1
\$\begingroup\$

I need a small surface mount or slightly larger transformer with a 1:3 turn ratio that will triple my voltage from 3.3 to 9.9 volts.

I have been literally searching the internet for such a thing and cannot find one. Can somebody please tell me what the heck you call the thing so that you can find it when searching for transformers?

I'm going to be running 2 PWM lines through it from an MSP430 controller and the PWMs will be doing a push-pull so that they generate +3.3 => -3.3 volts for a total of 6.6v peak to peak. They are also generating a nice 25khz square wave.

Anyway, I want to attach those 2 wires to the little transformer and have a nice pretty square wave coming out at around 18 volts peak to peak.

This needs to be very small. Preferably less than a centimeter square.

What do you call this thing and Where can I order one?

UPDATE (Clarification):

The purpose of this circuit is to drive a Piezo Transducer at 25khz and have it produce 110dB. In order to do this, I need a minimum of 10v Peak to Peak.

I am using an MPS430 controller which has 2 PWM pins. I'm using a push-pull algorithm to have those 2 pins produce opposite waveforms. So, I have 2 square waves being generated opposite of each other. This gives me +3.3v -> -3.3v or 6.6 volts Peak to Peak. The max voltage for the Piezo is 20, so I would like to double or triple my output voltage. One suggestion was a 3:1 Transformer. That seemed like the easiest way to do it. I am open to using whatever solution works, there have also been other recommendations. However, I would like to accomplish this with parts that do not exceed a total of $2.00. Also, the parts must be very small. This circuit needs to go on a board that is .75" x 1.5". It needs to fit in an enclosure that will attach to a dog's collar. The only power for the entire device is a small 3.7v rechargeable lithium battery.

Thanks !!

\$\endgroup\$
  • 1
    \$\begingroup\$ You call it a transformer with 1:3 turn ratio. Go somewhere where they sell transformers and search until you find one matching your specifications. \$\endgroup\$ – PlasmaHH Jun 30 '16 at 9:24
  • 2
    \$\begingroup\$ "Custom-wound". \$\endgroup\$ – Ignacio Vazquez-Abrams Jun 30 '16 at 9:25
  • \$\begingroup\$ I do understand what you mean: sometimes you find so few results on the Internet that you assume your search words must be wrong. In this case, there isn't a market for every kind of transformer so yours may just not be made. Incidentally, I would search for a coupling transformer but that's just me. I think you'll be left with winding your own or looking at a step-up DC-DC circuit or module. The latter is the common approach but may not meet your space or cost requirements. Good luck with it all. \$\endgroup\$ – TonyM Jun 30 '16 at 9:40
  • \$\begingroup\$ You verbal description of your circuit isn't really adequate for us to understand. How can it be "PWM" and "nice...square wave" at the same time? Why do you need an 18V signal? Why do you think it would be good to do that with a transformer? vs. an active circuit which may actually be smaller, more efficient and not require unobtanium components? \$\endgroup\$ – Richard Crowley Jun 30 '16 at 9:46
  • 1
    \$\begingroup\$ I keep thinking the solution I suggested would be much cheaper and much smaller than a transformer. Using dual P/N channel fets packaged in SOT-363 and multiple diodes in a single package, you can achieve what you want. You can double, triple, quadruple the voltage... Oh, well... \$\endgroup\$ – dim Jun 30 '16 at 10:19
8
\$\begingroup\$

Coilcraft are a turn to company and they do a few small ones at 1:3 ratio: -

enter image description here

All you have to do now is ensure that the primary current (25 kHz) doesn't saturate the core. V = L di/dt is the formula to use here. Coilcraft tell you the saturation current so I'll leave it to you to plug the numbers in. Let me know if you need help.

I think these might be too small and will saturate but check out what else coilcraft have to offer and maybe also consider upping the drive frequency to 100 kHz to make life easier on the core saturation.

\$\endgroup\$
  • \$\begingroup\$ OP seems to be converting voltage so I would suggest to use Vt=NAB for a calculation but since these are ready made, Ipeak=t*V/L. \$\endgroup\$ – winny Jun 30 '16 at 10:47
  • 2
    \$\begingroup\$ You also have to ensure that the MSP430 GPIO pins can supply teh required primary current, which I suspect will be a bigger problem. You might be better with 1 GPIO pin, driving a FET, and a 6:1 transformer. \$\endgroup\$ – Brian Drummond Jun 30 '16 at 11:11
  • \$\begingroup\$ The magnetization (imaginary) current would be high at 22uH and 25 kHz. A higher impedance may help, such as a 78250MC from Murata, murata-ps.com/data/magnetics/kmp_78250.pdf \$\endgroup\$ – scorpdaddy Jun 30 '16 at 14:35
  • \$\begingroup\$ @scorpdaddy yes I estimated it will saturate but this my answer is just guidance and hopefully the OP will use that advice to search a bit more or raise the frequency to maybe 100 kHz as I suggested in the answer. \$\endgroup\$ – Andy aka Jun 30 '16 at 17:05
  • \$\begingroup\$ Thanks !!! I've ordered several samples from them and should be getting things put together very soon. \$\endgroup\$ – Curtis Jul 1 '16 at 2:22

Not the answer you're looking for? Browse other questions tagged or ask your own question.