Electronics > Projects, Designs, and Technical Stuff
trailing edge MOSFET dimmer over current protection
(1/2) > >>
sascha32520:
Hi,

I designed a trailing edge MOSFET dimmer a while ago and recently I had an inductive load that was jammed and when increasing the power, the fuse probably blew and the entire energy of the motor went into my dimmer blowing the snubber, MOV and some traces.

So I thought about adding a over current protection that will prevent blowing the fuse in the first place. This might not help in all cases. I do not think it would be fast enough if a short occurs at the peak of the sine wave, but that would be something for experiments.

Now my problem is that the dimensions of the PCB are 56x32 and it is pretty cramped on there already. HLW8012 current sensor (not suitable for fast current measurements), ac dc buck regulator, atmega328p, esp12f module, zero crossing detection, mosfet driver, voltage regulator for mosfets, etc..

Adding a dual opamp to rectify the voltage at the MOSFETs would take quite some space. It could feed it directly to the atmega and trigger an interrupt if the recitifed and amplified voltage at the MOSFETs exceeds 0.7xVCC=~2.3V.

A more space saving variant is adding a diode and a voltage divider that basically does the same thing, but only for the positive half wave. It can be directly connected to the atmega pin, the protection diodes should prevent any damage from the high voltages that occur when the MOSETs are turned off. This might prevent damage if the load pulls too much current in general, but one half wave would not have any protection.

I added a ltspice simulation with 2 channels and the diodes + voltage divider. L is my GND for the atmega and I have 5V and 3.3V supply available.



Maybe there is a complete different approach to solve this problem. I also have a 0.005R current shunt for the HLW8012 in this curcuit that is connected to L. Adding something like an INA219 current sensor. That particular sensor is I2C and polling the bus is probably too slow, it needs to be something that can trigger an interrupt....
Circlotron:
Inductive load???
What is going to happen to the load current when the mosfet switches off?
sascha32520:
It will go into the snubber network but due to the jam it probably sent 100A through it instead of the regular 2-3A... adding another fuse at the output would have prevented any damage, but running a motor is not the regular use. Just how I created the magic smoke ;)
It would be a nice safety feature though.
Circlotron:
Depending on your snubber, I think an inductive load is a disaster waiting to happen. Could you post the power end of the circuit e.g. mosfet and snubber?
sascha32520:
Yes, in trailing edge mode this dimmer isn't really suitable for running large inductive loads. I mainly use it for LED dimming and occasionally for experiments.

In that particular case I had 6x100R and 6x220nF connected in parallel, ~17R/1.3uF, a pair of NCE65T360K and a 14D221K across the mosfets.



https://easyeda.com/sascha23095123423/iot_1ch_dimmer_copy_copy_copy

Navigation
Message Index
Next page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod