Overloading this transformer

[dknguyen]

I don't expect anyone to be able to give any hard answers, but just a "feel" to see if this okay (unless someone has hard data or knows the thermal time constants for transformers like this):

For the 40-50mW transformers on this page
https://www.trcelectronics.com/Tamura/microtran_audio_transformer_miniature_encapsulated.shtml

Do you think it would be okay to run them at 2% duty cycle with at 8W with an on-time of 2ms? The effective dissipation is 160mW at this level so I still might need to parallel 3 transformers. I don't know or have a feeling for the thermal time constant to know if the on-time is much shorter than the thermal time constant for this to work.

 

[audioguru]

If you feed those tiny transforemers anywhere near 200 times their power rating then the magnetic core will saturate and the output will be severely distorted DC.

 

[Sceadwian]

Why would you want to do that? This is one of those kinds of questions that makes you just go "Huh?!"

 

[dknguyen]

I see. So there is something else that instaneously limits the power. How is this tiny transformer working then? It's meant to be used for sonar.

https://www.electro-tech-online.com/custompdfs/2007/03/ranging20transformer20spec.pdf

I need a step up trasnformer that has a high low frequency bandwidth for driving electrostatic ulrasonic transducers. And it's difficult to find large transformers with 30kHz-100kHz bandwidths.

 

[Sceadwian]

It doesn't limit the power used, it limits the transfer of power. If you saturate the coil the rest is going to be wasted as heat. Why use such a dinky transformer, why not use one at least closely rated to the power you're going to be using?

 

[dknguyen]

I haven't been able to find any 10W transformers that specify bandwidth, and most aren't audio transformers and are designed for 50/60Hz use (either that, or they are enormous 100W+ audio transformers).

I've searched Pulse, Tamura, Hammond, Coilcraft, Plitron and the Digikey catalog, and more often that not the bandwidth isn't listed.

 

[Sceadwian]

What turns ratio pri/sec do you need?

 

[dknguyen]

The secondary AC voltage is 150-200Vpp. The primar side can be 5V, 10V-15V, or 20V. The insulation breakdown has to be able to withstand >400V though since there is a 200V bias on the secondary side.

So I guess anywhere between 50:1 and 10:1, something closer to 15:1 is preferable though so I can use a 12V or 15V supply to get 150Vpp.

 

[Sceadwian]

What's the maximum frequency?

 

[dknguyen]

Oops, could have sworn I said that somewhere. THe sonar is working between 40kHz-60kHz ideally, but is operable from 30kHz-100kHz so that would be nice. GLobalSpec seems to be saying I can only get something like that if it's custom made.

It seems no transformers exist for this frequency range. They are all 20Hz-30kHz (50kHz at the most), and then they are suddenly from 500kHz and up.

It's also going to be a pain to try and build my own since there's so many factors involved to what I'm trying to do already and I need 5 (maybe even 10). Don't want to debug 10 transformers. Bleck.

 

[Sceadwian]

Was just doing a little searching, looks like power transformers meant for switch mode power suppplies will work for you. They're meant to work at 100khz, and 8 watts isn't a lot of power, you might be able to get something useful out of a switch mode cell phone charger.

This is the first web site I ran across on google that had something that might suit you. Hope it helps.

**broken link removed**

 

[mneary]

Have you considered a simple switcher in "boost" configuration? There's a node in there that has your output voltage peak to peak. Isolate the DC with a capacitor. You can use the DC output for your 200V bias.

I haven't worked this all the way through, so I would appreciate hearing if I missed something important.

 

[dknguyen]

Switch mode transformer eh? I've been kind of avoiding them because I just know they're optimized for unipolar operation, and I didn't know how that affects regular sinusoidal distortion.

You are referring to using the capacitor so that I can apply DC bias directly to the transducer while preventing the DC from shorting across the transformer coil? I was considering that except it would not prevent the AC drive signal from going into the DC-DC converter. I am not sure how the DC-DC converter would behave it that happened.

It was also a problem because the capacitor would have to be far too large to pass 40kHz with less than a fraction of an ohm impedance. But that may not matter if I use that impedance for matching purposes.

I was just going to use a DC-DC converter to bias it (that was what you were referring to with the switching booster right?)

 

[ericgibbs]

hi dknguyen,

The RM7,8 and 10, ferrite transformers work well at these frequencies.
They can easily be wound for 12Vac to 200Aac step up.
Measure about 15mm thru 25mm dia.

I have used them for acoustic sonar from 33KHz thru 210KHz, square wave driven, tuned to resonant with the ceramic transducer.
Power 5Watts thru 100Watts, pulsed 100uSec thru 3mSec at 10pps.

Get the RM transformer data sheet.

EricG
EDIT:
If you don't want to wind all 10 txrs, design 1, test wind it, then look thru the trade catalogs for a local prototype coil winder.
Its not to expensive for 10 off.

 

[Hero999]

Can't you wind your own?

Most ferrite cores have bandwidth data available and switch mode power supplies seem to handle 500kHz with no problem whilst transferring hundreds of watts.