# Switching high currents

#### lumley32

Jul 9, 2012
14
hello! not sure if this is in the correct place but.....

I am trying to come up with a way of switching a high current load.

I have an electromagnet that I want to pulse on for around 60ms, it draws around 90A at 12V DC. it will only switch once every few seconds at the most.

I have looked at SSR's and it seems that I need one made for DC current, not zero switching. my question is do I need one rated for the full 90A or as its a short pulse can I get away with running a lower rated unit?

Any other suggestions?

#### Tha fios agaibh

Aug 11, 2014
2,239
It draws around 90A at 12V DC.
I guess the question is; What does it draw in 60ms?

I would be looking for a SCR to drive it.

I don't follow. Why only pulse an electromagnet for only 60ms every few seconds?
I don't see it delivering enough current for the magnetic field to build.

#### Harald Kapp

##### Moderator
Moderator
Nov 17, 2011
12,516
I would be looking for a SCR to drive it.
Hard to turn off an SCR while 90 A of DC flow through.
A MOSFET is much easier to control, e.g. an FD86367.

it draws around 90A at 12V DC
With respect to @Tha fios agaibh 's comment
I don't see it delivering enough current for the magnetic field to build.
When you apply 12 V to an inductor you'll see a rising current until the current is limited (hopefully) by the resistance of the inductor. I guess 90 A is the steady state current when you apply 12 V DC permanently to the coil. For switching the coil voltage with 60 ms pulses the max. current may be much less, depending on the inductance and the resistance of the coil.

Note also that an inductor when turned off will create a high flyback voltage which needs to be limited to protect the switch (MOSFET) from destruction.

#### Tha fios agaibh

Aug 11, 2014
2,239
Hard to turn off an SCR while 90 A of DC flow through.....flyback voltage which needs to be limited to protect the switch (MOSFET) from destruction.
Yes, I was also assuming 90A steady state current, so @ 60ms the current may only be a few amps or so?
I don't see a problem switching with a SCR but not nearly as quickly.
The irony is, a blocking diode to protect a mosfet from flyback will also slow down the coil from de-energizing.

#### WHONOES

May 20, 2017
1,215
A thyristor or SCR (they are the same thing) is not of much use for switching DC current. SCR's have a latching current, generally quite a small percentage of their carry current. Once the carry current has exceeded that, it will stay switched on regardless of what you do to the trigger terminal. The only way to turn them off is to reduce the current in the circuit to less than the latch current or encourage the voltage on the anode to be less than that on the cathode at which point it becomes reverse biased and current stops flowing.
A very large FET or IGBT would be preferable. If you can measure the inductance of the coil then you can calculate the current rise time to see if it reaches 90A in 60ms.
You are right regarding the need for a fly-back diode across the coil which will need a suitably large current rating. You are also correct in saying that the coil will hold up with a fly-back diode in parallel this is because the diode keeps the coil current present from the collapsing magnetic field circulating in the coil.

#### Minder

Apr 24, 2015
3,239
This may help in the selection.
M.

#### Attachments

• IGBT_Mosfet.pdf
51.5 KB · Views: 48

#### Hopup

Jul 5, 2015
253
You can get very big used industrial IGBT's from Ebay at very cheap prices. IGBT's should be best way to do it as far as I can see.

#### kellys_eye

Jun 25, 2010
4,775
You can also parallel a load of (low Rds(on)) MOSFETs to get mahoosive current-handling capability.

#### Harald Kapp

##### Moderator
Moderator
Nov 17, 2011
12,516
FD86367 is good for 100 A continuously - provided a good heatsink is used.
Rdson is 3.3 mΩ, causing power dissipation of 26 W at 90 A.
Thermal resistance Rthjunction-case is 0.66 °C/W. Add a heatsink with e.g. Rth = 2 °C/W plus some leeway for mounting etc. you arrive at a total Rth of ~ 3 °C/W. 26 W will drop ~80 °C across this combined Rth. Assuming a max. junction temperature of 160 °C (the datasheet gives no exact value, only from the derating curves one can assume up to 180 °C would be allowed) this combination will work up to ambient temperatures of 80 °C.
As the transistor is only pulsed there will be some headroom left.

Note that a snubbber circuit may be applicable to protect the MOSFET from transients.

#### WHONOES

May 20, 2017
1,215
FD86367 is good for 100 A continuously - provided a good heatsink is used.
Rdson is 3.3 mΩ, causing power dissipation of 26 W at 90 A.
Thermal resistance Rthjunction-case is 0.66 °C/W. Add a heatsink with e.g. Rth = 2 °C/W plus some leeway for mounting etc. you arrive at a total Rth of ~ 3 °C/W. 26 W will drop ~80 °C across this combined Rth. Assuming a max. junction temperature of 160 °C (the datasheet gives no exact value, only from the derating curves one can assume up to 180 °C would be allowed) this combination will work up to ambient temperatures of 80 °C.
As the transistor is only pulsed there will be some headroom left.

Note that a snubbber circuit may be applicable to protect the MOSFET from transients.
Don't forget that Rds will increase as the device warms up. Your chosen device datasheet should indicate by how much.