Problem with a MC34063: Can't Handle Any Load

[nathanpc]

I was testing a simple boost circuit based on the MC34063 to boost a 9V battery to 15V (100mA max). Here are the values that I've used:

L = 100uH, Ct = 200pF, Co = 47uF



Without any load I'm getting 15V, but as soon as I try to load it with a 1k resistor the output drops to 3V. Any ideas on why this could be happening?

[filip_cro]

Are you sure that you connected everything right? Are you sure that you are using 1k and on the right place?
You are using breadboard?

[nathanpc]

If I didn't connect everything right I wouldn't be getting 15V without load, correct? I'm putting the 1k resistor in Vout. I'm using a breadboard. Also, I'm using a brand new 9V battery to test the circuit.

[filip_cro]

0.76 ohm may be 76 instead and be limiting current to much lower value. Second, breadboards are known to not work good at higher frequency, and this may cause problems.

[wraper]

What inductor do you use? If it is small axial one or something like this, it will saturate.

[nathanpc]

Tried with 470uF, exactly the same result. I'm using 2x 1.5 ohm in parallel, I rechecked, they are 1.5. I'm using this inductor: http://www.digikey.com/product-detail/en/RLB0914-101KL/RLB0914-101KL-ND/2352772

[AndyC_772]

It's probably a layout issue to do with the current sense resistor. Breadboards are horrible when it comes to resistance and, especially, inductance.

I'd bet if you probe pin 7 with a scope, you won't see the nice, clean voltage ramp that the device requires, but instead you'll be seeing spikes and noise that will stop the closed loop feedback circuit from working properly. The more current flowing in the main loop, the worse those spikes will be, and the worse the error.

There's nothing for it but to improve the layout: very short wires, solid ground plane, all the usual advice applies. Ditch the breadboard, though.

[wraper]

Should work. Problem is likely breadbord or wrong wiring. Make it at least on proto board with connections as short as possible. Or make something like this:

[mariush]

Make sure the 9v battery can actually supply the peak currents. Test the circuit from a bench power supply or a 7.5v wall adapter.

Keep in mind 9v batteries will be dead at about 6.5v and full at about 10-10.5v so the circuit should go as low as 7.5-8v if you really want to power something from a 9v battery.

I have doubts a 9v battery can actually give you 15v 100mA   ... 15v x 100mA  is ideally 9v 160mA ... but you'll have probably peaks of 250-300mA pulled from battery.  Most 9v batteries can probably only do about 40-60mA before seriously dropping voltage

See datasheets : 

http://data.energizer.com/PDFs/522.pdf
http://users.tpg.com.au/users/p8king/files/va_9v.pdf

[c4757p]

Should work. Problem is likely breadbord or wrong wiring. Make it at least on proto board with connections as short as possible. Or make something like this:

Just don't use the same inductor I used...

[codeboy2k]

Should work. Problem is likely breadbord or wrong wiring. Make it at least on proto board with connections as short as possible. Or make something like this:

Just don't use the same inductor I used...

but I really like how you made the IC into a convertible.. top down, silicon flapping in the breeze

[c4757p]

Well, it sure did cool down once the top came off! (Before the top came off, it made a damn good simulation of the Big Bang...)

[matkar]

I don't know if this has any influence in boost mode, but it sure did in buck mode. I had a similar problem once. Power supply was not able to deliver calculated current. The reason was the diode. I firstly mounted a 1N4007 (or something similar). After changing with a schottky diode, everything was fine. I didn't think it will make such a difference. What diode do you use?

[nathanpc]

I used a 1N5817