PWM Parallel Power Transistors

SgtWookie

Joined Jul 17, 2007
22,230
Could you show me how that would be wired?

I have also attached a pic of the circuit layout complete with the PHAT tracings. Let me know if the layout is correct?
I'm still looking at this.

Are you going to attempt to make this board yourself?

You'll need 2 Oz/ft^2 copper minimum.

R4 will have to be a trace on the board, not a physical resistor.
An 0.4" (400 mil) wide trace 3.1" long of 2oz/ft^2 copper will have a resistance of 2.01 milliOhms, which will be just about right for this. A careful layout will minimize parasitics. I guess you need someone to do that for you, as the software you're using just isn't as capable as the Eagle Layout Editor.

I'll look at it when I have a bit of time.
 

cat3rn

Joined Jun 3, 2008
117
I'm still looking at this.

Are you going to attempt to make this board yourself?

You'll need 2 Oz/ft^2 copper minimum.

R4 will have to be a trace on the board, not a physical resistor.
An 0.4" (400 mil) wide trace 3.1" long of 2oz/ft^2 copper will have a resistance of 2.01 milliOhms, which will be just about right for this. A careful layout will minimize parasitics. I guess you need someone to do that for you, as the software you're using just isn't as capable as the Eagle Layout Editor.
I was planning on making the board mself but as of this moment I only have 1oz Copper clad in my possession.

I found a friend that has CM2000 he is going to bring it over tomorrow.
 

cat3rn

Joined Jun 3, 2008
117
Where it's shown in the schematics published on the site that I've already posted.

No. 222nH = 0.222uH. You need around 220uH, or 1,000 times as much.


You need magnet wire and a form to wind it on.

Basically, a spool. You can make a spool out of almost anything.

220uH is a fairly large value of inductance. You can make a 220 uH inductor that would be suitable for this project by using a spool that was 1" in diameter with flanges 1/2 inches apart, and winding on 81 turns of AWG 20 magnet wire. That'll take about 26 feet of wire. The outside diameter of the inductor will be about 1.4".

Of course, if you're trying to use anything different than what I've specified above, your mileage will vary - and considerably.

However, if you provide dimensions of a spool that you might have available, along with what AWG size of magnet wire that you have, I could tell you how many turns that you will need.

Basically, the larger the diameter of the spool, the higher the inductance.
The smaller the distance between the flanges, the higher the inductance.
Hey I have come to the conclusion that I need to send you my cell for testing it will be yours to keep. I am building a new cell that will be 8" X 8" with 25 plates. I need to have someone using my cell to help build a buck topology converter to regulate the current and make this a better project.

Where do I ship it to. pm me.

Bill
 

DickCappels

Joined Aug 21, 2008
10,152
One other detail...when you parallel them, use seprate gate damping resistors. If you were to try to us the same resistor, each gate will see the input capacitance of the other transistor.
 
Last edited:

John Luciani

Joined Apr 3, 2007
475
This circuit is a programmable current-sink. You will not be able to reduce the amount
of heat just spread the heat between multiple FETs. Are you using this a switch or
a resistor?

If you are using this as a resistor then the LM324 is fine provided that you do not
need higher currents than you can get with a resistance of Rds(on) + 0.0015Ohms. The Rds(on) is limited by the voltage swing of the LM324. If you do not need lower resistances than I wouldn't worry about it.

Do not parallel two FETs. The transconductance mismatch between devices can
be a problem especially at high Vds values. You will get uneven power dissipation
between devices. You need to add a second control loop.

If you are doing a new load I would get rid of the LM324 and get an op-amp with
better offset. Also using a +15V supply will give you higher current capability
(if you need it). A better op-amp will also let you get rid of the shunt amplifier circuit and
save you space. Using a small negative supply will let you completely turn
off the FET (if you need to).

You should be able to do a simple circuit like I have at
http://preview.tinyurl.com/6cbn6h (near the bottom)

(* jcl *)
 

KL7AJ

Joined Nov 4, 2008
2,229
As i can see you are using MOS so can easily parallel another one on it because MOS have a positive temperature coefficient thus the current will distribute equally on both of them. If one gets hot its resistance increases so less current pass through it and more through the other. This goes one until both have equal current flowing through them. Just take another one MOS and connect its source the the source of the existing one, its drain to the drain and its gate to the gate. Note that the turn on time will double because the input capacitance doubles. Reduce the 100ohm resistor to 50ohm to increase the turn on time. Another option is to drive them with an emitter follower to increase the turn on time.

Mik3

I'm glad you mention the temperature profile. I remember trying to balance huge banks of 2n3055s. By the time you got the equalizing resistors all tweaked, you usually ended up dissipating so much power you lost ground!

Eric
 
Top