Boost converter - Replacement for dual-winding transformer

[mindthomas]

Hi there.
I am currently trying to build a step-up converter also known as a boost converter. I have decided to give the MAX1605 chip a go to start with and later design my own controlling circuitry.
I the found this application note for the design which is just what I want to do - be able to generate around 60V from a 3V power supply.
**broken link removed**
The problem is now that I can't seem to find the custom designed transformer (or what it is called).

It is a SUMIDA CMD4D13 4365-T020 which is a custom designed but commonly used dual winding transformer.
I have asked our local distributors but they all say that this component is out of stock and will first be restocked in over 5 weeks.

Therefor I would like to know how I can find a replacement for this transformer. The specifications the transformer needs to fit is:

• Winding ratio (Np:Ns) ~ 1:5
• Primary inductance (Lp) ~ 4-5uH
• Saturation current (Isat) = 600mA max
• Resistance ~ 0.8 ohm

Do any of you guys know a transformer that fits these needs?
I have been searching a lot on the internet, both looking at Bourns and Coiltronics but I've been unable to find any transformers with a winding ratio of 1:5.

Please have in mind that the current needed at the 60V output of the boost converter is not much. It is only a couple of milliamps, so the transformer should be kept small (SMD version etc.)

I really hope some of you guys have a clue as I am really in deep water.

Best Regards
Thomas Jespersen

 

[mtwieg]

You should try looking at using coupled inductor arrays as an effective transformer. Usually all the windings have the same number of turns, but if there are enough windings you can configure some in series to give the desired effective turns ratio.

This is the closest thing I could find. A little expensive though, and it probably exceeds your power requirement.
https://www.digikey.com/product-detail/en/VP1-0059-R/513-1766-1-ND/2458728

 

[mindthomas]

You should try looking at using coupled inductor arrays as an effective transformer. Usually all the windings have the same number of turns, but if there are enough windings you can configure some in series to give the desired effective turns ratio.

This is the closest thing I could find. A little expensive though, and it probably exceeds your power requirement.
https://www.digikey.com/product-detail/en/VP1-0059-R/513-1766-1-ND/2458728

Wow! Yeah, that's a bit expensive - too expensive to be used in the product. Though maybe this could be used for the protoyping, buying a single one only?
So what you would do was to simple connect 5 of the coils in series and then leave a single one, and then use it like that?

Well I am still very open to other suggestions as I don't think it can be true that you can't find a coil with similar specs.
It is ok if the coil isn't 1:5 - it can also be 1:6 or 1:7 - as long as the saturation current and primary inductance is the same.

Thanks

 

[mtwieg]

Yes, by combining separate windings in the correct way (make sure the phasing is correct) you can make various transformer permutations.

I'd be surprised if you could find something similar off the shelf. Low voltage high frequency transformers is a niche market, and most parts will be custom made. Small coupled inductors are starting to be popular in high performance DC-DC converters, but most of those require equal turns on each winding. Hence why you can find lots of inductor arrays like that one. An actual 1:5 transformer will be much rarer.

 

[mindthomas]

Bump...

Any other suggestions for a transformer that would fit our need? Maybe one that is a bit smaller than the one mtwieg suggested?

 

[godfreyl]

Wouldn't it be easier to use a different design that just requires a simple inductor instead of a fancy transformer?

Disclaimer: That's just my 2c worth - I'm by no means an expert regarding SMPS, but in general I'd prefer a design that uses common off-the-shelf parts, rather than exotic unobtanium.

 

[mindthomas]

Wouldn't it be easier to use a different design that just requires a simple inductor instead of a fancy transformer?

Disclaimer: That's just my 2c worth - I'm by no means an expert regarding SMPS, but in general I'd prefer a design that uses common off-the-shelf parts, rather than exotic unobtanium.

Yes, that would be good and definitely better.
Unfortunately as the ratio between input and output voltage is that high it is difficult to make any efficient circuit that can generate the required output voltage and output current without using a transformer.

 

[BradtheRad]

One article about boost converters stated that when you want that large a step up multiple, that it works better to do it in two stages rather than one. This is based on using a simple coil per stage.

 

[mindthomas]

One article about boost converters stated that when you want that large a step up multiple, that it works better to do it in two stages rather than one. This is based on using a simple coil per stage.

We need to keep the component cost at a minimum, but it might be cheaper to do this instead of having to find a special, custom designed coil.
But the circuit is also a bit specific as the output voltage has to be controlled with a regulating loop measuring the output current.

If you should recommend a specific setup to generate 40-60V done in two steps and controlled by a PWM output by a microprocessor, what would you then suggest?

 

[ahsan_i_h]

Switching mosfet's (LX to GND) break down voltage is 32V.
So inductor can not be used.

 

[BradtheRad]

You can use a boost converter, then feed the outgoing pulses to a Villard style doubler or tripler.

A doubler means adding just one additional diode and capacitor.

The duty cycle may need some tinkering. Or the frequency. These may be handled by the driver IC.

You may need to divide down the reference voltage in order to get the desired performance from the driver IC.