex relationships. Coefficient of elasticity, density, speed of sound and acoustic impedance make up the basic equations.
 

    A

    ants

    Points: 2
    Helpful Answer Positive Rating
Re: How to find an audio transformer, step up from 8ohm to 1

There are lots of similar types in the UK at:

https://uk.farnell.com

Type "transformer" into the search box, then select "audio" at the next selection. There are many similar types near the bottom of the catalogue page. You might find a 100V line matching transformer is more suitable than one for driving a loudspeaker.

You don't need an account to buy from them, they take credit card payments.

Brian.
 

    A

    ants

    Points: 2
    Helpful Answer Positive Rating
I am rather late to this thread, but driving piezos for atomising is quite a common subject. There is some good information on Morgan Electro Ceramics web site. Download their "book" here:

**broken link removed**

I would say you should be using a simple drive (power MOSFET) and you will need to be winding your own transformers. I have designed quite a few. Ideally I would say you need to run pretty close to resonance, although there are different views at https://www.mpi-ultrasonics.com/ and I wouldn't doubt his experience. You will have to try and see what works for you. If you are way off resonance though, you will generate a lot of heat (in your transformer & transistor) and not much ultrasonic power. Also, you can end up with a device that makes a lot of noise but doesn't achieve the desired atomising effect.

Keith.
 

    A

    ants

    Points: 2
    Helpful Answer Positive Rating
Re: How to find an audio transformer, step up from 8ohm to 1

you will need to be winding your own transformers
Click to expand...
Yes. Everyone says it. Also many people know, why a standard audio transformer doesn't work for your application.

Regarding resonant operation, I understand that MPI-utrasonics's "MMM technology" can be useful for transducers and loads that actually have modes. I would rather expect pure thickness vibrations for a basic atomizer transducer. Operating out of any mode resonance means waste of energy, as said.
 

Thanks for your replies. You are all very helpful!

I need to conduct some experiments on transformers to learn their theory, I will buy some of the transformers on the links you gave. I will then try to make my own, I have had a look at the coilcraft site to make inductors previously. It is very good.

I have as we speak gone back to adapting my piezo humidifier circuit. I can get the tank frequency up and running at the moment, but I can't get the crystal to oscillate, just yet. I think the piezo transducer I have designed should work ok, it does involve a fair bit of labour though.

Thanks for the link to the book. I will have a read of that. The circuit I'm currently looking at does use a MOSFET.
 

As a practical note, up until the integrated audio amp
(LM380, LM386) just about every cheap transistor radio,
tape recorder, etc. had a 1K(ish) center tapped to 8 ohm
transformer in its output stage.

Whether they would be suitable in the MHz range I doubt,
but few hundred kHz out just to be a matter of losses with
the stacked-plate core probably being not-terrible. And if it
were, you could pull the core out of the bobbin and fill with
(say) powdered ferrite and just enough epoxy, or something.

Presuming you found a junk radio laying about.

Now, do you require spectral purity at your tone of interest,
or is square wave drive tolerable? Because a complementary
MOSFET driver would probably be happy to smack that
piezo around for you, and the right series inductor could make
it a fairly tuned circuit if that (tamping down the harmonics)
mattered.
 

    A

    ants

    Points: 2
    Helpful Answer Positive Rating
Hi dick_freebird,

I am currently fighting with a Mosfet to get it to resonate on an adapted humidifier circuit, from 1.6MHz down to 133KHz. I can get the tank circuit running at 42KHz, but the piezo isn't oscillating, yet!

The audio transformer/LM3886 is my prefered method, but it is a useful exercise making a circuit work. I'm quite certain the 8ohm to 1Kohm transformer will be fine at 100KHz and I would have plenty of spare power with it.

Changing the piezo seem to require that many other components are also changed, for impedance matching, taking into account the different capacitance of the piezo and altered frequency.

It is a big job, must buy that transformer for a rest!

Thanks,

Ant.
 

Re: How to find an audio transformer, step up from 8ohm to 1

ants said:
I'm quite certain the 8ohm to 1Kohm transformer will be fine at 100KHz and I would have plenty of spare power with it.
Click to expand...

I don't think an audio transformer will work very well at 100kHz if has a laminated core.

Keith,
 

    A

    ants

    Points: 2
    Helpful Answer Positive Rating
You see,
these "Audio trafo solution" will surly not sufficient, up to very bad, work!
Otherwise you can surly remember on "Si trafos"(I remember as max ca. M32--40)_ their are developed for "PTT-HF" transmissions developed for multi carrier telecomm channels, regularly are good for so 100-150 KHz, these Si-plates are more thin as usual trafo plates and some greany colored... I think the power is a limit at these type too.
I afraid only, that its no more to buy:-(, but some ferrite power cores are in all switcher built in, & in all years in more & more examplare!

If trafo (but for me its a heavy or inpossible way...), than Ferrite power cores(
K.
 

    A

    ants

    Points: 2
    Helpful Answer Positive Rating
Re: How to find an audio transformer, step up from 8ohm to 1

ants said:
I have found these transformers, do you think they will work?

https://uk.farnell.com/jsp/search/browse.jsp?N=6434+500006+1001637

Thanks.

Ant.
Click to expand...

No. They are meant for collision avoidance sensors or alarms, I think. They won't take any serious power. For an atomiser I assume you are looking at several watts.

The best match I could find on Farnell was https://uk.farnell.com/oep-oxford-e...2a1e/transformer-audio-6-3-6-3-1-1/dp/1172346 but it is only specified up to 30kHz

Ditto https://uk.farnell.com/vigortronix/vtx-101-001/transformer-audio-6-3-6-3-1-1/dp/1674300

I think you will really have to wind your own ferrites.

Keith.
 

Re: How to find an audio transformer, step up from 8ohm to 1

Basically, a transformer can be characterized by a set of electrical parameters:
- windings ratio
- main inductance (has to be transformed with n²)
- leak inductance (as above)
- windings resistances and current ratings (also considering skin/promity effect for HF)
- winding capacitances
- saturation current respectively maximum ∫udt (related to each other)
- frequency dependant core losses

Depending on the application, some parameters can be possibly neglected. You can also calculate derived parameters, e.g. power
rating or usable frequency range. Rather than guessing, if particular retail products fit your application, you should define your
specification. You may want to draw a transformer equivalent circuit together with your source and load for clarity.

As previously mentioned, I agree with Keith about building your own transformer. Some SMPS transformers available as a
standard product may fit anyway.
 

Re: How to find an audio transformer, step up from 8ohm to 1

FvM said:
As previously mentioned, I agree with Keith about building your own transformer. Some SMPS transformers available as a
standard product may fit anyway.
Click to expand...

True, you may be able to get away with one of the pre-wound Coilcraft ones if the turns ratio/inductance match your requirements. Have a look at these:

Keith.
 

Status
Not open for further replies.

Similar threads