Emitter-follower circuit for variable fan voltage control

[galapogos]

Hi,
I'm interested in building an emitter-follower circuit for variable fan voltage control like the ones shown in the Simply Transistors page, since it's easier and cheaper than using any sort of voltage regulator.

Are there any problems with using the single transistor circuit rather than a darlington, and using a higher powered NPN transistor such as the TIP31? This one seems to be the easiest to build and loses only 0.7V. With a TIP31, what would the maximum fan current rating before I damage the potentiometer tracks?

With a voltage drop of 0.7V, would I need any sort of boost regulator to boost the voltage back to 12V? If not, wouldn't the voltage control near 12V be non-linear and clip to 11.3V?

[cpemma]

With a TIP31, what would the maximum fan current rating before I damage the potentiometer tracks?

Swinging the question round, 16mm lin pots are commonly 0.2W, so a 1K will take 14mA, though with 12V across and the extra 1k series resistor shown to limit bottom end voltage, it would only be carrying 6mA in the linked circuit.

If your transistor base current demand went to the pot limit the max fan current would be around Hfe x 8mA, so a low-gain TIP31 would be less use than a medium-gain 2N2222A - and control would be very non-linear as you turned the pot.

If you can find the transistor, a KTD882 or KSD882 look to be far better than the TIP31, claimed 3A and typical gain well over 100.

I've seen another method using a pot unterminated at the ground end so it's just a current-limiting variable base resistor (add a 820R fixed in series for safety). Then you could get the full 14mA base current for a higher fan current, but again very non-linear.

Low-dropout regulators are to best way to get (well) over 11.3V at >500mA current levels.

[StarfishChris]

What's the big problem with 11.3? If you really need the little extra CFM you get at 12V then maybe you're looking at the wrong fan. For silence purposes 12V is too loud for most people, even with Nexus fans.

[galapogos]

Thanks. I guess I can live with 11.3, but I'm just wondering if the upper range of the voltage is clipped, or is the entire voltage range shifted down 0.7V, i.e. 6.3-11.3V rather than 7-12V for example.

I don't have access to KTD882/KSD882 transistors, but I do have access to 2N3804/TIP32C, so I guess I'll build the NPN/PNP circuit for a 0.7V drop

I have a couple of trimpots, but I don't know their specs, and haven't measured their resistance yet. 2 of them are MOUSER units and have A10K stamped on them, so they're probably 10Kohm units. The actual resistance doesn't matter right? Only the ratio of the 2, since it's a voltage divider?

Here's a pic of the pot:

[cpemma]

The actual resistance doesn't matter right? Only the ratio of the 2, since it's a voltage divider?

One thing schools and most textbooks gloss over with the simple potential-divider circuit is that as soon as you apply a load to it that's not a 100Gigohm meter, it affects the divider split (as a parallel resistance to the bottom half).

With the recommended 1k pot, the base current taken is small compared to the divider current, so it gives decent linearity.

Also, the A10K means it's a 10k log (audio) track, you want a B1K linear track.

[Gorsnak]

If you want a 12v option, it's easy enough to wire in a switch that will bypass the circuitry and feed the fan straight off the 12v rail.