Hi,

I'd like to power a 24V Fan from a 12V power supply line.

According to the label, the fan draws 0.27A at normal operation, but I'm assuming it's probably more than that on startup.

To run the fan at 24V, I breadboarded the following dc-dc step-up converter circuit and it looks to be working fine, allowing me to tweak the output voltage ±2V relative to the target 24V with the potentiometer (VR1).



Nevertheless, given that the fan is rated at 24V max, I'd also like to limit the output voltage so that it doesn't exceed this upper threshold.

In other words, in the final circuit I'd still like to be able to tweak the output voltage a bit, yet to have that ability limited with an upper-bound of 24V.

After Googling various alternatives, I'm considering adding a 24V/2W zener diode (DZ2424000L) in reverse bias on the output power line (marked as D5 in the above schematics, but currently not installed in the actual circuit).

However, I don't know if this is the optimal (or even correct) way of going about this.

So, my questions in this context are:

1) Is this indeed the correct way of achieving the goal of limiting the output voltage to 24V or is there a better way?

2) Would the 2W rating of the particular zener diode I plan on using high enough for this purpose?

3) Any additional advice on improving the circuit would be most appreciated

Thank you very much in advance for your time & advice,
SnowCrash

 

With the zener connected, if the output voltage is adjusted to more than 24V then the chips will supply the maximum current they can into the zener in an attempt to raise the output voltage higher.

What would be better is to connect a 22V zener in series with an ordinary diode (to get a total drop of 22.7V) between the output and the FB pin, Now once the output voltage is above 23.95V (approximately - 22V Zener + 0.7V diode + 1.25V FB voltage) the extra feedback via the diodes will limit the output.

 

You could put a zener diode across the fan. But what is going to protect the zener diode in the event of over-voltage or if the fan fails or is disconnected? The only thing that will prevent the zener diode from blowing is the internal resistance of the supply plus that of the zener.

You could install a 24V 1000W zener but that would be impractical.

A zener works in circuit with a current limiting resistor in series with the supply if you know the maximum over-voltage and the maximum current to be supplied.

As an example, if the max supply voltage is 26V and max current is 1A, then the series resistor is (26V - 24V)/1A = 2Ω @ 2W.
The zener has to be rated above 25W and the resistor above 2W.

One drawback is if the supply is below 24V, the zener will conduct no current but the resistor will drop the voltage to the fan by 2V @ 1A load or 0.5V @ 0.25A load.

What's a practical solution? You could install a 24V zener with the largest wattage rating that is practical and hope it doesn't blow.
Just my 2 cents worth.

 

Must be some expensive fan if you're worried about protecting it...

Some suggestions

• PTC fuse
• Regular fuse
• SCR crowbar with companion fuse that will blow when the SCR is triggered

Using a zener in the proposed emulates a poor TVS. TVS diodes are designed to carry significant current for brief periods of time; zeners aren't. You could use a TVS diode with a fuse that will blow from over current.

 

Albert's solution in post #2 seems to be the best.

 

Firstly, thank you very much to all who commented!

What would be better is to connect a 22V zener in series with an ordinary diode (to get a total drop of 22.7V) between the output and the FB pin, Now once the output voltage is above 23.95V (approximately - 22V Zener + 0.7V diode + 1.25V FB voltage) the extra feedback via the diodes will limit the output.

AlbertHall, your suggestion is brilliant and looks like exactly what I've been looking for! Much appreciated! I just hope I understood what you meant in terms of how to connect the 22V Zener and regular diode between the output and the feedback pin on the IC.

Is the following correct?



Also, a follow-up question if I may:

In the zener + regular diodes setup, is it possible to use a 22V zener with a lower power rating than 2W? (I've tried figuring out the way to calculate how much power dissipation would be needed but it all got a bit confusing...).

Thanks again!
SnowCrash

 

You still need the original resistors and pot on FB if you want it be adjustable. The diodes just set a maximum limit on the voltage. The zener can be low power and the diode can be a 1N4148 (any small diode).

 

You still need the original resistors and pot on FB if you want it be adjustable. The diodes just set a maximum limit on the voltage. The zener can be low power and the diode can be a 1N4148 (any small diode).

Thanks AlbertHall!

I'll order a few 22V zener diodes and see how goes.

 

Having recently built a small fan controller, this caught my attention. You probably don't have a data sheet for the fan but, if you do, you might find that 26V is not out of spec for the fan's motor. I went through about a dozen 24V fan datasheets in selecting mine and most gave a voltage range where 26V was not out of spec. Here's the ds for the one I bought. As you can see, that one would work fine at 26V. It looks like they have a 15% margin.

Also, 26V is only about 8% over the nominal spec so I suspect that most, if not all, 24V fans would work just fine at that voltage.

 

Having recently built a small fan controller, this caught my attention. You probably don't have a data sheet for the fan but, if you do, you might find that 26V is not out of spec for the fan's motor. I went through about a dozen 24V fan datasheets in selecting mine and most gave a voltage range where 26V was not out of spec. Here's the ds for the one I bought. As you can see, that one would work fine at 26V. It looks like they have a 15% margin.

Thanks for the input, philba!

You make a good point. The fan I've got is a decent NIdec Bestial D12E-24PH (38mm thickness/2700RPM with good air flow and pressure), and indeed I only went by the label on its back and didn't even consider looking at the datasheet... I've googled it now and only found an online spec on Nidec's website, as well as a rather abbreviated sort of datasheet that contains even less details than the spec. Interestingly, as you rightly suggested, it looks like my Nidec does indeed support voltage inputs ranging roughly between 20V - 26V, but this isn't mentioned/explained explicitly anywhere - just this range appearing next to the official 24V rating in the online spec. Nevertheless, it's certainly good to know

That said, in these kinds of mini-projects my intention is not only to make something useful (for myself at least), but also to learn new ways of getting things done. It always bugs me to have a question like the one I posted here float around in my head and not knowing the answer - hence, I'm grateful for AlbertHall's illuminating solution and everyone else - including yourself - who took part in the conversation.