Quick PWM fan connection question

So I’m connecting two Delta PWM computer fans to one PWM pin on a Photon in parallel. The Delta tech specs seem to indicate they should be fine with a 3.3V signal, and they do indeed work. The question I have is whether or not this is a safe long term solution for the Photon? That is to say do I need any protection between the Photon and the fans?

I’ve had some odd crashing that I believe was 100% related to the fans drawing more current than the power supply could handle (was using 24V fans and a Meanwell combo power supply that was 5V and 24V, but I think I’ve discerned that I was just above the current capability of the 24V side when the fans were at full bore and that was causing the 5V to dip). But I wanted to make sure I wasn’t also possibly causing a problem with that direct connection.

—Donnie

How much current are those fans pulling? I believe the Photon can only sink up to 25mA on a single GPIO pin. I’d imagine that those two fans are going to pull more than that. You need a MOSFET or some other mechanism to switch the power to the fans.

2 Likes

These fans have a PWM control signal wire. They get their power from separate power lines and don’t draw any significant current on the signal wire that I’m aware of. I suppose it’s possible that two is one too many, so I guess I should look to be sure.

But to be clear, I am not driving motors directly with a Photon.

—Donnie

1 Like

Good to know. It’s probably okay then if it’s just signal.

Hi @djb_rh

I am with @picsil on this one. I would most certainly make use off a MOSFET or NPN transistor between the Photon and any device drawing ‘relatively significant current’ to what the pin can deliver. I have 240V/20A relays on couple of devices and even though their coil power rating fall within range, I always put NPN transistors in-between as I have seen some intermitted odd behaviour.

Not sure if it is needed, but small capacitor and diode can’t hurt either :wink:

Hope this helps!

Regards,
Friedl.

Folks, the PWM input is most likely a high impedance, very low current control input. As such, the Photon GPIO pin doesn’t need any conditioning. You could, if you wish, put a small (200 ohm) resistor in series to limit the current but I don’t believe that’s necessary.

What you should focus on is the 24v/5v power supply which seems to be struggling when the fans are pulling a high current. Specifically, the 5v supply dipping is not good. You can either limit the amount of current going to the fans, or get a better power supply. Even a large 1000uF capacitor on the 5v supply won’t prevent the voltage dip if your fans are constantly drawing too much current.

One way to “limit” the current is to prevent your PMW output from driving the fans to the speed at which they draw too much current. This is like a software speed “limiter” on cars.

3 Likes

So to add a little more, first I had two 24V 1.05A fans. The PWM is 0-255 in range. At first I let them get to full speed when necessary and it worked for a while that way (it’s a night time air conditioner, basically). Then it started flaking out late morning where the thing would just lock up. I could see the last LCD update and knew every time this happened was at full bore and had suspicions.

So I dialed the max back to 235. And that seemed to help for a week or two, then it went back to dying late morning when it was at peak.

The power supply had plenty of 5V current, but was actually a 24V 2A supply. As you can see from above, at full speed I should have been drawing more like 2.1A, which is bad. But before I deployed this I had put a current meter on it and never saw more than about 1.8A total (either during acceleration to full speed nor while running at full speed).

But once this started happening at 235, I pulled it out and first thing I did was measure voltage at 235. It was dropping from 24V at under 100 to about 22.0V at 235, which means the current was likely climbing will above 2A. So I honestly think the 24V section was failing on that power supply.

So I ordered a new 5V/24V supply with a 4A 24V section. I installed it, but it was WORSE in that it was bouncing voltage all over the place when loaded or not.

So I’ve dropped back to a 5V/12V supply that does 3A on the 5V side with 12V 1.6A fans and strangely this setup works fine up to 255 with no issues whatsoever. Obviously I feel a little worried about this setup, too, but I plan to try yet another 24V supply for the 24V fans because they are a little higher output.

Anyway, just more to chew on.

–Donnie

1 Like

FWIW, I looked at the data sheet for the Delta PWM controlled fans and the current draw is listed as “less than 5mA.” So yeah, it looks like you can parallel four or maybe more off one Photon output pin with no buffer.

So since my last report, I noticed that the 12V fans definitely don’t have the output that the 24V do (not surprising if you look at the current draw of each) and cooling was noticeably affected. I have, sadly, also had one lockup. So I ordered a couple 5V power supplies and a separate 24V supply.

Today I swapped the 24V fans back into the system and replaced the single power supply with two independent ones. I powered it up with my voltmeter on the 24V lines and bam, this supply (a 3A one) is solid with the fans running full speed. DOES NOT DIP AT ALL. Sadly, I cost myself a lot of work by forgetting this time to tie the ground lines of the two power supplies together, and if you do that all sorts of weird things happen on the signal from the Photon to the fans. It’s strange that I did that because not only do I know that needs to happen, on the combo power supplies they actually do have separate ground posts by each power terminal and I had ohmed those out just to be SURE they were tied together internally. And then I forgot to tie them together with these two external supplies.

Anyway, fixed that and now I’ve got big output again. Fingers crossed that this fixes the lockups, because if it doesn’t I have no idea how to troubleshoot that.

–Donnie

3 Likes