Choice of switching element for 3kW DC?

[carpenter]

AC 230V max 16A -> diode rectifier -> mainly resistive load -> some switching element to control the power using MCU
The question is which one?

I would normally use a MOSFET, but when I look at the current prices and availability I am horrified..
A common MOSFET 600V 20A will have an RDSon of approx. 0.25 Ohm => for 1$A power lost over 45W => big and expensive cooler transistors with RDSon below 100mOhm are expensive.
Another option is an IGBT or thyristor.
I don't know much about IGBTs and thyristors...

In short, what would you choose?
Regulation is very slow and accuracy is not critical, Price plays a significant role

 

[KlausST]

Hi,

Why switching the DC side, why not the AC side.

Switching elements are specified by voltage and current (power relays additionally give some motor power information)
Current causes heat
Voltage causes arcing

An opening switch in combination with series inductance will cause high voltage spikes and adrcing. On AC systems the arc will be extinguished in the zero crossing if current. DC does not have a zero cross, thus it does not self-extinguish an arc. Thus DC is more difficult to seitch OFF and thus the DC voltage rating usually is lower than the AC voltage rating.

What switching device you want to use is your decision:
There are ready to buy mechanical relays, there are triac based and MOSFET electronic relays (soloid state relays).
Triac based are for AC only.
And you may design your own electronic switch.

Which one to chose depends on your requirements. Besides I and V, switching count, frequency, zero cross switching or not ...
But also: cost, size, power dissipation, ruggedness, overvoltage protection, control power, humidity, dirt, safety level, isolation or not, ....

The simplest way maybe is a low side logic level MOSFET...
But as said: It depends on your requirements.

Klaus

 

[carpenter]

Originally, I did not consider the possibility of regulation on AC. Why.
See very simplified replacement load scheme
If the voltage on the load is lower than 170V, it does no useful work.
In fact, the voltage must be higher than 205V because the current must be higher than I1
The parasitic load capacity is approx. 900uF.

 

[d123]

Hi,

Saw a SiC circuit for HV DC of 1200V, it was e.g. three 600V FETs in series, controlled by a (relatively speaking) low-power Si MOSFET at the bolottom of the chain. That way you might use three * cheaper lower voltage MOSFETs instead of one * horribly expensive MOSFET.

Figure 1 in the article sort of shows the idea, only sort of. If I can find a document or article that shows the exact idea, I'll post it/the link.

Series-connected MOSFETs increase voltage & power handling​

Not ideal, but gives the idea if adaptable... Figure 1, again.

 

[Easy peasy]

Triac or back to back SCR's on the AC side will provide all the control you need ...