Hi everybody,
I am trying to develop a project where I need to use a step up converter.
My main power supply is 15V. I need a boost converter to get an additional 19V output. I have wired the VIN- input to GND and VIN+ input to a mechanical switch, which allows me to give 15V when on. When off, GND remains connected to VIN- but VIN+ is not connected.The output of the boost converter is not currently connected to anything.
The issue is, the boost converter shorts after a few power on and off cycles. Connecting it the first time works as expected, but after switching a few times the mechanical switch the power supply switches off. After disconnecting the boost circuit, I read 2 Ohm resistance between the 2 input terminals. Why does this happen? Am I wiring anything incorrectly?
I can provide a schematic if it’s needed, but I believe the circuit is pretty simple.
The boost converter I used is this.
This is a picture of the items used. Both experienced the same issue.
I know the wiring colors are not standard, but I am already using red and black elsewhere. Do not worry, it’s 15V DC and the polarity is correct.
I measure 19V on the output before it starts acting up.
Thing is, initially the module has something like 2K resistance, then it drops to 2 after it starts malfunctioning.
This is also suggested in some other listings for the same MT3608 boost circuit. However, it feels silly to me to just put a fake load (resistance) across the output leads even when there is no required load.
I have bought some XL6009 circuits which will hopefully fit my needs better. I will report on that.
It may seem "silly" to you, but the "generalized circuit" drawing reminded me that you can't pump power into a capacitor forever, you need something to consume it. Should be built in to the converter, but hey, without a schematic for it, who knows? Rather than "silly", it may be a design input.
Or course, I just meant that in my project the converter sits in an enclosure and 0.1A at 19V is 1.9W, which is a rather big amount of power to dissipate. I can see as it is not a problem in other use cases.
I guess the device fails due to an overshoot at power up. Do you have a scope? I would add a cap at the output to slow down the output voltage rise an give the converter time to compensate.
(The converter tries to quickly rise output voltage; this means the current is high and so energy stored in the coil is high. When the output reaches the target value the converter stops switching but all the energy stored in the coil is dumped into the output cap. That causes the overshoot. If the output capacitor is small compared to the coil the overshoot may be so large it damages the IC.)
Another possibility is the output capacitor is too large and this causes to high inrush current (when the output cap is empty the current is limited only be the coil resistance). You may try to add a small series resistor to the input side to reduce the inrush current.
+1
A boost converter has a high startup current that the supply (or buffer capacitor) must provide.
From cold start (no voltage on the output), the circuit is not ready yet to drive the mosfet.
After a start, and too low output voltage, the fet wants to do it's job (is 'on').
Then you will now measure the resistance of L1 (diagram post#4).
Try a 1000uF buffer capacitor on the 15volt supply, just before the switch.
Leo..
I expect OP measures the 2 Ohm resistance using a DMM. Typical DMM in Ohm range does not provide voltage high enough for the MOSFET to turn on and so it should measure quite high resistance. Also OP states the new modules had much higher resistance. I agree OP is measuring the coil resistance but this is because the MOSFET is destroyed and no longer can switch off.
I think I missed the fact that it is destroyed after a while.
No wonder if you are trying to get 15 + 19 = 34volt out ot a module that only is specified for 24volt max output.
Leo..
This one module solution to produce 19V (adjustable from 4-25V) is a quality product from a sound company that provides excellent customer and product support.
Thank for the suggestion. Next time I’ll consider it for sure. The documentation is extremely useful and companies should be rewarded for such a service.
Meanwhile, the XL6009 circuit works perfectly. Higly suggested instead of the MT3608.