How to make a huge powerful transistor with your own hands

One day, for one of my projects I needed a powerful mosfet. A large current of about 500 Amperes had to be switched. And then the idea came to mind to make a separate transistor, modular, composite.
From low-power mosfets it was necessary to assemble one powerful one. Fortunately, these transistors are well arranged and work perfectly if they are connected in parallel to each other.

Will need

Transistors - 50N06 - 12 pieces -
Metal adhesive, cold welding -
Epoxy resin.
Textolite board for installation.
Flux and solder.

The process of making a powerful transistor with your own hands

We make the board using any available method.

We etch it in ferric chloride and then rinse it well.

We solder the technological jumpers and tin the board tracks.

We solder the field-effect transistors. This compound transistor uses 12 mosfets of the 50N06 series in a TO-252 package. All parts are connected in parallel without any correction circuits.

The contacts are made of thick copper wire.

Before soldering, you can temporarily connect thick wire with thin wire and then remove it.

We wash the board from flux.

We glue the base of the contacts with epoxy resin.

We cut out a heat-conducting substrate for the transistor from a thick sheet of aluminum.

We drill a hole for future mounting on the radiator.

We “plant” the board on heat-conducting glue - cold welding.

To give full visibility to the transistor, we will make a housing for pouring from plexiglass.

Mix epoxy resin in the proportion indicated on the instructions.

Fill in the form. and leave to dry.

We remove the plexiglass body.

Almost ready to do it.

All that remains is to apply the markings. Let's do this with a laser.

We rub the indentations with paint.

The power MOSFET transistor is ready for use.

Watch the video

come back

Similar master classes

Powerful transformer from three low-power ones

Radiator for low-power transistors

How to make a powerful electric bike using 4 low-power motors

DIY powerful strobe light

Powerful electromagnet from a microwave oven

How to make a mega powerful flashlight from old batteries from

Particularly interesting

Cable antenna for digital TV in 5 minutes

A selection of simple and effective schemes.

Three-phase voltage from single-phase in 5 minutes

Starting a three-phase motor from a single-phase network without a capacitor

Eternal flashlight without batteries

How to make an inexpensive but very powerful LED lamp

Comments (25)

#1 Dmitriy Guests 10 July 2020 15:22
14

Is it possible to connect semiconductors in parallel? It is not recommended due to the variation in parameters. As a result, there will be different currents through different transistors, possibly differing by tens of times.
- Answer
1. #2 Guest Vasily Guests 11 July 2020 20:53
  2
  
  There are devices that allow you to select pairs of transistors according to parameters. with great accuracy.Naturally, for this you need to have as many transistors from one batch as possible.
  
  Answer
2. #3 Gridin Maxim Guests 19 July 2020 19:48
  4
  
  Yes, you can use mosfets if they are from the same batch.
  /Only the sources need to be through separate small resistors of a few ohms
  
  But in general, of course, it’s nonsense. Why spend money on such a massive body?
  
  More on ALI
  1) there are practically no good mosfets, especially high-current ones, and you won’t be able to save money on them. I bought many, many times and then had to open disputes.. Unfortunately, it’s better to buy from us offline, no matter how painful it may be.
  2) there are ready-made boards for paralleling. For DC spot welders. They are relatively inexpensive. They can then even be connected to each other in stacks, also paralleling the power contacts with bolts
  
  Answer
3. #4 Dmitriy Guests 24 August 2020 21:47
  1
  
  Field-effect transistors can be connected in parallel. This is exactly what they do to increase power.
  
  Answer
4. #5 Sherpa Guests 30 August 2020 13:06
  0
  
  You can, but only be careful!
  
  Answer
#6 AL_Elektronik Guests 11 July 2020 16:53
18

Dmitry, compare mosfets transitions with resistances. But the conclusions and tracks are rather weak, as for me at 600 Amps
- Answer
1. #7 Aritenko Sergey Guests 14 August 2020 09:36
  2
  
  What is the cross-section of a 600A fuse? I think these legs will withstand 400 amperes for the lead length if they are not unsoldered by heat and the melting point of copper is much higher than solder
  
  Answer
#8 Vitaly Guests 11 July 2020 22:30
18

It will all burn out at the first decent current. It is unlikely to withstand even 100 amperes. Remember the wires of a normal car cigarette lighter - they are as thick as a finger. There, too, by the way, up to 600 amperes at the start.
- Answer
1. #9 Guest Victor Guests 23 July 2020 20:45
  14
  
  Well, of course everything will burn! It will burn because the thermal resistance of the fiberglass will not allow the transistors to be effectively cooled. The heat sink must be copper and the transistor housings are directly soldered to the heat sink; for this purpose, there are special D2pak housings. Parallel connection of field-effect transistors is the ABC of circuit design, attractive because there is no need to use equalizing resistors. The second century is used in power electronics!
  
  Answer
#10 Cat Guests 11 July 2020 23:16
28

I’m wondering how many seconds this “transistor” will work, and what will be the first to not withstand overheating, the tracks or the transistors???? A track 5cm long and 1mm wide (track on the right) is a good fuse, but I’m afraid it’s less than 600A))) Well, the thermal conductivity of the PCB will not allow heat to reach the aluminum “substrate”)))
And so, it turned out to be a beautiful trinket, you can put it on a shelf and be proud of it
- Answer
#11 Vyacheslav Guests 14 July 2020 16:53
15

Purely to amuse...
1- textolite conducts heat poorly
2-tracks and pins are not designed for the declared current
3- The huge gate input capacitance will not allow this transistor to operate effectively at frequencies above 50 Hz.
4- the gate driver for this product must provide 50 amperes per charge of the gate capacitance and will be at least the size of this transistor....
- Answer
1. #12 Sherpa Guests 30 August 2020 13:13
  2
  
  Your command of the question is 5. I agree with you 100%
  
  Answer
#13 Michael Guests 15 July 2020 14:30
7

Dmitry, it is possible with equalizing resistors (the same “correction circuits” that the author did not install). Without them, yes, the current will be different, and the transistors will very quickly “fly out” one after another.
- Answer
#14 Guest Maxim Guests July 23, 2020 00:49
7

Well, it’s not like transferring heat to the radiator through ordinary textolite, that’s just nonsense.It’s the same as if these transistors are hanging in the air when the current is pumping, the instantaneous power dissipation is zero!! I don’t even want to comment on everything else, it’s not serious. This craft has a place on the souvenir shelf; you can show it off to friends while drunk, pretending to be Nikola Tesla.
- Answer
#15 Guest Victor Guests 23 July 2020 20:33
2

Field transistors switch very well in a parallel circuit! Since the resistance of the source-drain junction has a directly proportional dependence on temperature, several parallel transistors adjust to each other, forming an ideal option!
- Answer
#16 Abgamych Guests 27 July 2020 10:33
3

Is it possible to place mosfits in series to increase voltage?
- Answer
#17 Dmitriy Guests 5 August 2020 13:18
4

Firstly, transistors cannot be connected in this way. When switching direct current or very low frequency, it will work for some time, but not for long. Second: in essence, the design excludes normal cooling, i.e. under real load it will burn out almost instantly. Third: a printed circuit board and pins of this design, in principle, cannot provide an equal load on individual keys.
In short, the result was a working prototype of a transistor in a large package. Not for work, but to show.
- Answer