2
\$\begingroup\$

I made a boost converter based on the below schematic. I want to use this to create 18V.

Boost convertor based on MT3608L

When I connect a 12V powersupply to the connector P3IN, I only measure the same 12V at the output on connector PWR3. Changing the potmeter R52 seems not to change anything.

I have 2 PCB's (ordered via JLCPCB - minimum qty for SMT assembly is 2), so I tried both. Unfortunately, on my second board the MT3608L blew up. So there must definitely be something wrong.

Does somebody has a clue?

Edit 29JUL2022

Some more information as requested in the below comments:

  1. The inductor is SWPA6045S220MT. I choose it directly from the JLCPCB components, which shows chinese data sheet. But when I google, you can also find it at DigiKey.

SWPA6045S220MT datasheet on DigiKey

  1. The layout of my PCB is shown below. As this was only an experimental PCB, the boost convertor is just a complete separate section on my PCB. Just ignore the USB connector DC1. I only show 2 layers, but this is a 4 layer board. One of the mid-layers is GND, which is not visible, but that obviously does connect all the GND together. The GND pins of connectors P3IN and PWR3 are connected through that inner layer.

Boost Convertor PCB

Edit 01AUG2022

Good news, the boost convertor works as a charm. It was a matter of tuning the potmeter correctly. I removed the potmeter from the my PCB, measured a correct 63.8kOhm, and connected it with some wires back in the circuit in such way that R52 becomes 63.8kOhm, and FB is connected between R51 and R52. After that, I got almost exactly 18V.

The circuit I made is indeed very dangerous, because if the resistor wiper gets to close to the output, you might exceed the max 6V rating on the FB pin, which destroys the chip (which was the case in my first test). As Verbal Kint commented, it's indeed better to use a fixed resistor divider if I only want one fixed voltage.

Also worth mentioning is that I didn't need any extra load to make it work.

Consider this as solved.

\$\endgroup\$
10
  • 6
    \$\begingroup\$ Which inductor are you using? Can you link to the datasheet? Can you add a picture of your PCB layout? Have you checked to be sure the polarity of your diode is correct? Have you looked at the switch node waveform on a 'scope at startup? \$\endgroup\$
    – John D
    Commented Jul 28, 2022 at 17:23
  • 3
    \$\begingroup\$ Looks like a very crude boost controller to me, never heard of this brand. A boost cannot operate unloaded unless there is a specific circuitry bringing the duty ratio to 0% which does not seem to be the case. So, first off, load this converter before turning on the power. Second, don't use a potentiometer as you may bring FB to \$V_{out}\$ and I suspect the max rating of the pin is 5 V or so. Immediate destruction of course. If the wiper is too close to the ground, then you can't regulate and \$V_{out}\$ runs away: the SW pin breaks. Put a fixed resistive divider then to have 0.6 V on FB. \$\endgroup\$ Commented Jul 28, 2022 at 18:28
  • 4
    \$\begingroup\$ Please show your layout. \$\endgroup\$
    – winny
    Commented Jul 28, 2022 at 20:43
  • 1
    \$\begingroup\$ I see that your circuit is essentially that which is shown on the device instruction sheet: utmel.com/components/…. So for any input voltage greater than 6V, there should be a resistor on the high side of the potentiometer R52 that limits the FB voltage to 6V. \$\endgroup\$
    – PStechPaul
    Commented Jul 28, 2022 at 21:13
  • \$\begingroup\$ @JohnD I edited my post and show the conductor and PCB layout. I did use my oscilloscope to see some switch node waveform, but couldn't see anything else than a flat line. \$\endgroup\$ Commented Jul 29, 2022 at 6:24

1 Answer 1

7
\$\begingroup\$

The MT3608L doesn't tolerate a voltage greater than 6V on it's FB pin. If you know that you need a 18V output, I'll would change the potentiometer by a fixed value resistor (the formula should be in the datasheet), or I would at least add a fixed resistor in serie.

The problem (maybe not the only one) of your circuit is that the voltage on the FB pin can be as high as the output (so over the FB pin limitation) if the potentiometer is not well set.

Also, John D pointed some good points

[EDIT] : Seeing you connected your output to several connector's pins, you maybe want to connect your ground to several ones too. The output current will also flow through the ground. Depending on your connector's current capabilities, it may not be an issue for the 2.5 Amps of the MT3608L, but it's good practice if you intend to work with higher power.

\$\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.