2

I am trying to connect a photovoltaic panel directly to a heating element (coil) without using a battery or an inverter and switch it on or off by using a transistor or a thyristor.

I am well aware that the power won't be constant throughout the day and not have power at all in the night, so you don't have to warn me about that.

The panel supplies a max of 37V and a max current of 8.3A. First I was looking for a high current MOSFET and found quite a few. At home I had IRF3205 which are rated for currents up to 110A and 55V.

Just to be on the safe side, I connected two of them in parallel by following two schematics I found on stack overflow. You can see them in the pictures below:

And here's the link to the stack overflow post: Parallel MOSFETs

I placed the MOSFETS on a large heatsink and then I connected the photovoltaic panels. The MOSFETS have a diode between source and drain which got shorted in just a few seconds after I connected the power from the panels.

I also have a large diode connected to the heating element to prevent reverse current.

I never connected the +5V DC to the gate of the transistor which means it was always closed.

I have also tried using TIP35C which shorted out just like the IRF3502.

Can anyone suggest a solution to this problem with any kind of transistor, or maybe using thyristors and an optocoupler?

Just to clarify some things, the transistor or thyristor will be connected to a microcontroller which will control the switching of power from the photovoltaic panel to the heating element.

Goran Kostoski
  • 23
  • 4
  • 2
    Welcome to EE.SE. (1) "*37V of power*"; 37 V is a measure of voltage, not power, and we just say "a max of 37 V". (You wouldn't say "13 m of distance".) (2) If the MOSFET diodes conducted then you have them connected the wrong way around. (3) "*I never connected the +5V DC to the gate of the transistor which means it was always closed.*" Not so. The gate resistance is so high that they are susceptible to stray voltage. You need a pull-down resistor to turn them off properly. (4) Thyristors are generally not suitable for DC circuits as there is no easy way to turn them off. – Transistor Jan 28 '20 at 19:55
  • I appreciate your welcome! (1) Sorry about that, I didn't realize I typed it like that. It is fixed now. (2) I triple checked that the diode wasn't the other way around. After burning the first 2 MOSFETs, I connected two more, and the same thing happened. I'm sure that the diode is the right way, even checked on the internet to see if I made a mistake. (3) I am aware of the "_stray voltage_" but I didn't think it might be enough to open the transistor. I will try it again and give you an update. (4) I thought of turning the thyristor off with another less powerful transistor, will that work? – Goran Kostoski Jan 28 '20 at 20:15
  • 3
    When the switch is off, what defines the gate voltage? I see no pulldown resistor. –  Jan 28 '20 at 20:20
  • When i was trying to do this with the TIP35C I did put a pulldown resistor, but the transistor was shorted in matter of seconds. When I was working with the IRF3205 I did not put a pulldown resistor, because that's what I saw on these schematics here on StackExchange. – Goran Kostoski Jan 28 '20 at 20:32
  • 2
    It might be time to post the real schematics. We're looking at his of schematics with no solar panel, etc., and we all guessing. – Transistor Jan 28 '20 at 20:51
  • 1
    The panel is *37V* but *5V* is shown on the schematic - was 37V accidentally applied to the MOSFET gates? If so, the gate pin can only accept a range of +/-20V. – rdtsc Jan 28 '20 at 22:53
  • 1
    As shown with load in source lead, Vgate must be driven to Vload + Vgson to turn FET on. For > 20V supply FET will be destroyed.| SO Place load in Drain side of cct. PV+ - load - FET_drain ... . | Floating gate may easily be driven to Vin by leakage. FET will be destroyed. SO Connect gates to ground with a resistor at all times (say 10k but anything smaller also OK). Consider connecting a say 12V zener gate to ground so gates can NEVER be driven above about 12V. – Russell McMahon Jan 29 '20 at 00:20

1 Answers1

3

As shown with load in source lead, Vgate must be driven to Vload + Vgson to turn FET on.
For > 20V supply FET will be destroyed.
SO Place load in Drain side of cct. PV+ - load - FET_drain ... .

Floating gate may easily be driven to Vin by leakage.
FET will be destroyed.
SO Connect gates to ground with a resistor at all times (say 10k but anything smaller also OK).

Consider connecting a say 12V zener gate to ground so gates can NEVER be driven above about 12V.

5V gate drive is marginally low for IRF3205.
Probably OK but not as fully enhanced (on) as you may like.
Fig 1 & 2 in the datasheet show Vdson vs Vgs and Id.
At 8A with Vgs = 5V expect about 0.1 Ohm Vds FET dissipation = V x I = 0.1 x 8 = 0.8W - say about 1 Watt/FET.
You have not said what package you are using.
A modest amount of heatsinking will be advisable.

schematic

simulate this circuit – Schematic created using CircuitLab

http://www.irf.com/product-info/datasheets/data/irf3205.pdf

Russell McMahon
  • 147,325
  • 18
  • 210
  • 386
  • Thank you so much! I was finally able to test it and your schematics works perfectly. Connecting the load to the drain side of the MOSFET, pulling the gate to ground and placing a 12V zener diode solved the problem. The IRF3205 doesn't even get warm. Once again, THANK YOU! – Goran Kostoski Feb 03 '20 at 10:26