Troubleshooting a fan speed controller UPDATE

Thread Starter

Tonyr1084

Joined Sep 24, 2015
6,263
Yeah, perhaps a PNP would work better. 2N3906 should do. Move the resistor between sink and the transistor base. If the IGBT doesn't turn off fast enough I can add a resistor to pull it down. Thanks.

I would like to learn how to calculate the needed resistance values.
 

MisterBill2

Joined Jan 23, 2018
9,267
The circuit should work, the drawing is OK. Some folks just don't get it, they may NEVER get it. But as I looked at the spec sheet I see that the saturation voltage is about 3 volts, which is a big portion of that 12 volt supply. So you do need a serious heat sink 3 volts x 30 amps is 90 watts of heat, and that can lead to a warm transistor in a hurry.
 

Wolframore

Joined Jan 21, 2019
2,342
yep it needed a pull down. Now it makes sense but driving it with a PNP is an improvement, you don’t need or want a pull down with a PNP You will need a pull up at the emitter of the PNP. You don’t need much current to switch the gate, start with a uA or two. It’ll be fine.

the gate of the IGBT has capacitance without the pull down, as drawn (with the NPN), the IGBT will not completely turn off when the BJT stops conducting.
 
Last edited:

MisterBill2

Joined Jan 23, 2018
9,267
yep it needed a pull down. Now it makes sense but driving it with a PNP is an improvement, you don’t need or want a pull down with a PNP You will need a pull up at the emitter of the PNP. You don’t need much current to switch the gate, start with a uA or two. It’ll be fine.

the gate of the IGBT has capacitance without the pull down, as drawn (with the NPN), the IGBT will not completely turn off when the BJT stops conducting.
At the current levels that the IGBT will be operating in this application switching time, that interval in the linear region, will be rather important. Thus either serious heat sinking or fast switching will be important. And I am still concerned about that 3 volt VceSAT value. I may not be reading the data sheet right, that is what I think that I saw.
 

Alec_t

Joined Sep 17, 2013
12,201
I agree that you need a pull-down resistor (or transistor) from the IGBT gate to ground in the circuit shown. Without it there is nothing to turn the IGBT off (other than leakage current), so the PWM will be ineffective.
 
Last edited:

shortbus

Joined Sep 30, 2009
8,805
@shortbus I don't recall stating clockwise or counterclockwise. Maybe I did. Nevertheless, for the sake of clarification; CW = brightening CCW = dimming. However, because of the way the circuit (post #1) is set up - I would expect that (eventually) the motor will spin faster with CCW. I think.

[edit] OK, I finished reviewing all my posts. I didn't see anywhere where I said clockwise or counterclockwise. Maybe I missed something. But I don't think I ever said that.
I got the direction from the inference in post #36.
 

Thread Starter

Tonyr1084

Joined Sep 24, 2015
6,263
Following up on recommendations; redrawn the schematic with a PNP transistor. To be honest I don't know what I'm doing. I believe the circuit, in theory, is workable. However, I have no idea how to calculate the correct resistance.

Go ahead and be critical. I learn nothing if nobody points out my mistakes and shows me "how to fish". I'm not looking for someone to give me the right answers, I want to know how to get them for myself.

PNP Variable Blower.jpg
 

Chris65536

Joined Nov 11, 2019
270
The Emitter of the PNP should go straight to +12v. The Collector goes to the Gate of the IGBT. And a 1K resistor from the Gate to ground. The Base resistor could also be much larger - maybe 10K.

Honestly though, your first circuit should have worked too. Maybe double check the pinout on the IGBT?
 
Last edited:

Thread Starter

Tonyr1084

Joined Sep 24, 2015
6,263
@Chris65536 Thanks. The pinout on the iggy is verified. Tested it for functionality. The gate is definitely the gate.

Now it's time to add the new circuitry and give it a test. Again, to those who worry a lot - I'm testing with an auto bulb. If it can do that then it can control the speed of the blower.
 

Wolframore

Joined Jan 21, 2019
2,342
Ok keep us updated, sometimes you just need to plug it in before we work through It... don’t take lunch until you get it working today :p
 

MisterBill2

Joined Jan 23, 2018
9,267
To maintain the logic sense the circuit is correct as drawn. The on time of the PWM is when the output of the controller is SINKING current, but the on mode of the IGBT is when the gate is pulled positive relative to the emitter. At least that is what I think I see. In that case the circuit as shown will work, but in reverse. So the advice in post #72 will provide the opposite operation, but I would use 1K instead of ten K for the base resistor and the resistor moved to the collector side of the 2N3906.
 

shortbus

Joined Sep 30, 2009
8,805
@Tonyr1084, why not substitute the mosfet inside the PWM module with your IGBT??? If your so worried about the mosfet not holding up, which is a wrong way to think.:) The mosfet will be quite happy doing it though.

Where did you come up with the amp readings on the blower? If it is from the fuse ratting in the car for it, I kind of think it isn't going to be the same as the way your using the blower. In the car it was moving air through a restriction, where as with no restriction the amperage will be much lower.
 

Thread Starter

Tonyr1084

Joined Sep 24, 2015
6,263
@shortbus I've already explained that the drain current for that FET is 25 amps. 100 amps peak at a duration of 10µS or less. (see post #19 where @Dodgydave posted a data sheet) There is no heat sink inside the PWM module, no more than just the spade lug it's riveted to. On top of that the copper traces on the board do not look like they can handle all that current. And why risk blowing up my PWM dimmer? If I blow up a scrap IGBT - so what. Cost me nothing. And in the event of a failure - I can redesign and rebuild and still have the dimmer in good working order.

As for the current measurements, they were taken in actual - unloaded conditions. The blower is a centrifugal compressor type blower. It blows air into the exhaust stream of a running engine. So it can't just blow, it has to push the air into the exhaust system. I won't be using it under such heavy demands. Hence, the reason for slowing the blower down. It's going to stoke a fire with fresh air. The fire will burn hotter and cleaner.

Why do I want to do that? Because I want to. I want to get more heat energy out of an outdoor fireplace. Hotter fires radiate more energy. There's another possible use for it - I may make a heat extractor that goes into the fireplace. I need air flowing through the plumbing drawing heat out of the fire and putting it into the room. But that's not so likely to happen. Still, there are reasons for wanting to control the amount of air moving through the system, which ever system I build. I MAY EVEN BUILD A SECOND UNIT for melting ice from the porch and driveway. Why? Because. Sometimes we don't build to solve a problem, sometimes we build because we can.
 

shortbus

Joined Sep 30, 2009
8,805
I agree with @shortbus for 12v mosfets are more efficient. I did find references to similar current readings for blower motors At least for running condition.
But most blowers are under some type of restriction, but he is going to use this blowing into open air little or no restriction. Fuses in cars many(most?) times control more than one thing and are rated for all things being on at the same time. A better way to tell is to look at the motor leads, and figure the amp handling capacity of the wire.
 
Top