2
\$\begingroup\$

I have an LED string that requires 140V that I plan to use in a circuit that has a 5V power supply (capable of up to 4A of current.) The LEDS glow at an acceptable brightness with as little as 50mA. So the LED string will require 7W of energy minimum.

When it comes to generating the required voltage, I explored two options.

Inductor Boost

I got a successful simulation in LTspice of this circuit, but it involved using an insanely high duty cycle of 96% and the peak currents surpassed what my power supply could handle. I also received advice that an inductor boost of a magnitude greater than 5 can lead to issues.

Flyback Converter

I'm exploring this option but a couple of things are not clear. For one, I have not come across a transformer that has a 1:30 step up ratio. Does such a thing exists? Additionally, I have come across conflicting advice on whether I could use one of those 240V to 5V transformers in reverse (paying special attention to impedance.)

I also came across this previous question, where the OP was trying to do something similar. They used the 750315832 transformer to boost from 5V to 200V. That doesn't make sense to me. The primary coil of that transformer doesn't even support anything less than 9V, and the output looks like it's rated at only 100V.

\$\endgroup\$
6
  • 4
    \$\begingroup\$ Transformers don't have a lower limit on the input voltage. The only limit on the output voltage is the breakdown voltage of the insulation on the coils. \$\endgroup\$
    – JRE
    Sep 21, 2022 at 19:05
  • \$\begingroup\$ Oh that explains how the OP from the previous post used 5V in. I'm still confused on getting 200V out when it's rated at 100V. \$\endgroup\$
    – Movieboy
    Sep 21, 2022 at 19:40
  • \$\begingroup\$ Anything from 1:10 to 1:25 would be fine. \$\endgroup\$
    – Andy aka
    Sep 21, 2022 at 19:53
  • 1
    \$\begingroup\$ As you discovered in your experiments, a boost converter is a very poor choice for such a high output-to-input voltage ratio. Flyback and forward converters are typically the way to go for a job like this. I wouldn't use a boost converter for producing more than a few times the input voltage, generally. \$\endgroup\$
    – Hearth
    Sep 21, 2022 at 19:56
  • 1
    \$\begingroup\$ Transformers don't care how they're wired -- look into offline (AC-DC) converter transformers, e.g. 90-265V primary, 5V secondary. Just ignore the extra (10-15V) aux winding they usually come with. And the ratio can be pretty loose so anything from 5 to 12V secondary would likely be fine, likewise anywhere 90 to 265V primary. \$\endgroup\$ Sep 21, 2022 at 23:05

2 Answers 2

2
\$\begingroup\$

Flyback Converter

I'm exploring this option but a couple of things are not clear. For one, I have not come across a transformer that has a 1:30 step up ratio. Does such a thing exists?

This is a misconception, you do not need a 1:30 step-up ratio. For a push-pull converter you need a 1:30 transformer, for a flyback converter you do not.

You already know that, if one abruptly interrupts the current flow in an inductor, the inductor will generate a large "kickback" voltage. Because \$ V = L \frac{di}{dt} \$. The generated open-circuit voltage depends only on the inductance, current, and the turn-off speed of the switch. This is a boost converter.

If you want to obtain an even higher voltage, you can add a second winding to the inductor, it now becomes a "flyback transformer" (some prefer to call it a coupled inductor) to boost the high kickback voltage to an even higher value with the help of the transformer's turn-ratio. This is essentially how a flyback converter works.

In theory, the output open-circuit voltages of both a boost and flyback converter are unlimited and does not depend on the turn ratio. Though, in a real circuit, ultimately the highest achievable voltage depends on the load and circuit parasitics.

As Andy aka has commented, anything from 1:10 to 1:25 would be fine.

Additionally, I have come across conflicting advice on whether I could use one of those 240V to 5V transformers in reverse (paying special attention to impedance.)

This is another result from the same misconception. A free-running transformer solely uses the winding inductance for current-limiting, so if you use a 60 Hz transformer on 50 Hz, overheating is a risk. If you connect the 12 V low-voltage side of the transformer to 220 V, the transformer will be destroyed due to overcurrent.

On the other hand, most flyback converters work by sensing and regulating the primary current, with cycle-by-cycle peak current limit. So impedance is a non-issue. Operating frequency is, though.

They used the 750315832 transformer to boost from 5V to 200V. That doesn't make sense to me. The primary coil of that transformer doesn't even support anything less than 9V, and the output looks like it's rated at only 100V.

This is another misconception.

If you open the website of an electronics vendor and search for transformers, you'll see most transformers have specified input and output voltages, such as 230 V input and 12 V output. But it's important to realize that these numbers are not the actual ratings, they are just vendor's recommended applications.

What's happening here is that, if you're designing a power supply, you usually want to use a standardized transformer and not a custom-made part. Thus, the transformer vendors would take a look at the customer demands, then go to design transformers for some particular applications in mind - AC/DC 110V/220V off-line power supply, DC/DC converters, and gate drivers. Often even with a reference design of a complete power supply.

Also, because chipmakers know the customers won't be happy if they know the latest chip needs a custom transformer that doesn't exist. Thus, sometimes, semiconductor vendors collaborate with transformer vendors to get some transformers made as well. This is where the "specs" of a transformer came from.

But there's nothing preventing you from being creative and use these transformers in different circuits (as long as the turn ratio, winding inductance, current rating, and withstand voltage are suitable for your application). Using the transformers in reverse is also okay, as long as the ratings are appropriate.

Take a look at your Wurth Elektronik 750315832 transformer transformer again. It's not a suitable transformer due to low turn-ratio, but notice that the specification reads:

  • Dielectric:
    • 1875 VAC, 1 second
    • 1500 VAC, 60 seconds

This is the actual "voltage rating" of the transformer. Not quite, since it's only a Hi-Pot stress test, not a continuous working voltage rating. But as long as you stay significantly below the Hi-Pot voltage, you'll be fine (though, if the transformer is meant to protect human safety - for example, in an isolated AC/DC or DC/DC power supply - it must be rated and certified for reinforced insulation).

Transformer Selection

So what transformers should you use? Here are some guidelines:

  1. Select "flyback" topology.

    This is important, flyback converters are meant to use transformers as inductors, so the transformers are designed to allow a high current flow without saturation. Meanwhile, forward and push-pull converters use transformers as ideal transformers, transformers optimized for these topologies easily saturate and cannot be used for flyback.

    A good way of distinguishing transformers for flyback and forward converters is observing the current rating. Flyback transformers are usually rated in peak current, say Ipk = 2 A. Other transformers are rated in volt-second products.

  2. Ignore the voltage ratings of the primary and secondary.

  3. Look for anything with the turn-ratio you need, e.g. both 1:10 and 1:0.1. Since the latter one can be used in reverse.

  4. Check the inductance ratings of the windings.

    In particular, some step-down flyback transformers won't work very well when wired in reverse, because the primary inductance (marked as secondary) inductance would be too low.

    Inductance ratio is the square of the turn ratio or voltage ratio. A 1:0.1 step-down transformer with 200 μH primary inductance will have a secondary inductance of just 2 μH, and will be a terrible idea to be used in reverse. The flyback controller senses the peak current and immediately turns off the switch, forcing it to work with an uncomfortably short on-time, and little energy can be transferred in each cycle.

  5. Check the current rating of the windings, also take a look at the resistance (DCR) while you're on it.

    Peak current rating (Ipk) is especially important, all inductors stop having an inductance and become nearly a short-circuit (for hard-saturation materials) after the core is saturated.

  6. Check the Hi-Pot rating between the primary and secondary side of the transformer. Make sure your operating voltage is significantly under this value.

    For safety applications, there are additional special requirements. But for a "functional" circuit without safety requirement, just leave a good margin.

Happy searching.

Examples

Here's some potential (no pun intended) transformer candidates. I don't guarantee they are suitable choices, but they worth taking a look.

1. Capacitor Charger Transformers

Years ago, Linear Technology made a high-voltage flyback converter for charging photoflash capacitors, called the LT3750 and LT3751. Because Linear didn't want to leave their customers without off-the-shelf transformers, so they contacted various transformer vendors to create a few models, all are still available.

Linear Technology LT3750's Recommended Transformers

Here's Coilcraft's offerings.

Coilcraft's capacitor charger transformers

2. WE-FLEX HV Flexible Transformer for DC/DC Converter

Wurth Elektronik has a bunch of "uncommitted" flyback transformers in 1:1:1:1:1 arrangement. It's possible to get a 1:5 turn-ratio by connecting some turns in series.

3. Coilcraft low-profile 1:N coupled power inductors

Coilcraft has some 1:N coupled power inductors for flyback applications.

Coilcraft's low-profile coupled power inductors

\$\endgroup\$
1
\$\begingroup\$

The “inductor boost” is not good at all – it makes very controversial requirements for switch element: high current and high voltage at the same time, so using transformer is the right way to go. It will reduce duty cycle, peak switch voltage and peak diode current. At the same time it will increase backward voltage on the output rectifying diode. That given, turns ratio may vary quite wide: in somewhere 5…30. Transformer may have non-isolated windings (tap configuration).

The turns ratio is not the main parameter to select. The most critical one is saturation current. For your application it must be at least 3 A for the primary winding (with least turns). Of second importance is primary inductance, which will define the working frequency. 10 uH will give about 20 kHz minimum frequency. And the turns ratio is only the third important parameter.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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