MC3406A Problem

[aaronsnoswell]

Hi all.

I'm trying to develop a portable fume extractor, based on the design at http://jumperone.com/2011/07/portable-fume-extractor/.

I've come up with the below circuit, and have built it in a breadboard to test. When I put in 5V input with no output load, I get 13.4V output, as expected, but when I attach the computer fan I want to drive, the output voltage drops to 8V. I kind of expected this would happen, but I don't understand why. As far as I can tell, I followed the design guidelines for the MC3406A correctly. What's going on?

[tszaboo]

We are going to need the load current, type of L and D1. The title is wrong. Also, try increasing CT maybe it is just a frequency problem.

[aaronsnoswell]

Hi! Thanks for your comments.

The fan I'm using has a label indicating it runs at 12V, 2.6W, giving 220mA current. I checked with my meter and it was 100mA when running, I assume the 220mA is a peak current. 220mA is what I used in the design formulae for Iout.

The inductor I'm using is one of a 'Colour Wheel Inductor' pack I ordered on eBay - see the picture. They are supposedly 0.5W, and the markings indicate 10% tolerance. Would the type of inductor winding method have an effect on this circuit?

The diode is a 1N5819 Schottky, which from datasheets should have a forward voltage drop of 0.6V (this is what I used in the design formulae).

The circuit was actually for an adjustable supply, but I swapped R2 and R1 out for non-adjustable resistors to simplify debugging and for when I posted this question.

Finally, I tried upping Ct with no noticeable effect. 330pF gives 7.6V, 470pF gives 7.7V and 820pF gave 7.1V.

Thanks again for your comments - I hope with this info you can make some more suggestions!

[free_electron]

those are chokes. not usable for SMPS operation. there is no magnetic mass...

[aaronsnoswell]

free_electron, thanks - what should I be using instead? A manually wound piece of wire? I don't have any ferrite core material on hand (nor an inductance meter to test with - although I could make a simple testing circuit using an Arduino).

The plot thickens - while doing more testing today I noticed that when the fan was running, and I probed the output leads, the meter was momentarily reading 13V, then dropping suddenly to 6-7V. It's an Extech instruments meter - one of the ones Dave recommended actually. So I think that the circuit is probably outputting 13V as it should, but my meter is unable to give me a good reading for some reason. Why would that be?

[tszaboo]

If you absolutlely must use through hole components, use something like Panasonic ELC10D330E, Würth 744772330, Murata 18R333C or similar. I think you got the picture. Also, breadboarding SMPS is not recommended.